U.S. patent number 7,202,776 [Application Number 11/034,325] was granted by the patent office on 2007-04-10 for method and system for detecting objects external to a vehicle.
This patent grant is currently assigned to Intelligent Technologies International, Inc.. Invention is credited to David S. Breed.
United States Patent |
7,202,776 |
Breed |
April 10, 2007 |
Method and system for detecting objects external to a vehicle
Abstract
Method and system for obtaining information about objects in the
environment outside of and around a vehicle and preventing
collisions involving the vehicle includes directing a laser beam
from the vehicle into the environment, receiving from an object in
the path of the laser beam a reflection of the laser beam at a
location on the vehicle, and analyzing the received laser beam
reflections to obtain information about the object from which the
laser beam is being reflected. Analysis of the laser beam
reflections preferably entails range gating the received laser beam
reflections to limit analysis of the received laser beam
reflections to only those received from an object within a defined
(distance) range such that objects at distances within the range
are isolated from surrounding objects.
Inventors: |
Breed; David S. (Boonton
Township, Morris County, NJ) |
Assignee: |
Intelligent Technologies
International, Inc. (Denville, NJ)
|
Family
ID: |
46303694 |
Appl.
No.: |
11/034,325 |
Filed: |
January 12, 2005 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20050134440 A1 |
Jun 23, 2005 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
10822445 |
Apr 12, 2004 |
7085637 |
|
|
|
10118858 |
Apr 9, 2002 |
6720920 |
|
|
|
09177041 |
Oct 22, 1998 |
6370475 |
|
|
|
10118858 |
Apr 9, 2002 |
6720920 |
|
|
|
09679317 |
Oct 4, 2000 |
6405132 |
|
|
|
09523559 |
Mar 10, 2000 |
|
|
|
|
09177041 |
Oct 22, 1998 |
6370475 |
|
|
|
09909466 |
Jul 19, 2001 |
6526352 |
|
|
|
60123882 |
Mar 11, 1999 |
|
|
|
|
60062729 |
Oct 22, 1997 |
|
|
|
|
Current U.S.
Class: |
340/435; 342/70;
342/118; 340/557; 701/514 |
Current CPC
Class: |
G01S
19/42 (20130101); G08G 1/167 (20130101); B60W
30/04 (20130101); G01S 17/86 (20200101); G08G
1/161 (20130101); G08G 1/165 (20130101); G05D
1/024 (20130101); B60N 2/2863 (20130101); B60W
50/00 (20130101); G08G 1/166 (20130101); B60W
30/18154 (20130101); G01S 19/50 (20130101); G01S
17/18 (20200101); G08G 1/163 (20130101); G08G
1/164 (20130101); G01S 17/89 (20130101); G01S
19/17 (20130101); B60W 30/16 (20130101); B60W
40/06 (20130101); G01S 17/931 (20200101); B60W
30/18172 (20130101); G01S 2013/9318 (20200101); B60W
2556/50 (20200201); G01S 2007/4975 (20130101); G01S
2013/9316 (20200101); G01S 2013/9329 (20200101); B60W
2050/0071 (20130101); B60W 40/076 (20130101); G01S
2013/93272 (20200101); G01S 2013/932 (20200101); G01S
2013/9322 (20200101); G01S 2013/93271 (20200101); B60W
10/18 (20130101); B60W 10/20 (20130101); G01S
2013/93273 (20200101); G05D 2201/0213 (20130101); G01S
2013/9315 (20200101); G01S 2013/9321 (20130101); G01S
2013/93185 (20200101); B60W 30/02 (20130101); B60W
50/14 (20130101); G01S 2013/9319 (20200101); B60W
2050/0073 (20130101); B60W 10/06 (20130101); B60W
30/14 (20130101); B60W 2050/0078 (20130101); G01S
19/41 (20130101); B60W 40/072 (20130101) |
Current International
Class: |
B60Q
1/00 (20060101) |
Field of
Search: |
;340/435,557,901,905,907,995.1 ;342/70,118,357.06 ;348/113,118
;701/200,207,213,223,51,58,59,65 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
0353200 |
|
Jan 1990 |
|
EP |
|
0750202 |
|
Dec 1996 |
|
EP |
|
2141890 |
|
Jan 1984 |
|
GB |
|
Other References
Ofer David et al., "Range gated active night vision system for
automobilies", Applied Optics, vol. 45, No. 28, Oct. 1, 2006, pp.
7248-7254. cited by other.
|
Primary Examiner: Mullen; Thomas
Attorney, Agent or Firm: Roffe; Brian
Parent Case Text
CROSS REFERENCE TO RELATED APPLICATIONS
This application is a continuation-in-part of U.S. patent
application Ser. No. 10/822,445 filed Apr. 12, 2004, now U.S. Pat.
No. 7,085,637, which is a continuation-in-part of:
1) U.S. patent application Ser. No. 10/118,858 filed Apr. 9, 2002,
now U.S. Pat. No. 6,720,920, which is: A) a continuation-in-part of
U.S. patent application Ser. No. 09/177,041 filed Oct. 22, 1998,
now U.S. Pat. No. 6,370,475, which claims priority under 35 U.S.C.
.sctn.119(e) of U.S. provisional patent application Ser. No.
60/062,729 filed Oct. 22, 1997; B) a continuation-in-part of U.S.
patent application Ser. No. 09/679,317 filed Oct. 4, 2000, now U.S.
Pat. No. 6,405,132, which is a continuation-in-part of U.S. patent
application Ser. No. 09/523,559 filed Mar. 10, 2000, now abandoned,
which claims priority under 35 U.S.C. .sctn.119(e) of U.S.
provisional patent application Ser. No. 60/123,882 filed Mar. 11,
1999, and which is a continuation-in-part of U.S. patent
application Ser. No. 09/177,041 filed Oct. 22, 1998, now U.S. Pat.
No. 6,370,475, which claims priority under 35 U.S.C. .sctn.119(e)
of U.S. provisional patent application Ser. No. 60/062,729 filed
Oct. 22, 1997; and C) a continuation-in-part of U.S. patent
application Ser. No. 09/909,466 filed Jul. 19, 2001, now U.S. Pat.
No. 6,526,352; and
2) U.S. patent application Ser. No. 10/216,633 filed Aug. 9, 2002,
now U.S. Pat. No. 6,768,944, which is a continuation-in-part of
U.S. patent application Ser. No. 10/118,858 filed Apr. 9, 2002, now
U.S. Pat. No. 6,720,920.
All of the above applications are incorporated by reference herein.
Claims
What is claimed is:
1. A method for obtaining information about objects in the
environment outside of and around a vehicle, comprising: directing
a laser beam from the vehicle into the environment; receiving from
an object in the path of the laser beam a reflection of the laser
beam at a location on the vehicle; forming at least one image from
the received laser beam reflections; and analyzing the at least one
image formed from the received laser beam reflections to obtain
information about the object from which the laser beam is being
reflected, the analyzing step comprising range gating the received
laser beam reflections to limit analysis of the at least one image
formed from the received laser beam reflections to only images
formed from laser beam reflections received from an object within a
defined range such that objects at distances within the range are
isolated from surrounding objects.
2. The method of claim 1, wherein the analyzing step obtains
information about the distance between the vehicle and the
object.
3. The method of claim 1, wherein the laser beam is infrared.
4. The method of claim 1, further comprising controlling the
direction of the laser beam.
5. The method of claim 1, further comprising: providing a digital
map including information relating to roads on which the vehicle
can travel or is traveling; and defining a field into which the
laser beam will be directed based on the map; the step of directing
the laser beam into the environment comprising directing the laser
beam into the defined field.
6. The method of claim 1, further comprising: scanning with the
laser beam at a high scanning speed; and scanning with an
additional laser beam at a slower scanning speed.
7. The method of claim 1, wherein the analyzing step further
comprises analyzing the at least one image formed from the received
laser beam reflections to detect the presence of objects
potentially affecting operation of the vehicle, the range gating
step being performed once the presence of each object is detected
and the range being determined to encompass any objects whose
presence has been detected.
8. The method of claim 7, wherein the analyzing step further
comprises narrowing the range such that laser beam reflections from
only the object whose presence is detected and other objects in the
same range are analyzed.
9. The method of claim 7, wherein the analysis of the at least one
image formed from the received laser beam reflections to detect the
presence of objects is performed using a pattern recognition
algorithm.
10. The method of claim 7, wherein the analyzing step further
comprises ascertaining the identity of or identifying each object
and proceeding to obtain information about the distance between the
object and the vehicle based on the identity or identification of
the object.
11. The method of claim 1, further comprising alerting a driver of
the vehicle if the information obtained about an object in the
environment outside of and around the vehicle indicates that a
collision with the object is about to occur.
12. The method of claim 1, wherein the laser beam reflections are
received by an image sensor which forms the at least one image.
13. A method for avoiding collisions between a vehicle and another
object, comprising: mounting a laser beam projector on the vehicle;
directing a laser beam from the projector outward from the vehicle;
determining whether an object is present in the path of the laser
beam based on reception of reflections of the laser beam caused by
the presence of the object in the path of the laser beam; when an
object is determined to be present, setting a distance range
including a distance between the vehicle and the object; processing
only received reflections of the laser beam emanating from objects
in the set distance range to determine whether each object may
impact the vehicle; and if a determination is made that the object
may impact the vehicle, effecting a countermeasure with a view
toward preventing the impact.
14. The method of claim 13, further comprising: providing a digital
map including information relating to roads on which the vehicle
can travel or is traveling; and defining a field into which the
laser beam will be directed based on the map; the step of directing
the laser beam outward from the vehicle comprising directing the
laser beam into the defined field.
15. The method of claim 13, further comprising: scanning with the
laser beam at a high scanning speed; and scanning with an
additional laser beam at a slower scanning speed.
16. The method of claim 13, wherein the processing step comprises
applying a pattern recognition technique to ascertain the identity
of or identify each object in the set distance range and assessing
the potential for and consequences of an impact between the vehicle
and the object based on the identity or identification of the
object.
17. The method of claim 13, wherein the countermeasure effected
entails alerting a driver of the vehicle about the possible impact
and/or altering the travel of the vehicle.
18. A system for avoiding collisions between a vehicle and another
object, comprising: a laser beam projector arranged on the vehicle
to direct a laser beam outward from the vehicle; a receiving unit
for receiving reflections of the laser beam which reflect off of
objects in the path of the laser beam; and a processor coupled to
said receiving unit and arranged to process any received
reflections to determine whether an object is present in the path
of the laser beam and when an object is determined to be present,
said processor being arranged to set a distance range including a
distance between the vehicle and the object, process only received
reflections of the laser beam emanating from objects in the set
distance range to determine whether each object may impact the
vehicle, and if a determination is made that the object may impact
the vehicle, cause a countermeasure to be effected with a view
toward preventing the impact.
19. The system of claim 18, wherein said processor includes a
pattern recognition algorithm which ascertains the identity of or
identifies each object in the set distance range and assesses the
potential for and consequences of an impact between the vehicle and
the object based on the identity or identification of the
object.
20. The system of claim 18, further comprising a driver
notification system or a vehicle control system, the countermeasure
caused by said processor being activation of said driver
notification system to alert the driver of the impending impact or
activation of said vehicle control system to vary the travel of the
vehicle to avoid the impending impact.
21. A method for obtaining information about objects in the
environment outside of and around a vehicle, comprising: providing
a digital map including information relating to roads on which the
vehicle can travel or is traveling; defining a field in the
environment into which a laser beam will be directed based on the
map; directing a laser beam from the vehicle into the defined
field; receiving from an object in the path of the laser beam a
reflection of the laser beam at a location on the vehicle; and
analyzing the received laser beam reflections to obtain information
about the object from which the laser beam is being reflected, the
analyzing step comprising range gating the received laser beam
reflections to limit analysis of the received laser beam
reflections to only those received from an object within a defined
range such that objects at distances within the range are isolated
from surrounding objects.
22. A method for obtaining information about objects in the
environment outside of and around a vehicle, comprising: directing
two laser beams from the vehicle into the environment; receiving
from an object in the path of the laser beams a reflection of the
laser beams at a location on the vehicle; and analyzing the
received laser beam reflections to obtain information about the
object from which the laser beam is being reflected, the directing
step comprising scanning with a first one of the laser beams at a
high scanning speed and scanning with a second one of the laser
beams at a slower scanning speed, the analyzing step comprising
range gating the received laser beam reflections to limit analysis
of the received laser beam reflections to only those received from
an object within a defined range such that objects at distances
within the range are isolated from surrounding objects.
23. A method for obtaining information about objects in the
environment outside of and around a vehicle, comprising: directing
a laser beam from the vehicle into the environment; receiving from
an object in the path of the laser beam a reflection of the laser
beam at a location on the vehicle; and analyzing the received laser
beam reflections to obtain information about the object from which
the laser beam is being reflected, the analyzing step comprising
analyzing the received laser beam reflections to detect the
presence of objects potentially affecting operation of the vehicle
and range gating the received laser beam reflections to limit
analysis of the received laser beam reflections to only those
received from an object within a defined range such that objects at
distances within the range are isolated from surrounding objects,
the range gating step being performed once the presence of each
object is detected and the range being determined to encompass any
objects whose presence has been detected.
24. The method of claim 23, wherein the analyzing step further
comprises narrowing the range such that laser beam reflections from
only the object whose presence is detected and other objects in the
same range are analyzed.
25. The method of claim 23, wherein analysis of the received laser
beam reflections to detect the presence of objects is performed
using a pattern recognition algorithm.
26. The method of claim 23, wherein the analyzing step further
comprises ascertaining the identity of or identifying each object
and proceeding to obtain information about the distance between the
object and the vehicle based on the identity or identification of
the object.
Description
FIELD OF THE INVENTION
This invention is in the fields of automobile safety, intelligent
highway safety systems, accident avoidance, accident elimination,
collision avoidance, blind spot detection, anticipatory sensing,
automatic vehicle control, intelligent cruise control, vehicle
navigation, vehicle-to-vehicle communication,
vehicle-to-non-vehicle communication and non-vehicle-to-vehicle
communication and other automobile, truck and train safety,
navigation, communication and control related fields.
The invention relates generally to methods for vehicle-to-vehicle
communication and communication between a vehicle and non-vehicles
and more particularly to apparatus and methods using coded spread
spectrum, ultrawideband, noise radar or similar technologies. The
coding scheme can use may be implemented using multiple access
communication methods analogous to frequency division multiple
access (FDMA), time division multiple access (TDMA), or code
division multiple access (CDMA) in a manner to permit simultaneous
communication with and between multiple vehicles generally without
the use of a carrier frequency.
The invention also relates generally to an apparatus and method for
precisely determining the location and orientation of a host
vehicle operating on a roadway and location of multiple moving or
fixed obstacles that represent potential collision hazards with the
host vehicle to thereby eliminate collisions with such hazards. In
the early stages of implementation of the apparatus and method and
when collisions with such hazards cannot be eliminated, the
apparatus and method will generate warning signals and possibly
initiate avoidance maneuvers to minimize the probability of a
collision and the consequences thereof. More particularly, the
invention relates to the use of a Global Positioning System
("GPS"), differential GPS ("DGPS"), other infrastructure-based
location aids, cameras, radar, laser radar, terahertz radar and an
inertial navigation system as the primary host vehicle and target
locating system with centimeter level accuracy. The invention is
further supplemented by a processor to detect, recognize and track
all relevant potential obstacles, including other vehicles,
pedestrians, animals, and other objects on or near the roadway.
More particularly, the invention further relates to the use of
centimeter-accurate maps for determining the location of the host
vehicle and obstacles on or adjacent the roadway. Even more
particularly, the invention further relates to an inter-vehicle and
vehicle-to-infrastructure communication systems for transmitting
GPS or DGPS position data, velocities, headings, as well as
relevant target data to other vehicles for information and control
action. The present invention still further relates to the use of
Kalman filters, neural networks, combination neural networks and
neural-fuzzy rule sets or algorithms for recognizing and
categorizing obstacles and generating and developing optimal
avoidance maneuvers where necessary.
BACKGROUND OF THE INVENTION
All of the patents, patent applications, technical papers and other
references referenced below are incorporated herein by reference in
their entirety. Various patents, patent applications, patent
publications and other published documents are discussed below as
background of the invention. No admission is made that any or all
of these references are prior art and indeed, it is contemplated
that they may not be available as prior art when interpreting 35
U.S.C. .sctn.102 in consideration of the claims of the present
application.
There are numerous components described and disclosed herein. Many
combinations of these components are described but to conserve
space, the inventors have not described all combinations and
permutations of these components but the inventors intend that each
such combination and permutation is an invention to be considered
disclosed by this disclosure. The inventors further intend to file
continuation and continuation-in-part applications to cover many of
these combinations and permutations.
Automobile accidents are one of the most serious problems facing
society today, both in terms of deaths and injuries, and in
financial losses suffered as a result of accidents. The suffering
caused by death or injury from such accidents is immense. The costs
related to medical treatment, permanent injury to accident victims
and the resulting loss of employment opportunities, and financial
losses resulting from damage to property involved in such accidents
are staggering. Providing the improved systems and methods to
eventually eliminate these deaths, injuries and other losses
deserves the highest priority. The increase in population and use
of automobiles worldwide with the concomitant increased congestion
on roadways makes development of systems for collision avoidance
and elimination even more urgent. While many advances have been
made in vehicle safety, including, for example, the use of
seatbelts, airbags and safer automobile structures, much room for
improvement exists in automotive safety and accident prevention
systems.
There are two major efforts underway that will significantly affect
the design of automobiles and highways. The first is involved with
preventing deaths and serious injuries from automobile accidents.
The second involves the attempt to reduce the congestion on
highways. In the first case, there are approximately forty two
thousand (42,000) people killed each year in the United States by
automobile accidents and another several hundred thousand are
seriously injured. In the second case, hundreds of millions of
man-hours are wasted every year by people stuck in traffic jams on
the world's roadways. There have been many attempts to solve both
of these problems; however, no single solution has been able to do
so.
When a person begins a trip using an automobile, he or she first
enters the vehicle and begins to drive, first out of a parking
space and then typically onto a local or city road and then onto a
highway. In leaving the parking space, he or she may be at risk
from an impact of a vehicle traveling on the road. The driver must
check his or her mirrors to avoid such an event and several
electronic sensing systems have been proposed which would warn the
driver that a collision is possible. Once on the local road, the
driver is at risk of being impacted from the front, side and rear,
and electronic sensors are under development to warn the driver of
such possibilities. Similarly, the driver may run into a
pedestrian, bicyclist, deer or other movable object and various
sensors are under development that will warn the driver of these
potential events. These various sensors include radar, optical,
terahertz or other electromagnetic frequencies, infrared,
ultrasonic, and a variety of other sensors, each of which attempts
to solve a particular potential collision event. It is important to
note that as yet, in none of these cases is there sufficient
confidence in the decision that the control of the vehicle is taken
away from the driver. Thus, action by the driver is still
invariably required.
In some proposed future Intelligent Transportation System (ITS)
designs, hardware of various types is embedded into the highway and
sensors which sense this hardware are placed onto the vehicle so
that it can be accurately guided along a lane of the highway. In
various other systems, cameras are used to track lane markings or
other visual images to keep the vehicle in its lane. However, for
successful ITS, additional information is needed by the driver, or
the vehicle control system, to take into account weather, road
conditions, congestion etc., which typically involves additional
electronic hardware located on or associated with the highway as
well as the vehicle. From this discussion, it is obvious that a
significant number of new electronic systems are planned for
installation ontovehicles. However, to date, no product has been
proposed or designed which combines all of the requirements into a
single electronic system. This is one of the intents of some
embodiments of this invention.
The safe operation of a vehicle can be viewed as a process in the
engineering sense. To achieve safe operation, first the process
must be designed and then a vehicle control system must be designed
to implement the process. The goal of a process designer is to
design the process so that it does not fail. The fact that so many
people are being seriously injured and killed in traffic accidents
and the fact that so much time is being wasted in traffic
congestion is proof that the current process is not working and
requires a major redesign. To design this new process, the
information required by the process must be identified, the source
of that information determined and the process designed so that the
sources of information can communicate effectively with the user of
the information, which will most often be a vehicle control system.
Finally, the process must have feedback that self-corrects the
process when it is tending toward failure.
Although it is technologically feasible, it is probably socially
unacceptable at this time for a vehicle safety system to totally
control the vehicle. An underlying premise of embodiments of this
invention, therefore, is that people will continue to operate their
vehicle and control of the vehicle will only be seized by the
control system when such an action is required to avoid an accident
or when such control is needed for the orderly movement of vehicles
through potentially congested areas on a roadway. When this
happens, the vehicle operator will be notified and given the choice
of exiting the road at the next opportunity. In some cases,
especially when this invention is first implemented on a trial
basis, control will not be taken away from the vehicle operator but
a warning system will alert the driver of a potential collision,
road departure or other infraction.
Let us consider several scenarios and what information is required
for the vehicle control process to prevent accidents. In one case,
a driver is proceeding down a country road and falls asleep and the
vehicle begins to leave the road, perhaps heading toward a tree. In
this case, the control system would need to know that the vehicle
was about to leave the road and for that, it must know the position
of the vehicle relative to the road. One method of accomplishing
this would be to place a wire down the center of the road and to
place sensors within the vehicle to sense the position of the wire
relative to the vehicle, or vice versa. An alternate approach would
be for the vehicle to know exactly where it is on the surface of
the earth and to also know exactly where the edge of the road
is.
These approaches are fundamentally different because in the former
solution every road in the world would require the placement of
appropriate hardware as well as the maintenance of this hardware.
This is obviously impractical. In the second case, the use of the
global positioning satellite system (GPS), augmented by additional
systems to be described below, will provide the vehicle control
system with an accurate knowledge of its location. While it would
be difficult to install and maintain hardware such as a wire down
the center of the road for every road in the world, it is not
difficult to survey every road and record the location of the
edges, and the lanes for that matter, of each road. This
information must then be made available through one or more of a
variety of techniques to the vehicle control system.
Another case might be where a driver is proceeding down a road and
decides to change lines while another vehicle is in the driver's
blind spot. Various companies are developing radar, ultrasonic or
optical sensors to warn the driver if the blind spot is occupied.
The driver may or may not heed this warning, perhaps due to an
excessive false alarm rate, or he or she may have become
incapacitated, or the system may fail to detect a vehicle in the
blind spot and thus the system will fail.
Consider an alternative technology where again each vehicle knows
precisely where it is located on the earth surface and additionally
can communicate this information to all other vehicles within a
certain potential danger zone relative to the vehicle. Now, when
the driver begins to change lanes, his or her vehicle control
system knows that there is another vehicle in the blind spot and
therefore will either warn the driver or else prevent him or her
from changing lanes thereby avoiding the accident.
Similarly, if a vehicle is approaching a stop sign, other traffic
marker or red traffic light and the operator fails to bring the
vehicle to a stop, if the existence of this traffic light and its
state (red in this example) or stop sign has been made available to
the vehicle control system, the system can warn the driver or seize
control of the vehicle to stop the vehicle and prevent a potential
accident. Additionally, if an operator of the vehicle decides to
proceed across an intersection without seeing an oncoming vehicle,
the control system will once again know the existence and location
and perhaps velocity of the oncoming vehicle and warn or prevent
the operator from proceeding across the intersection.
Consider another example where water on the surface of a road is
beginning to freeze. Probably the best way that a vehicle control
system can know that the road is about to become slippery, and
therefore that the maximum vehicle speed must be significantly
reduced, is to get information from some external source. This
source can be sensors located on the highway that are capable of
determining this condition and transmitting it to the vehicle.
Alternately, the probability of icing occurring can be determined
analytically from meteorological data and a historical knowledge of
the roadway and communicated to the vehicle over a LEO or GEO
satellite system, the Internet or an FM sub-carrier or other means.
A combination of these systems can also be used.
Studies have shown that a combination of meteorological and
historic data can accurately predict that a particular place on the
highway will become covered with ice. This information can be
provided to properly equipped vehicles so that the vehicle knows to
anticipate slippery roads. For those roads that are treated with
salt to eliminate frozen areas, the meteorological and historical
data will not be sufficient. Numerous systems are available today
that permit properly equipped vehicles to measure the coefficient
of friction between the vehicle's tires and the road. It is
contemplated that perhaps police or other public vehicles will be
equipped with such a friction coefficient measuring apparatus and
can serve as probes for those roadways that have been treated with
salt. Information from these probe vehicles will be fed into the
information system that will then be made available to control
speed limits in those areas.
Countless other examples exist; however, from those provided above,
it can be seen that for the vehicle control system to function
without error, certain types of information must be accurately
provided. These include information permitting the vehicle to
determine its absolute location and means for vehicles near each
other to communicate this location information to each other.
Additionally, map information that accurately provides boundary and
lane information of the road must be available. Also, critical
weather or road-condition information is necessary. The road
location information need only be generated once and changed
whenever the road geometry is altered. This information can be
provided to the vehicle through a variety of techniques including
prerecorded media such as CD-ROM or DVD disks or through
communications from transmitters located in proximity to the
vehicle, satellites, radio and cellular phones.
Consider now the case of the congested highway. Many roads in the
world are congested and are located in areas where the cost of new
road construction is prohibitive or such construction is
environmentally unacceptable. It has been reported that an accident
on such a highway typically ties up traffic for a period of
approximately four times the time period required to clear the
accident. Thus, by eliminating accidents, a substantial improvement
of the congested highway problem is obtained. This of course is
insufficient. On such highways, each vehicle travels with a
different spacing, frequently at different speeds and in the wrong
lanes. If the proper spacing of the vehicles could be maintained,
and if the risk of an accident could be substantially eliminated,
vehicles under automatic control could travel at substantially
higher velocities and in a more densely packed configuration
thereby substantially improving the flow rate of vehicles on the
highway by as much as a factor of 3 to 4 times. This not only will
reduce congestion but also improve air pollution. Once again, if
each vehicle knows exactly where it is located, can communicate its
location to surrounding vehicles and knows precisely where the road
is located, then the control system in each vehicle has sufficient
information to accomplish this goal.
Again, an intention of the system and process described here is to
totally eliminate automobile accidents as well as reduce highway
congestion. This process is to be designed to have no defective
decisions. The process employs information from a variety of
sources and utilizes that information to prevent accidents and to
permit the maximum vehicle throughput on highways.
The information listed above is still insufficient. The geometry of
a road or highway can be determined once and for all, until erosion
or construction alters the road. Properly equipped vehicles can
know their location and transmit that information to other properly
equipped vehicles. There remains a variety of objects whose
location is not fixed, which have no transmitters and which can
cause accidents. These objects include broken down vehicles,
animals such as deer which wander onto highways, pedestrians,
bicycles, objects which fall off of trucks, and especially other
vehicles which are not equipped with location determining systems
and transmitters for transmitting that information to other
vehicles. Part of this problem can be solved for congested highways
by restricting access to these highways to vehicles that are
properly equipped. Also, these highways are typically in urban
areas and access by animals can be effectively eliminated. Heavy
fines can be imposed on vehicles that drop objects onto the
highway. Finally, since every vehicle and vehicle operator becomes
part of the process, each such vehicle and operator becomes a
potential source of information to help prevent catastrophic
results. Thus, each vehicle should also be equipped with a system
of essentially stopping the process in an emergency. Such a system
could be triggered by vehicle sensors detecting a problem or by the
operator strongly applying the brakes, rapidly turning the steering
wheel or by activating a manual switch when the operator observes a
critical situation but is not himself in immediate danger. An
example of the latter case is where a driver witnesses a box
falling off of a truck in an adjacent lane.
To solve the remaining problems, therefore, each vehicle should
also be equipped with an anticipatory collision sensing system, or
collision forecasting system, which is capable of identifying or
predicting and reacting to a pending accident. As the number of
vehicles equipped with the control system increases, the need for
the collision forecasting system will diminish.
Once again, the operator will continue to control his vehicle
provided he or she remains within certain constraints. These
constraints are like a corridor. As long as the operator maintains
his vehicle within this allowed corridor, he or she can operate
that vehicle without interference from the control system. That
corridor may include the entire width of the highway when no other
vehicles are present or it may be restricted to all eastbound
lanes, for example. In still other cases, that corridor may be
restricted to a single line and additionally, the operator may be
required to keep his vehicle within a certain spacing tolerance
from the preceding vehicle. If a vehicle operator wishes to exit a
congested highway, he could operate his turn signal that would
inform the control system of this desire and permit the vehicle to
safely exit from the highway. It can also inform other adjacent
vehicles of the operator's intent, which could then automatically
cause those vehicles to provide space for lane changing, for
example. The highway control system is thus a network of individual
vehicle control systems rather than a single highway resident
computer system.
Considering now the U.S. Department of Transportation (DOT) policy,
in the DOT FY 2000 Budget in Brief Secretary Rodney Slater states
that "Historic levels of federal transportation investment . . .
are proposed in the FY 2000 budget." Later, Secretary Slater states
that "Transportation safety is the number one priority." DOT has
estimated that $165 billion per year are lost in fatalities and
injuries on U.S. roadways. Another $50 billion are lost in wasted
time of people on congested highways. Presented herein is a plan to
eliminate fatalities and injuries and to substantially reduce
congestion. The total cost of implementing this plan is minuscule
compared to the numbers stated above. This plan has been named the
"Road to Zero Fatalities.TM.", or RtZF.TM. for short.
The DOT Performance Plan FY 2000, Strategic Goal: Safety, states
that "The FY 2000 budget process proposes over $3.4 billion for
direct safety programs to meet this challenge." The challenge is to
"Promote the public health and safety by working toward the
elimination of traffic related deaths, injuries and property
damage". The goal of the RtZF.TM., as described below and which is
a part of the present invention, is the same and herein a plan is
presented for accomplishing this goal. The remainder of the DOT
discussion centers around wishful thinking to reduce the number of
transportation-related deaths, injuries, etc. However, the
statistics presented show that in spite of this goal, the number of
deaths is now increasing. As discussed below, this is the result of
a failed process.
Reading through the remainder of the DOT Performance Plan FY 2000,
one is impressed by the billions of dollars that are being spent to
solve the highway safety problem coupled with the enormous
improvement that has been made until the last few years. It can
also be observed that the increase in benefits from these
expenditures has now disappeared. For example, the fatality rate
per 100 million vehicle miles traveled fell from 5.5 to 1.7 in the
period from the mid-1960s to 1994. But this decrease has now
substantially stopped! This is an example of the law of diminishing
returns and signals the need to take a totally new approach to
solving this problem.
The U.S. Intelligent Vehicle Initiative (IVI) policy states that
significant funds have been spent on demonstrating various ITS
technologies. It is now time for implementation. With over 40,000
fatalities and almost four million people being injured every year
on U.S. roadways, it is certainly time to take affirmative action
to stop this slaughter. The time for studies and demonstrations is
past. However, the deployment of technologies that are inconsistent
with the eventual solution of the problem will only delay
implementation of the proper systems and thereby result in more
deaths and injuries.
A primary goal of the Intelligent Vehicle Initiative was to reduce
highway related fatalities per 100 million vehicle miles traveled
from 1.7 in 1996 to 1.6 in 2000. Of course, the number of
fatalities may still increase due to increased road use. If this
reduction in fatalities comes about due to slower travel speeds,
because of greater congestion, then has anything really been
accomplished? Similar comments apply to the goal of reducing the
rate of injury per 100 million vehicle miles from 141 in 1996 to
128 in 2000. An alternate goal, as described herein, is to have the
technology implemented on all new vehicles by the year 2010 that
will eventually eliminate all fatalities and injuries. As an
intermediate milestone, it is proposed to have the technology
implemented on all new vehicles by 2007 to reduce or eliminate
fatalities caused by road departure, center (yellow) line crossing,
stop sign infraction, rear end and excessive speed accidents.
Inventions described herein will explain how these are goals can be
attained.
In the IVI Investment Strategy, Critical Technology Elements And
Activities of the DOT, it says "The IVI will continue to expand
these efforts particularly in areas such as human factors, sensor
performance, modeling and driver acceptance". An alternate, more
effective, concentration for investments would be to facilitate the
deployment of those technologies that will reduce and eventually
eliminate highway fatalities. Driver acceptance and human factors
will be discussed below. Too much time and resources have already
been devoted to these areas. Modeling can be extremely valuable and
sensor performance is in a general sense a key to eliminating
fatalities.
On Jul. 15, 1998, the IVI light vehicle steering committee met and
recommended that the IVI program should be conducted as a
government industry partnership like the PNGV. This is believed to
be quite wrong and it is believed that the IVI should now move
vigorously toward the deployment of proven technology.
The final recommendations of the committee was "In the next five
years, the IVI program should be judged on addressing selected
impediments preventing deployment, not on the effect of IVI
services on accident rates." This is believed to be a mistake. The
emphasis for the next five years should be to deploy proven
technologies and to start down the Road to Zero Fatalities.TM..
Five years from now technology should be deployed on production
vehicles sold to the public that have a significant effect toward
reducing fatalities and injuries.
As described in the paper "Preview Based Control of A Tractor
Trailer Using DGPS For Preventing Road Departure Accidents" the
basis of the technology proposed has been demonstrated.
DISCUSSION AND REVIEW OF RELEVANT ART
1. Vehicle Collision Warning and Control
The world is experiencing an unacceptable growth in traffic
congestion and attention is increasingly turning to smart highway
systems to solve the problem. It has been estimated that
approximately $240 billion will be spent on smart highways over the
next 20 years. All of the initiatives currently being considered
involve a combination of vehicle-mounted sensors and sensors and
other apparatus installed in or on the roadway. Such systems are
expensive to install, difficult and expensive to maintain and will
thus only be used on major highways, if at all. Although there will
be some safety benefit from such systems, it will be limited to the
highways which have the system and perhaps to only a limited number
of lanes.
The RtZF.TM. system in accordance with the invention eliminates the
shortcomings of the prior art by providing a system that does not
require modifications to the highway. The information as to the
location of the highway is determined, as discussed above, by
mapping the edges of the roadway and the edges of the lanes of the
roadway using a process whereby the major roads of the entire
country can be mapped at very low cost. Thus, the system has the
capability of reducing congestion as well as saving lives on all
major roads, not just those which have been selected as high-speed
guided lanes.
The ALVINN project of Carnegie Mellon University (Jochem, Todd M.,
Pomerleau, Dean A., and Thorpe, Charles E., "Vision-Based Neural
Network Road and Intersection Detection and Traversal", IEEE
Conference on Intelligent Robots and Systems, Aug. 5 9, 1995,
Pittsburgh, Pa., USA)) describes an autonomous land vehicle using a
neural network. The neural network is trained based on how a driver
drives the vehicle given the output from a video camera. The output
of the neural network is the direction that the vehicle should
travel in based on the input information from the video camera and
the training based on what a good driver would do. A similar system
can be used in some embodiments of the present invention to guide a
vehicle to a safe stop in the event that the driver becomes
incapacitated or some other emergency situation occurs wherein the
driver is unable to control the vehicle. The input to the neural
network in this case would be the map information rather than a
video camera. Additionally, a laser radar or terahertz radar
imaging system of this invention could also be an input to the
system. This neural network system can additionally take over in
the event that an accident becomes inevitable. Simple neural
networks are probably not sufficient for this purpose and neural
fuzzy, modular neural networks or combination neural networks are
probably required.
U.S. Pat. No. 5,479,173 to Yoshioka, et al. uses a steering angle
sensor, a yaw rate sensor and a velocity of the vehicle sensor to
predict the path that the vehicle will take. It uses a radar unit
to identify various obstacles that may be in the path of the
vehicle, and it uses a CCD camera to try to determine that the road
is changing direction in front of the vehicle. No mention is made
of the accuracy with which these determinations are made. It is
unlikely that sub-meter accuracy is achieved. If an obstacle is
sensed, the brakes can be automatically activated.
U.S. Pat. No. 5,540,298 to Yoshioka, et al. is primarily concerned
with changing the suspension and steering characteristics of the
vehicle in order to prevent unstable behavior of the vehicle in
response to the need to exercise a collision avoidance maneuver.
The collision anticipation system includes an ultrasonic unit and
two optical laser radar units.
U.S. Pat. No. 5,572,428 to Ishida is concerned with using a radar
system plus a yaw rate sensor and a velocity sensor to determine
whether a vehicle will collide with another vehicle based on the
area occupied by each vehicle. Since radar cannot accurately
determine this area, it has to be assumed by the system.
U.S. Pat. No. 5,613,039 to Wang, et al. is a collision warning
radar system utilizing a real time adaptive probabilistic neural
network. Wang discusses that about 60% of roadway collisions could
be avoided if the operator of the vehicle was provided warning at
least one-half second prior to a collision. The radar system used
by Wang uses two separate frequencies. The reflective radar signals
are analyzed by a probabilistic neural network that provides an
output signal indicative of the likelihood and threat of a
collision with a particular object. A Fourier transform circuit
converts the digitized reflective signal from a time series to a
frequency representation. It is important to note that in this
case, as in the others above, true collision avoidance will not
occur since, without a knowledge of the roadway, two vehicles can
be approaching each other on a collision course, each following a
curved lane on a highway and yet the risk of collision is minimal
due to the fact that each vehicle remains in its lane. Thus, true
collision avoidance cannot be obtained without an accurate
knowledge of the road geometry.
U.S. Pat. No. 5,983,161 to Lemelson describes a GPS-based collision
avoidance and warning system that contains some of the features of
embodiments of the present invention. This patent is primarily
concerned with using centimeter-accuracy DGPS systems to permit
vehicles on a roadway to learn and communicate their precise
locations to other vehicles. In that manner, a pending collision
can, in some cases, be predicted.
Lemelson does not use an inertial navigation system for controlling
the vehicle between GPS updates. Thus, the vehicle can travel a
significant distance before its position can be corrected. This can
lead to significant errors. Lemelson also does not make use of
accurate map database and thus it is unable to distinguish cases
where two cars are on separate lanes but on an apparent collision
course. Although various radar and lidar systems are generally
discussed, the concept of range gating is not considered. Thus, the
Lemelson system is unable to provide the accuracy and reliability
required by the Road to Zero Fatalities.TM. system described
herein.
Since many of the concepts disclosed in the inventions herein make
use of neural networks, a background of neural networks is
important to the reader. The theory of neural networks including
many examples can be found in several books on the subject
including: (1) Techniques and Application of Neural Networks,
edited by Taylor, M. and Lisboa, P., Ellis Horwood, West Sussex,
England, 1993; (2) Naturally Intelligent Systems, by Caudill, M.
and Butler, C., MIT Press, Cambridge Mass., 1990; (3) J. M. Zaruda,
Introduction to Artificial Neural Systems, West publishing Co.,
N.Y., 1992, (4) Digital Neural Networks, by Kung, S. Y., PTR
Prentice Hall, Englewood Cliffs, N.J., 1993, Eberhart, R., Simpson,
P., (5) Dobbins, R., Computational Intelligence PC Tools, Academic
Press, Inc., 1996, Orlando, Fla., (6) Cristianini, N. and
Shawe-Taylor, J. An Introduction to Support Vector Machines and
Other Kernel-Based Learning Methods, Cambridge University Press,
Cambridge England, 2000; (7) Proceedings of the 2000 6.sup.th IEEE
International Workshop on Cellular Neural Networks and their
Applications (CNNA 2000), IEEE, Piscataway N.J.; and (8) Sinha, N.
K. and Gupta, M. M. Soft Computing & Intelligent Systems,
Academic Press 2000 San Diego, Calif. The neural network pattern
recognition technology is one of the most developed of pattern
recognition technologies. The invention described herein uses
combinations of neural networks to improve the pattern recognition
process.
2. Accurate Navigation
U.S. Pat. No. 5,504,482 to Schreder describes an automobile
equipped with an inertial and satellite navigation system as well
as a local area digitized street map. The main use of this patent
is for route guidance in the presence of traffic jams, etc.
Schreder describes how information as to the state of the traffic
on a highway can be transmitted and utilized by a properly equipped
vehicle to change the route the driver would take in going to his
destination. Schreder does not disclose sub-meter vehicle location
accuracy determination, nevertheless, this patent provides a good
picture of the state of the art as can be seen from the following
quoted paragraphs:
" . . . there exists a wide range of technologies that have
disadvantageously not been applied in a comprehensive integrated
manner to significantly improve route guidance, reduce pollution,
improve vehicular control and increase safety associated with the
common automobile experience. For example, it is known that gyro
based inertial navigation systems have been used to generate
three-dimensional position information, including exceedingly
accurate acceleration and velocity information over a relatively
short travel distance, and that GPS satellite positioning systems
can provide three-dimensional vehicular positioning and epoch
timing, with the inertial system being activated when satellite
antenna reception is blocked during "drop out" for continuous
precise positioning. It is also known that digitized terrain maps
can be electronically correlated to current vehicular transient
positions, as have been applied to military styled transports and
weapons. For another example, it is also known that digitally
encoded information is well suited to RF radio transmission within
specific transmission carrier bands, and that automobiles have been
adapted to received AM radio, FM radio, and cellular
telecommunication RF transmissions. For yet another example, it is
further known that automobile electronic processing has been
adapted to automatically control braking, steering, suspension and
engine operation, for example, anti-lock braking, four-wheel
directional steering, dynamic suspension stiffening during turns
and at high speeds, engine governors limiting vehicular speed, and
cruise control for maintaining a desired velocity. For still
another example, traffic monitors, such as road embedded magnetic
traffic light sensor loops and road surface traffic flow meters
have been used to detect traffic flow conditions. While these
sensors, meters, elements, systems and controls have served limited
specific purposes, the prior art has disadvantageously failed to
integrate them in a comprehensive fashion to provide a complete
dynamic route guidance, dynamic vehicular control, and safety
improvement system."
"Recently, certain experimental integrated vehicular dynamic
guidance systems have been proposed. Motorola has discussed an
Intelligent Vehicle Highway System in block diagram form in
copyright dated 1993 brochure. Delco Electronics has discussed
another Intelligent Vehicle Highway System also in block diagram
form in Automotive News published on Apr. 12, 1993. These systems
use compass technology for vehicular positioning. However,
displacement wheel sensors are plagued by tire slippage, tire wear
and are relatively inaccurate requiring recalibration of the
current position. Compasses are inexpensive, but suffer from
drifting particularly when driving on a straight road for extended
periods. Compasses can sense turns, and the system may then be
automatically recalibrated to the current position based upon
sensing a turn and correlating that turn to the nearest turn on a
digitized map, but such recalibration, is still prone to errors
during excessive drifts. Moreover, digitized map systems with the
compass and wheel sensor positioning methods operate in two
dimensions on a three dimensional road terrain injecting further
errors between the digitized map position and the current vehicular
position due to a failure to sense the distance traveled in the
vertical dimension."
"These Intelligent Vehicle Highway Systems appear to use GPS
satellite reception to enhance vehicular tracking on digitized road
maps as part of a guidance and control system. These systems use
GPS to determine when drift errors become excessive and to indicate
that recalibration is necessary. However, the GPS reception is not
used for automatic accurate recalibration of current vehicular
positioning, even though C-MIGITS and like devices have been used
for GPS positioning, inertial sensing and epoch time monitoring,
which can provide accurate continuous positioning."
"These Intelligent Vehicle Highway Systems use the compass and
wheel sensors for vehicular positioning for route guidance, but do
not use accurate GPS and inertial route navigation and guidance and
do not use inertial measuring units for dynamic vehicular control.
Even though dynamic electronic vehicular control, for example,
anti-lock braking, anti-skid steering, and electronic control
suspension have been contemplated by others, these systems do not
appear to functionally integrate these dynamic controls with an
accurate inertial route guidance system having an inertial
measuring unit well suited for dynamic motion sensing. There exists
a need to further integrate and improve these guidance systems with
dynamic vehicular control and with improved navigation in a more
comprehensive system."
"These Intelligent Vehicle Highway Systems also use RF receivers to
receive dynamic road condition information for dynamic route
guidance, and contemplate infrastructure traffic monitoring, for
example, a network for road magnetic sensing loops, and contemplate
the RF broadcasting of dynamic traffic conditions for dynamic route
guidance. The discussed two-way RF communication through the use of
a transceiver suggests a dedicated two-way RF radio data system.
While two-way RF communication is possible, the flow of necessary
information between the vehicles and central system appears to be
exceedingly lopsided. The flow of information from the vehicles to
a central traffic radio data control system may be far less than
the required information from traffic radio data control system to
the vehicles. It seems that the amount of broadcasted dynamic
traffic flow information to the vehicles would be far greater than
the information transmitted from the vehicles to the central
traffic control center. For example, road side incident or accident
emergency messages to a central system may occur far less than the
occurrences of congested traffic points on a digitized map having a
large number of road coordinate points."
"Conserving bandwidth capacity is an objective of RF communication
systems. The utilization of existing infra structure
telecommunications would seem cost-effective. AT&T has recently
suggested improving the existing cellular communication network
with high-speed digital cellular communication capabilities. This
would enable the use of cellular telecommunications for the purpose
of transmitting digital information encoding the location of
vehicular incidents and accidents. It then appears that a vehicular
radio data system would be cost-effectively used for unidirectional
broadcasting of traffic congestion information to the general
traveling public, while using existing cellular telecommunication
systems for transmitting emergency information. The communication
system should be adapted for the expected volume of information.
The Intelligent Vehicular Highway Systems disadvantageously suggest
a required two-way RF radio data system. The vast amount of
information that can be transmitted may tend to expand and
completely occupy a dedicated frequency bandwidth. To the extent
that any system is bi-directional in operation tends to
disadvantageously require additional frequency bandwidth capacity
and system complexity."
2.1 GPS
Referring to FIG. 1, the presently implemented Global Positioning
System with its constellation of 24 satellites 2 is truly
revolutionizing navigation throughout the world. The satellites
orbit the Earth in six orbits 4. However, in order to reach its
full potential for navigation, GPS needs to be augmented both to
improve accuracy and to reduce the time needed to inform a vehicle
driver of a malfunction of a GPS satellite, the so-called integrity
problem.
The Global Positioning System (GPS) is a satellite-based navigation
and time transfer system developed by the U.S. Department of
Defense. GPS serves marine, airborne and terrestrial users, both
military and civilian. Specifically, GPS includes the Standard
Positioning Service (SPS) that provides civilian users with 100
meter accuracy as to the location or position of the user. It also
serves military users with the Precise Positioning Service that
provides 20-meter accuracy for the user. Both of these services are
available worldwide with no requirement for any local
equipment.
2.2 DGPS, WAAS, LAAS and Pseudolites
Differential operation of GPS is used to improve the accuracy and
integrity of GPS. Differential GPS places one or more high quality
GPS receivers at known surveyed locations to monitor the received
GPS signals. This reference station(s) estimates the slowly varying
components of the satellite range measurements, and forms a
correction for each GPS satellite in view. The correction is
broadcast to all DGPS users within the coverage area of the
broadcast facilities.
For a good discussion of DGPS, several are reproduced from
OMNISTAR: in U.S. patent application Ser. No. 10/822,445, now U.S.
Pat. No. 7,085,637, and incorporated by reference herein.
The above description is provided to illustrate the accuracy which
can be obtained from the DGPS system. It is expected that the WAAS
system when fully implemented will provide the same benefits as
provided by the OMNISTAR system. However, when the standard
deviation of approximately 0.5 meter is considered, it is evident
that this WAAS system is insufficient by itself and will have to be
augmented by other systems to improve the accuracy at least at this
time.
GLONASS is a Russian system similar to GPS. This system provides
accuracy that is not as good as GPS.
The Projected Position Accuracy of GPS and GLONASS, Based on the
Current Performance is:
TABLE-US-00001 Horizontal Error (m) Vertical Error (m) (50%) (95%)
(95%) GPS 7 18 34 GLONASS 10 26 45
The system described here will achieve a higher accuracy than
reported in the above table due to the combination of the inertial
guidance system that permits accurate changes in position to be
determined and through multiple GPS readings. In other words, the
calculated position will converge to the real position over time.
The addition of DGPS will provide an accuracy improvement of at
least a factor of 10, which, with the addition of a sufficient
number of DGPS stations in some cases is sufficient without the use
of the carrier frequency correction. A further refinement where the
vehicle becomes its own DGPS station through the placement of
infrastructure stations at appropriate locations on roadways will
further significantly enhance the system accuracy to the required
level.
Multipath is the situation where more than one signal from a
satellite comes to a receiver with one of the signals resulting
from a reflection off of a building or the ground, for example.
Since multipath is a function of geometry, the system can be
designed to eliminate its effects based on highway surveying and
appropriate antenna design. Multipath from other vehicles can also
be eliminated since the location of the other vehicles will be
known.
As discussed below, the Wide Area Augmentation System (WAAS) is
being installed by the US Government to provide DGPS for airplane
landings. The intent is to cover the entire CONUS. This may be
useful for much of the country for the purposes of this invention.
Another alternative would be to use the cellular phone towers,
since there are so many of them, if they could be programmed to act
as pseudolites.
An important feature of DGPS is that the errors from the GPS
satellites change slowly with time and therefore, only the
corrections need be sent to the user from time to time. Using
reference receivers separated by 25 120 km, accuracies from 2 cm to
1 m are achievable using local area DGPS which is marginal for
RtZF.TM.. Alternately, through the placement of appropriate
infrastructure as described below even better accuracies are
obtainable.
A type of wide area DGPS (WADGPS) system has been developed spans
the entire U.S. continent which provides position RMS accuracy to
better than about 50 cm. This system is described in the Bertiger,
et al, "A Prototype Real-Time Wide Area Differential GPS System,"
Proceedings of the National Technical Meeting, Navigation and
Positioning in the Information Age, Institute of Navigation, Jan.
14 16, 1997 pp. 645 655. A RMS error of 50 cm would be marginally
accurate for RtZF.TM.. Many of the teachings of this invention,
especially if the road edge and lane location error were much less,
could be accomplished using more accurate surveying equipment. The
OmniSTAR system is another WADGPS system that claims 6 cm
(1.sigma.) accuracy.
A similar DGPS system which is now being implemented on a
nationwide basis is described in "DGPS Architecture Based on
Separating Error Components, Virtual Reference Stations and FM
Subcarrier Broadcast", by Differential Corrections Inc., 10121
Miller Ave., Cupertino, Calif. 95041. The system described in this
paper promises an accuracy on the order of about 10 cm.
Suggested DGPS update rates are usually less than twenty seconds.
DGPS removes common-mode errors, those errors common to both the
reference and remote receivers (not multipath or receiver noise).
Errors are more often common when receivers are close together
(less than 100 km). Differential position accuracies of 1 10 meters
are possible with DGPS based on C/A code SPS signals.
Using the CNET commercial system, 1 foot accuracies are possible if
base stations are no more than 30 miles from the vehicle unit. This
would require approximately 1000 base stations to cover CONUS.
Alternately, the same accuracy is obtainable if the vehicle can
become its own DGPS system every 30 miles as described herein.
Unfortunately, the respective error sources mentioned above rapidly
decorrelate as the distances between the reference station and the
vehicle increases. Conventional DGPS is the terminology used when
the separation distances are sufficiently small that the errors
cancel. The terms single-reference and multi-reference DGPS are
occasionally used in order to emphasize whether there is a single
reference station or whether there are multiple ones. If it is
desired to increase the area of coverage and, at the same time, to
minimize the number of fixed reference receivers, it becomes
necessary to model the spatial and temporal variations of the
residual errors. Wide Area Differential GPS (WADGPS) is designed to
accomplish this. Funds have now been appropriated for the U.S.
Government to deploy a national DGPS system.
The Wide Area Augmentation System (WAAS) is being deployed to
replace the Instrument Landing System used at airports across the
country. The WAAS system provides an accuracy of from about 1 to 2
meters for the purpose of aircraft landing. If the vertical
position of the vehicle is known, as would be in the case of
automobiles at a known position on a road, this accuracy can be
improved significantly. Thus, for many of the purposes of this
invention, the WAAS can be used to provide accurate positioning
information for vehicles on roadways. The accuracy of the WAAS is
also enhanced by the fact that there is an atomic clock in every
WAAS receiver station that would be available to provide great
accuracy using carrier phase data. With this system, sub-meter
accuracies are possible for some locations.
The WAAS is based on a network of approximately 35 ground reference
stations. Signals from GPS satellites are received by aircraft
receivers as well as by ground reference stations. Each of these
reference stations is precisely surveyed, enabling each to
determine any error in the GPS signals being received at its own
location. This information is then passed to a wide area master
station. The master station calculates correction algorithms and
assesses the integrity of the system. This data is then put into a
message format and sent to a ground earth station for uplink to a
geostationary communications satellite. The corrective information
is forwarded to the receiver on board the aircraft, which makes the
needed adjustments. The communications satellites also act as
additional navigation satellites for the aircraft, thus, providing
additional navigation signals for position determination.
This system will not meet all of FAA's requirements. For category
III landings, the requirement is 1.6-m vertical and horizontal
accuracy. To achieve this, FAA is planning to implement a network
of local area differential GPS stations that will provide the
information to aircraft. This system is referred to as the Local
Area Augmentation System (LAAS).
The WAAS system, which consists of a network of earth stations and
geo-synchronous satellites, is currently being funded by the U.S.
Government for aircraft landing purposes. Since the number of
people that die yearly in automobile accidents greatly exceeds
those killed in airplane accidents, there is clearly a greater need
for a WAAS-type system for solving the automobile safety problem
using the teachings of this invention. Also, the reduction in
required highway funding resulting from the full implementation of
this invention would more than pay for the extension and tailoring
of the WAAS to cover the nation's highways.
The Local Area Augmented System (LAAS) is also being deployed in
addition to the WAAS system to provide even greater coverage for
the areas surrounding major airports. According to Newsletter of
the Institute of Navigation, 1997, "the FAA's schedule for (LAAS)
for Category II and III precision instrument approaches calls for
development of standards by 1998 that will be sufficient to
complete a prototype system by 2001. The next step will be to work
out standards for an operational system to be fielded in about
2005, that could serve nationwide up to about 200 runways for Cat
II III approaches."
In a country like the United States, which has many airfields, a
WAAS can serve a large market and is perhaps most effective for the
control of airplane landings. The best way for other countries,
with fewer airports, to participate in the emerging field of
GPS-based aviation aids may be to build LAAS. In countries with a
limited number of airports, LAAS is not very expensive while the
costs of building a WAAS to get Category I type accuracy is very
expensive. However, with the added benefit of less highway
construction and greater automobile safety, the added costs for a
WAAS system may well be justified for much of the world.
For the purposes of the RtZF.TM. system, both the WAAS and LAAS
would be useful but probably insufficient unless the information is
used in a different mathematical system such as used by the
OmniSTAR.TM. WADGPS system. Unlike an airplane, there are many
places where it might not be possible to receive LAAS and WAAS
information or, even more importantly, the GPS signals themselves
with sufficient accuracy and reliability. Initial RtZF.TM. systems
may therefore rely on the WAAS and LAAS but as the system develops
more toward the goal of zero fatalities, road-based systems which
permit a vehicle to pinpoint its location will be preferred.
However, there is considerable development ongoing in this field so
that all systems are still candidates for use with RtZF.TM. system
and the most cost effective will be determined in time.
Pseudolites are artificial satellite like structures, located on
the earth surface, that can be deployed to enhance the accuracy of
the DGPS system. Such structures could become part of the RtZF.TM.
system.
2.3 Carrier Phase Measurements
An extremely accurate form of GPS is Carrier Based Differential
GPS. This form of GPS utilizes the 1.575 GHz carrier component of
the GPS signal on which the Pseudo Random Number (PRN) code and the
data component are superimposed. Current versions of Carrier Based
Differential GPS involve generating position determinations based
on the measured phase differences at two different antennas, a base
station or pseudolite and the vehicle, for the carrier component of
a GPS signal. This technique initially requires determining how
many integer wave-lengths of the carrier component exist between
the two antennas at a particular point in time. This is called
integer ambiguity resolution. A number of approaches currently
exist for integer ambiguity resolution. Some examples can be found
in U.S. Pat. No. 5,583,513 and U.S. Pat. No. 5,619,212. Such
systems can achieve sub-meter accuracies and, in some cases,
accuracies of about 1 cm or less. U.S. Pat. No. 5,477,458 discusses
a DGPS system that is accurate to about 5 cm with the base stations
located on a radius of about 3000 km. With such a system, very few
base stations would be required to cover the CONUS. This system
still suffers from the availability of accurate signals at the
vehicle regardless of its location on the roadway and the location
of surrounding vehicles and objects. Nevertheless, the principle of
using the carrier frequency to precisely determine the location of
a vehicle can be used with the highway-based systems described
below to provide extreme location accuracies.
Several attempts to improve the position accuracy of GPS are
discussed here, for example, the Wide Area Augmentation System
(WAAS), the Local Area Augmentation System (LAAS) and various
systems that make use of the carrier phase.
A paper by S. Malys et al., titled "The GPS Accuracy Improvement
Initiative" provides a good discussion of the errors inherent in
the GPS system without using differential corrections. It is there
reported that the standard GPS provides a 9-meter RMS 3-D
navigational accuracy to authorize precise positioning service
users. This reference indicates that there are improvements planned
in the GPS system that will further enhance its accuracy. The
accuracies of these satellites independently of the accuracies of
receiving units is expected to be between 1 and 1.5 meters RMS.
Over the past eight years of GPS operations, a 50% (4.6 meter to
2.3 meter) performance improvement has been observed for the signal
in space range errors. This, of course, is the RMS error. The
enhancements contained in the accuracy improvement initiative will
provide another incremental improvement from the current 2.3 meters
to 1.3 meters and perhaps to as low as 40 centimeters.
Pullen, Samuel, Enge, Per and Parkinson, Bradford,
"Simulation-Based Evaluation of WAAS Performance: Risk and
Integrity Factors" discusses the accuracy that can be expected from
the WAAS system. This paper indicates that the standard deviation
for WAAS is approximately 1 meter. To get more accurate results
requires more closely spaced differential stations. Using DGPS
stations within 1,500 kilometers from the vehicle, high accuracy
receivers can determine a location within 3 meters accuracy for
DGPS according to the paper. Other providers of DGPS corrections
claim considerably better accuracies.
From a paper by J. F. Zumberge, M. M. Watkins and F. H. Webb,
titled "Characteristics and Applications of Precise GPS Clock
Solutions Every 30 Seconds", Journal of the Institute of
Navigation, Vol. 44, No. 4, Winter 1997 1998, it appears that by
using the techniques described in this reference, the WAAS system
could eventually be improved to provide accuracies in the
sub-decimeter range for moving vehicles without the need for other
DGPS systems. This data would be provided every 30 seconds.
W. I. Bertiger et al., "A Real-Time Wide Area Differential GPS
System", Journal of the Institute of Navigation, Vol. 44, No. 4,
Winter 1997 1998. This paper describes the software that is to be
used with the WAAS System. The WAAS System is to be completed by
2001. The goal of the research described in this paper is to
achieve sub-decimeter accuracies worldwide, effectively equaling
local area DGPS performance worldwide. The full computation done on
a Windows NT computer adds only about 3 milliseconds. The
positioning accuracy is approximately 25 centimeters in the
horizontal direction. That is, the RMS value so that gives an error
at +3 sigma of 1.5 meters. Thus, this real time wide area
differential GPS system is not sufficiently accurate for the
purposes of some embodiments of this invention. Other systems claim
higher accuracies.
According to the paper by R. Braff, titled "Description of the
FAA's Local Area Augmentation System (LAAS)", Journal of the
Institute of Navigation, Vol. 44, No. 4, Winter 1997 1998, the LAAS
System is the FAA's ground-based augmentation system for local area
differential GPS. It is based on providing corrections of errors
that are common to both ground-based and aircraft receivers. These
corrections are transmitted to the user receivers via very high
frequency (VHF), line of sight radio broadcast. LAAS has the
capability of providing accuracy on the order of 1 meter or better
on the final approach segment and through rollout. LAAS broadcasts
navigational information in a localized service volume within
approximately 30 nautical miles of the LAAS ground segment.
O'Connor, Michael, Bell, Thomas, Elkaim, Gabriel and Parkinson,
Bradford, "Automatic Steering of Farm Vehicles Using GPS" describes
an automatic steering system for farm vehicles where the vehicle
lateral position error never deviated by more than 10 centimeters,
using a carrier phase differential GPS system whereby the
differential station was nearby.
The following quote is from Y. M. Al-Haifi et al., "Performance
Evaluation of GPS Single-Epoch On-the Fly Ambiguity Resolution",
Journal of the Institute of Navigation, Vol. 44, No. 4, Winter 1997
1998. This technique demonstrates sub-centimeter precision results
all of the time provided that at least five satellites are
available and multipath errors are small. A resolution of 0.001
cycles is not at all unusual for geodetic GPS receivers. This leads
to a resolution on the order of 0.2 millimeters. In practice,
multipath affects, usually from nearby surfaces, limit the accuracy
achievable to around 5 millimeters. It is currently the case that
the reference receiver can be located within a few kilometers of
the mobile receiver. In this case, most of the other GPS error
sources are common. The only major problem, which needs to be
solved to carry out high precision kinematic GPS, is the integer
ambiguity problem. This is because at any given instant, the whole
number of cycles between the satellite and the receiver is unknown.
The recovery of the unknown whole wavelengths or integer
ambiguities is therefore of great importance to precise phase
positioning. Recently, a large amount of research has focused on
so-called "on the fly" (OTF) ambiguity resolution methodologies in
which the integer ambiguities are solved for while the unknown
receiver is in motion.
The half-second processing time required for this paper represents
44 feet of motion for a vehicle traveling at 60 mph, which would be
intolerable unless supplemented by an inertial navigation system.
The basic guidance system in this case would have to be the laser
or MEMS gyro on the vehicle. With a faster PC, one-tenth a second
processing time would be achievable, corresponding to approximately
10 feet of motion of the vehicle, putting less reliance on the
laser gyroscope. Nowhere in this paper is the use of this system on
automobiles suggested. The technique presented in this paper is a
single epoch basis (OTF) ambiguity resolution procedure that is
insensitive to cycle slips. This system requires the use of five or
more satellites which suggests that additional GPS satellites may
need to be launched to make the smart highway system more
accurate.
F. van Diggelen, "GPS and GPS+GLONASS RTK", ION-GPS, September 1997
"New Products Descriptions", gives a good background of real time
kinematic systems using the carrier frequency. The products
described in this paper illustrate the availability of centimeter
level accuracies for the purposes of the RtZF.TM. system. The
product described in F. van Diggelen requires a base station that
is no further than 20 kilometers away.
A paper by J. Wu and S. G. Lin, titled "Kinematic Positioning with
GPS Carrier Phases by Two Types of Wide Laning", Journal of the
Institute of Navigation, Vol. 44, No. 4, Winter 1997 discusses that
the solution of the integer ambiguity problem can be simplified by
performing other constructs other than the difference between the
two phases. One example is to use three times one phase angle,
subtracted from four times another phase angle. This gives a
wavelength of 162.8 centimeters vs. 86.2 for the single difference.
Preliminary results with a 20-kilometer base line show a success
rate as high as 95% for centimeter level accuracies.
A paper by R. C. Hayward et al., titled "Inertially Aided GPS Based
Attitude Heading Reference System (AHRS) for General Aviation
Aircraft" provides the list of inertial sensors that can be used
with the teachings of embodiments of this invention.
K. Ghassemi et al., "Performance Projections of GPS IIF", describes
the performance objectives for a new class of GPS 2F satellites
scheduled to be launched in late 2001.
Significant additional improvement can be obtained for the WAAS
system using the techniques described in the paper "Incorporation
of orbital dynamics to improve wide-area differential GPS" by J.
Ceva, W. Bertinger, R. Mullerschoen, T. Yunck and B. Parkinson,
Institute on Navigation, Meeting on GPS Technology, Palm Springs,
Calif., September 1995.
Singh, Daljit and Grewal, Harkirat, "Autonomous Vehicle using
WADGPS", discusses ground vehicle automation using wide-area DGPS.
Though this reference describes many of the features of embodiments
of the present invention, it does not disclose sub-meter accuracy
or sub-meter accurate mapping.
U.S. Pat. No. 5,272,483 to Kato describes an automobile navigation
system. This system attempts to correct for the inaccuracies in the
GPS system through the use of an inertial guidance, geomagnetic
sensor, or vehicle crank shaft speed sensor. However, it is unclear
as to whether the second position system is actually more accurate
than the GPS system. This combined system, however, cannot be used
for sub-meter positioning of an automobile.
U.S. Pat. No. 5,383,127 to Shibata uses map matching algorithms to
correct for errors in the GPS navigational system to provide a more
accurate indication of where the vehicle is or, in particular, on
what road the vehicle is. This procedure does not give sub-meter
accuracy. Its main purpose is for navigation and, in particular, in
determining the road on which the vehicle is traveling.
U.S. Pat. No. 5,416,712 to Geier, et al. relates generally to
navigation systems and more specifically to global positioning
systems that use dead reckoning apparatus to fill in as backup
during periods of GPS shadowing such as occur amongst obstacles,
e.g., tall buildings in large cities. This patent shows a method of
optimally combining the information available from GPS even when
less than 3 or 4 satellites are available with information from a
low-cost, inertial gyro, having errors that range from 1 5%. This
patent provides an excellent analysis of how to use a modified
Kalman filter to optimally use the available information.
U.S. Pat. No. 5,606,506 to Kyrtsos provides a good background of
the GPS satellite system. It describes a method for improving the
accuracy of the GPS system using an inertial guidance system. This
is based on the fact that the GPS signals used by Kyrtsos do not
contain a differential correction and the selective access feature
is on. Key paragraphs from this application that describe subject
matter applicable to embodiments of the instant invention
follow.
"Several national governments, including the United States (U.S.)
of America, are presently developing a terrestrial position
determination system, referred to generically as a global
positioning system (GPS). A GPS is a satellite-based
radio-navigation system that is intended to provide highly accurate
three-dimensional position information to receivers at or near the
surface of the Earth.
"The U.S. government has designated its GPS the "NAVSTAR." The
NAVSTAR GPS is expected to be declared fully operational by the
U.S. government in 1993. The government of the former Union of
Soviet Socialist Republics (USSR) is engaged in the development of
a GPS known as "GLONASS". Further, two European systems known as
"NAVSAT" and "GRANAS" are also under development." For ease of
discussion, the following disclosure focuses specifically on the
NAVSTAR GPS. The invention, however, has equal applicability to
other global positioning systems.
"In the NAVSTAR GPS, it is envisioned that four orbiting GPS
satellites will exist in each of six separate circular orbits to
yield a total of twenty-four GPS satellites. Of these, twenty-one
will be operational and three will serve as spares. The satellite
orbits will be neither polar nor equatorial but will lie in
mutually orthogonal inclined planes."
"Each GPS satellite will orbit the Earth approximately once every
12 hours. This coupled with the fact that the Earth rotates on its
axis once every twenty-four hours causes each satellite to complete
exactly two orbits while the Earth turns one revolution."
"The position of each satellite at any given time will be precisely
known and will be continuously transmitted to the Earth. This
position information, which indicates the position of the satellite
in space with respect to time (GPS time), is known as ephemeris
data."
"In addition to the ephemeris data, the navigation signal
transmitted by each satellite includes a precise time at which the
signal was transmitted. The distance or range from a receiver to
each satellite may be determined using this time of transmission
which is included in each navigation signal. By noting the time at
which the signal was received at the receiver, a propagation time
delay can be calculated. This time delay when multiplied by the
speed of propagation of the signal will yield a "pseudorange" from
the transmitting satellite to the receiver."
"The range is called a "pseudorange" because the receiver clock may
not be precisely synchronized to GPS time and because propagation
through the atmosphere introduces delays into the navigation signal
propagation times. These result, respectively, in a clock bias
(error) and an atmospheric bias (error). Clock biases may be as
large as several milliseconds."
"Using these two pieces of information (the ephemeris data and the
pseudorange) from at least three satellites, the position of a
receiver with respect to the center of the Earth can be determined
using passive triangulation techniques."
"Triangulation involves three steps. First, the position of at
least three satellites in "view" of the receiver must be
determined. Second, the distance from the receiver to each
satellite must be determined. Finally, the information from the
first two steps is used to geometrically determine the position of
the receiver with respect to the center of the Earth."
"Triangulation, using at least three of the orbiting GPS
satellites, allows the absolute terrestrial position (longitude,
latitude, and altitude with respect to the Earth's center) of any
Earth receiver to be computed via simple geometric theory. The
accuracy of the position estimate depends in part on the number of
orbiting GPS satellites that are sampled. Using more GPS satellites
in the computation can increase the accuracy of the terrestrial
position estimate."
"Conventionally, four GPS satellites are sampled to determine each
terrestrial position estimate. Three of the satellites are used for
triangulation, and a fourth is added to correct for the clock bias
described above. If the receiver's clock were precisely
synchronized with that of the GPS satellites, then this fourth
satellite would not be necessary. However, precise (e.g., atomic)
clocks are expensive and are, therefore, not suitable for all
applications."
For a more detailed discussion on the NAVSTAR GPS, see Parkinson,
Bradford W. and Gilbert, Stephen W., "NAVSTAR: Global Positioning
System--Ten Years Later, Proceedings of the IEEE, Vol. 71, No. 10,
October 1983; and GPS: A Guide to the Next Utility," published by
Trimble Navigation Ltd., Sunnyvale, Calif., 1989, pp. 147. For a
detailed discussion of a vehicle positioning/navigation system
which uses the NAVSTAR GPS, see commonly owned U.S. patent
application Ser. No. 7/628,560, entitled "Vehicle Position
Determination System and Method," filed Dec. 3, 1990."
"The NAVSTAR GPS envisions two modes of modulation for the carrier
wave using pseudorandom signals. In the first mode, the carrier is
modulated by a "C/A signal" and is referred to as the
"Coarse/Acquisition mode". The Coarse/Acquisition or C/A mode is
also known as the "Standard Positioning Service". The second mode
of modulation in the NAVSTAR GPS is commonly referred to as the
"precise" or "protected" (P) mode. The P-mode is also known as the
"Precise Positioning Service".
The P-mode is intended for use only by Earth receivers specifically
authorized by the U.S. government. Therefore, the P-mode sequences
are held in secrecy and are not made publicly available. This
forces most GPS users to rely solely on the data provided via the
C/A mode of modulation (which results in a less accurate
positioning system) "In addition to the clock error and atmospheric
error, other errors which affect GPS position computations include
receiver noise, signal reflections, shading, and satellite path
shifting (e.g., satellite wobble). These errors result in
computation of incorrect pseudoranges and incorrect satellite
positions. Incorrect pseudoranges and incorrect satellite
positions, in turn, lead to a reduction in the precision of the
position estimates computed by a vehicle positioning system."
U.S. Pat. No. 5,757,646 to Talbot, et al. illustrates the manner in
which centimeter level accuracy on the fly in real time is
obtained. It is accomplished by double differencing the code and
carrier measurements from a pair of fixed and roving GPS receivers.
This patent also presents an excellent discussion of the problem
and various prior solutions as in the following paragraphs:
"When originally conceived, the global positioning system (GPS)
that was made operational by the United States Government was not
foreseen as being able to provide centimeter-level position
accuracies. Such accuracies are now commonplace."
"Extremely accurate GPS receivers depend on phase measurements of
the radio carriers that they receive from various orbiting GPS
satellites. Less accurate GPS receivers simply develop the
pseudoranges to each visible satellite based on the time codes
being sent. Within the granularity of a single time code, the
carrier phase can be measured and used to compute range distance as
a multiple of the fundamental carrier wavelength. GPS signal
transmissions are on two synchronous, but separate carrier
frequencies "L1" and "L2", with wavelengths of nineteen and
twenty-four centimeters, respectively. Thus, within nineteen or
twenty-four centimeters, the phase of the GPS carrier signal will
change 360.degree.."
"However the numbers of whole cycle (360.degree.) carrier phase
shifts between a particular GPS satellite and the GPS receiver must
be resolved. At the receiver, every cycle will appear the same.
Therefore there is an "integer ambiguity". The computational
resolution of the integer ambiguity has traditionally been an
intensive arithmetic problem for the computers used to implement
GPS receivers. The traditional approaches to such integer ambiguity
resolution have prevented on-the-fly solution measurement updates
for moving GPS receivers with centimeter accurate outputs. Very
often such highly accurate GPS receivers have required long periods
of motionlessness to produce a first and subsequent position
fix."
"There are numerous prior art methods for resolving integer
ambiguities. These include integer searches, multiple antennas,
multiple GPS observables, motion-based approaches, and external
aiding. Search techniques often require significant computation
time and are vulnerable to erroneous solutions when only a few
satellites are visible. More antennas can improve reliability
considerably. If carried to an extreme, a phased array of antennas
results whereby the integers are completely unambiguous and
searching is unnecessary. But for economy the minimum number of
antennas required to quickly and unambiguously resolve the
integers, even in the presence of noise, is preferred."
"One method for integer resolution is to make use of the other
observables that modulate a GPS timer. The pseudo-random code can
be used as a coarse indicator of differential range, although it is
very susceptible to multipath problems. Differentiating the L1 and
L2 carriers provides a longer effective wavelength, and reduces the
search space. However dual frequency receivers are expensive
because they are more complicated. Motion-based integer resolution
methods make use of additional information provided by platform or
satellite motion. But such motion may not always be present when it
is needed."
This system is used in an industrial environment where the four
antennas are relatively close to each other. Practicing teachings
of this invention permits a navigational computer to solve for the
position of the rover vehicle to within a few centimeters on the
fly ten times a second. An example is given where the rover is an
airplane.
The above comments related to the use of multiple antennas to
eliminate the integer ambiguity suggest that if a number of
vehicles are nearby and their relative positions are known, the
ambiguity can be resolved in this manner.
2.4 Inertial Navigation System
An example of how various sensors other than the GPS and PPS
systems described in this invention can be found in "Magnetometer
and Differential Carrier-Phase GPS-Aided INS for Advanced Vehicle
Control", IEEE Transactions on Robotics and Automation, Vol. 19,
No. 2, April 2003.
3. Maps and Mapping
It is intended that the map database of embodiments of the instant
invention will conform to the open GIS specification. This will
permit such devices to additionally obtain on-line consumer
information services such as driving advisories, digital yellow
pages that give directions, local weather pictures and forecasts
and video displays of local terrain since such information will
also be in the GIS database format.
A paper by O'Shea, Michael and Shuman, Valerie entitled "Looking
Ahead: Map Databases in Predictive Positioning and Safety Systems"
discusses map databases which can assist radar and image-processing
systems of this invention since the equipped vehicle would know
where the road ahead is and can therefore distinguish the lane of
the preceding vehicle. No mention, however, is made in this
reference of how this is accomplished through range gating or other
means. This reference also mentions that within five years it may
be possible to provide real time vehicle location information of
one-meter accuracy. However, it mentions that this will be limited
to controlled access roads such as interstate highways. In other
words, the general use of this information on all kinds of roads
for safety purposes is not contemplated. This reference also states
that "road geometry, for example, may have to be accurate to within
one meter or less as compared to the best available accuracy of 15
meters today". This reference also mentions the information about
lane configuration that can be part of the database including the
width of each lane, the number of lanes, etc., and that this can be
used to determine driver drowsiness. This reference also states
that "at normal vehicle speeds, the vehicle location must be
updated every few milliseconds". It is also stated that the
combination of radar and map data can help to interpret radar
information such as the situation where a radar system describes an
overpass as a semi truck. Image processing in this reference is
limited to assessing road conditions such as rain, snow, etc. The
use of a laser radar system, for example, is not contemplated by
this reference. The use of this information for road departures
warnings is also mentioned, as is lane following. The reference
also mentions that feedback from vehicles can be used to improve
map configurations.
A great flow of commercially available data will begin with the new
generation of high resolution (as fine as about 1 meter) commercial
earth imaging satellites from companies like EarthWatch and SPOT
Image. Sophisticated imaging software is being put in place to
automatically process these imaging streams into useful data
products. This data can be used to check for gross errors in the
map database.
According to Al Gore, in "The Digital Earth: Understanding our
Planet in the 21.sup.st Century", California Science Center, Jan.
31, 1998, the Clinton Administration licensed commercial satellites
to provide one meter resolution imaging beginning in 1998. Such
imaging can be combined with digital highway maps to provide an
accuracy and reality check.
U.S. Pat. No. 5,367,463 to Tsuji describes a vehicle azimuth
determining system. It uses regression lines to find the vehicle on
a map when there are errors in the GPS and map data. This patent
does not give sub-meter accuracy. The advantage of this invention
is that it shows a method of combining both map matching data and
GPS along with a gyroscope and vehicle velocity and odometer data
to improve the overall location accuracy of the vehicle.
4. Precise Positioning
For the purposes herein, a Precise Positioning Station, or PPS,
will mean any system that involves the existence of or placement of
a detectable infrastructure on or near a roadway that when used in
conjunction with an accurate map permits a vehicle to determine its
precise location. In other words, PPS can be any system that can
recognize anything in or on the infrastructure and thereby, in
conjunction with an accurate map, can locate the vehicle. Such
detectable infrastructure can comprise a MIR triad, radar
reflectors, SAW devices, RFID devices, devices or marks detectable
visibly such as bar codes or other recognizable objects including
edges of buildings, poles, signs or the like, magnetic markers or
any other object whose position is precisely known and/or is
detectable in a manner that permits the vehicle to determine its
position relative to the device or absolutely and where the object
is noted on a map database residing within the vehicle. An
alternative procedure is to map the reflective signature of the
road environment and, using a laser, radar, terahertz or similar
system, a vehicle can compare the sensed reflective signature with
that recorded and thereby determine its location. As the
environment changes with the seasons, there will be segments of the
signature that are unreliable but since a reasonable adjustment
distance might be once per mile it is quite likely that somewhere
during a mile of travel that the reflective signature will be
invariant over time. Bridge abutments, roadside signs or light
poles, for example, would not typically change from one time of the
year to another and thus could be used as quite accurate markers of
position along the road. Such a system has the advantage that no
additions to the infrastructure would be required. When PPS or
Precise Positioning Station is referred to below, it is generally
intended to include all of these devices and/or methods.
If two vehicles are traveling near each other and have established
communications, and assuming that each vehicle can observe at least
four of the same GPS satellites, each vehicle can send the
satellite identification and the time of arrival of the signal at a
particular epoch to the other. Then, each vehicle can determine the
relative position of the other vehicle as well as the relative
clock error. As one vehicle passes a Precise Positioning Station
(PPS), it knows exactly where it is and thus the second vehicle
also knows exactly where it is and can correct for satellite
errors. All vehicles that are in communication with the vehicle at
the PPS similarly can determine their exact position and the system
approaches perfection. This concept is based on the fact that the
errors in the satellite signals are identical for all vehicles that
are within a mile or so of each other. Furthermore, each vehicle
can set its onboard clock since the vehicle passing the PPS can do
so, and communicate the exact time to the others, and then each
vehicle can know the carrier phase of each satellite signal at the
PPS and thus invoke carrier phase DGPS.
When the operator begins operating his vehicle with a version of
the RtZF.TM. system of this invention, he or she will probably not
be near a reference point as determined by one of the radar
reflector, MIR or RFID or other landmark locator systems as
discussed below as part of this invention, for example. In this
situation, he or she will use the standard GPS system with the WAAS
or other DGPS corrections such as available from OmniStar.TM., the
U.S. Government or other provider. This will provide accuracy of
between a few meters to 6 centimeters. This accuracy might be
further improved as he or she travels down the road through
map-matching or through communication with other vehicles. The
vehicle will know, however, that is not operating in the high
accuracy mode. As soon as the vehicle (vehicle #1) passes a radar
reflector, SAW, MIR, RFID or equivalent precise positioning system,
it will be able to calculate exactly where it is within a few
centimeters and the vehicle will know that it is in the accurate
mode. Similarly, when another vehicle passes through a precise
positioning station and learns its precise location it can
communicate this fact with other vehicles in its vicinity (5 miles,
for example) along with the latest GPS satellite transmissions.
Each other vehicle will then be able to calculate its relative
location extremely accurately and thus know its position almost as
accurately as the vehicle that just passed through the precise
positioning station. Furthermore, if vehicle #1 also has an
accurate clock, as further described below, it can record the phase
of each carrier wave from each satellite and predict that phase for
perhaps an hour into the future. This then permits vehicle #1 to
switch to carrier phase DGPS and know its precise position relative
to the precise positioning station, and thus on the earth, until
the clock accuracy degrades its knowledge of the carrier phase at
the precise positioning station. Through continuous communication
between vehicle #1 and other vehicles, all vehicles in the vicinity
can similarly operate in the carrier phase DGPS mode without the
need for the installation and maintenance of local DGPS stations.
Thus, the addition of a few precise positioning stations at very
low cost permits each vehicle traveling on the road to know its
precise location on the earth and for the system to approach
perfection, a necessary requirement for achieving zero fatalities.
For high-speed travel on a controlled highway, frequent precise
positioning stations can be inexpensively provided and each vehicle
can thereby be accurately contained within its proper corridor.
Also, the size of the corridors that the vehicle is permitted to
travel in can be a function of the accuracy state of the
vehicle.
A paper by Han, Shaowei entitled "Ambiguity Recovery For Long-Range
GPS Kinematic Positioning" appears to say that if a mobile receiver
is initially synchronized with a fixed receiver such that there is
no integer ambiguity, and if the mobile receiver then travels away
from the fixed receiver, and during the process it loses contact
with the satellites for a period of up to five minutes, that the
carrier phase can be recovered and the ambiguity eliminated,
providing again centimeter-range accuracies. Presumably, the fixed
station is providing the differential corrections. This is
important for embodiments of the instant invention since the
integer ambiguity can be eliminated each time the vehicle passes a
Precise Positioning Station (PPS) as explained below. After that, a
five-minute loss of GPS signals should never occur. Thus, carrier
phase accuracies will eventually be available to all vehicles. Note
that the integer ambiguity problem disappears when the GPS
satellites provide more frequencies. If, for example, each
satellite would broadcast two frequencies with each frequency being
a prime number of cycles per second, there would be no integer
ambiguity problem. Due to the problem of identifying large prime
numbers, other schemes can be used such that the relative phase of
one carrier to the other does not repeat in the space from the
vehicle to the satellite or if it does repeat, it repeats only a
few times. This problem becomes simpler as more frequencies are
added as for three frequencies, for example, the phase relation
between any two can repeat as long as the phase relationships
between all three don't repeat very often. Also, with multiple
frequencies the DGPS corrections become less important and in some
cases may not be needed. This is because each frequency is
diffracted a different amount by the ionosphere and therefore the
diffraction or cash frequency can be determined. A new civilian
frequency is scheduled to be introduced by the U.S. Government as
part of the NAVSTAR system and the forthcoming European GALILEO
system is planned to have multiple frequencies for civilian
use.
U.S. Pat. No. 5,361,070 to McEwan, although describing a motion
detector, discusses technology which is used as part of a system to
permit a vehicle to precisely know where it is on the face of the
earth at particular locations. The ultra wideband 200 picosecond
radar pulse emitted by the low power radar device of McEwan is
inherently a spread spectrum pulse which generally spans hundreds
of megahertz to several gigahertz. A frequency allocation by the
FCC is not relevant. Furthermore, many of these devices may be
co-located without interference. The concept of this device is
actually discussed in various forms in the following related
patents to McEwan. The following comments will apply to these
patents as a group.
U.S. Pat. No. 5,510,802 to McEwan describes a time of flight
radio-location system similar to what is described below. In this
case, however, a single transmitter sends out a pulse, which is
received by three receivers to provide sub-millimeter resolution.
The range of this device is less than about 10 feet.
The concept described in McEwan's U.S. Pat. No. 5,519,400 is that
the MIR signal can be modulated with a coded sequence to permit
positive identification of the sending device. In an additional
McEwan patent, U.S. Pat. No. 5,589,838, a short-range
radio-location system is described. Additionally, in U.S. Pat. No.
5,774,091, McEwan claims that the MIR system will operate to about
20 feet and give resolutions on the order of 0.01 inches.
5. Radar and Laser Radar Detection and Identification of Objects
External to the Vehicle
The RtZF system described herein can include an energy beam or
flood that is projected from the vehicle into the environment for
the purpose of illuminating the environment around the vehicle and
objects therein. In some cases this can be a beam of radar
operating at 24 GHz or 77 GHz, for example. In other cases, this
can be a laser beam in the infrared portion of the spectrum. Other
frequencies can also be used and there are particularly interesting
developments in the terahertz frequency range. Terahertz devices
are under development that can create a terahertz beam of radiation
using laser technology. Similarly devices are now available for
sensing terahertz radiation with an array of pixels. The terahertz
frequency is interesting for interrogating the vicinity of a
vehicle since it can be transmitted in a very narrow beam like a
laser and yet it has the ability to penetrate fog, for example,
more like radar, but not nearly as good as radar, thus providing
the advantages of both systems. In the form of a flood light to
illuminate areas closer to the vehicle for blind spot interrogation
or for use as a headlight for animal and pedestrian identification
is also interesting since such a system would work in both daylight
and at night since there is little natural radiation in the
terahertz part of the electromagnetic spectrum. When used as a
beam, terahertz will be referred herein as terahertz radar. For the
purposes herein, the terahertz frequency range will be taken as the
range from about 300 GHz (0.3 THz) to about 3000 GHz (3 THz) which
is about where the infrared range begins.
Lasers, such as infrared lasers, can be used in beams of varying
diameter and divergence angles through the appropriate optics
providing the energy of the beam per square millimeter remains
below the eye safety limits set by the U.S. Government. Thus, a
very narrow beam can be used in a scanning fashion, in which case,
in the limit a single pixel can be used as the detector such as a
photodiode or avalanche diode. In other cases, a high-powered diode
laser can still emit radiation below the eye safe limits if the
beam is expanded through appropriate optics, in which case, a
multi-pixel detector such as a CCD or CMOS imager can be used. In
both cases, a particular range from the vehicle can be interrogated
and imaged, as discussed below herein, through range gating.
Although this concept was originally disclosed in the patents and
patent applications referenced above and assigned to the current
assignee of this patent application, several recent patents and
publications have also disclosed some features. Some of these
related art patents and publications will now be discussed.
One method of achieving range gating is disclosed in WO9701111 and
that publication discloses other prior art range gating methods
that have been used to obtain three dimensional information of a
scene. As mentioned elsewhere herein, other systems use liquid
crystals, garnet crystals, Pockel and Kerr cells. Prior to the
disclosure in the assignee's patents, none of these methods have
been used in the automotive environment for obtaining
three-dimensional information about objects within or outside of a
vehicle. Nevertheless, the use of the optical ranging apparatus and
techniques disclosed in this and other patents and publications of
3DV Systems Ltd. are among the preferred methods used in practicing
the teachings of the instant invention. Other patents and
publications of 3DV include: WO9701112, WO9701113, U.S. Pat. No.
6,327,073, U.S. Pat. No. 6,483,094 and U.S. Pat. No.
20020185590.
U.S. Pat. No. 5,791,757, U.S. Pat. No. 5,857,770, U.S. Pat. No.
5,890,796, U.S. Pat. No. 5,971,578 and U.S. Pat. No. 6,036,340 are
patents assigned to Ford Global Technologies, LLC relating to the
use of a high power laser diode in the visible portion of the
spectrum for automotive headlight and tail light application. These
patents are significant in that they show how to implement such a
device so that little space is occupied. These patents do not
disclose the use of a laser diode for interrogating the space
adjacent to or at a distance from the vehicle for any purpose such
as collision avoidance. The optics system illustrated could be
usable in implementing one or more of the inventions disclosed
herein as are other optical systems.
U.S. Pat. No. 6,690,017, also assigned to Ford Global Technologies,
LLC, relates to the use of a high-power infrared laser diode in
conjunction with a display for night vision applications. In order
to avoid blinding a similar system of an approaching vehicle, the
illumination created is synchronized based of the GPS clock and the
direction that the vehicle is traveling. In at least one invention
disclosed herein, the GPS clock is also utilized to control the
time of transmission of an IR interrogating illumination but since
the distance to the object being interrogated in important, and
since vehicles traveling in the same direction as the subject
vehicle may also have similar apparatus, the transmission is
synchronized so as not to interfere with such similar systems as
discussed below.
U.S. Pat. No. 6,725,139, also assigned to Ford Global Technologies,
LLC, relates to a method of controlling the direction of infrared
illumination based on the steering direction of the vehicle. This
night vision system is used in conjunction with a display to
supplement the normal headlights. Although inventions described
herein disclose changing the direction of projected illumination,
in general it is not tied to the direction of the steering
wheel.
U.S. Pat. No. 6,730,913, U.S. Pat. No. 6,774,367 and U.S. Pat. No.
6,809,870, similarly assigned to Ford Global Technologies, LLC,
describe a night vision system using range gating to determine the
location of objects in the field of view. The general disclosure of
these patents is believed to be anticipated by assignee's patents
referenced above. The patents describe three methods of obtaining
equal illumination for varying distances, varying the intensity of
the transmitted pulses, varying the camera sensitivity or varying
the number of pulses used to interrogate a particular range. It is
well known in the art that to obtain an image of sufficient
brightness to permit display or image analysis, sufficient
illumination must be supplied. These patents, however, purport to
provide equal brightness for all objects regardless of their
distance from the camera. Also, these patents are based on the
concept that a series of gradually increasing ranges will always be
sequentially interrogated whereas in the instant invention the
interrogation method will not necessarily follow such a scheme and
the location of the roadway as known from accurate maps will often
be used to determine the range and direction of the interrogating
illumination. Another variable, not discussed in the '913 patent is
the variation of the transmission angular field of view of the
illumination as is used in some applications of the current
invention discussed below. Further, at least one of the current
inventions disclosed herein can be used for illuminating and
identifying objects external to the vehicle both in daytime and at
night. Note also that the concept of triggering illumination out of
phase to prevent one vehicle's system from interfering with
another's as disclosed in the '367 patent was previously disclosed
in assignee's patents and patent applications.
U.S. Pat. No. 6,429,429 and patent application publications U.S.
20030034462, U.S. 20030036881 and U.S. 20030155513 also assigned to
Ford Global Technologies, LLC, describe various night vision
systems some of which use range gating and time-of-flight methods
as first disclosed in the assignee's patents cross referenced
above.
A paper by Amamoto, Naohiro and Matsumoto, Koji entitled
"Obstruction Detector By Environmental Adaptive Background Image
Updating" describes a method for distinguishing between moving
object pixels, stationary object pixels, and pixels that change due
to illumination changes in a video image. This paper appears to
handle the case of a camera fixed relative to the earth, not one
mounted on a vehicle. This allows the system to distinguish between
a congested area and an area where cars are moving freely. The
video sampling rate was 100 milliseconds.
A paper by Doi, Ayumu, Yamamomo, Yasunori, and Butsuen, Tetsuro
entitled "Development Of Collision Warning System and Its Collision
Avoidance Effect" describes a collision warning system that has
twice the accuracy of conventional systems. It uses scanning a
laser radar. In the system described in this paper, the authors do
not appear to use phase measurements, range gating or time of
flight to separate one vehicle from another.
A paper by Min, Joon, Cho, Hyung, and Choi, Jong, entitled "A
Learning Algorithm Using Parallel Neuron Model" describes a method
of accurately categorizing vehicles based on the loop in the
highway. This system uses a form of neural network, but not a back
propagation neural network. This would essentially be categorizing
a vehicle by its magnetic signature. Much information is lost in
this system, however, due to the lack of knowledge of the vehicle's
velocity.
Work has been done at JPL (Jet Propulsion Laboratories) to develop
a target recognition system. Neural networks play a key role in
this target recognition process. The recognition of vehicles on a
roadway is a considerably simpler process. Most of the cluttering
information can be eliminated through range gating. The
three-dimensional image obtained as described below will permit
simple rotations of the image to artificially create a frontal view
of the object being investigated. Also, the targets of interest
here are considerably closer than was considered by JPL.
Nevertheless, the techniques described in this reference and in the
references cited by this reference, are applicable here in a
simplified form. The JPL study achieved over a 90% success rate at
60 frames per minute.
U.S. Pat. No. 4,521,861 to Logan describes a method and apparatus
for enhancing radiometric in-aging and a method and apparatus for
enhancing target detection through the utilization of an imaging
radiometer. The radiometer, which is a passive thermal receiver,
detects the reflected and emitted thermal radiation of targets.
Prior to illumination, foliage will appear hot due to its high
emissivity and metals will appear cold due to their low
emissivities. When the target is momentarily illuminated foliage
appears dark while metals appear hot. By subtracting the
non-illuminated image from the illuminated image, metal targets are
enhanced. The teachings of this patent thus have applicability to
embodiments of the instant invention as discussed below.
U.S. Pat. No. 5,463,384 to Juds uses a plurality of infrared beams
to alert a truck driver that a vehicle is in his blind spot when he
begins to turn the vehicle. The system is typically activated by
the vehicle's turn signal. No attempt is made to measure exactly
where the object is, only whether it is in the blind spot or
not.
U.S. Pat. No. 5,467,072 to Michael relates to a phased array radar
system that permits the steering of a radar beam without having to
rotate antennas. Aside from that, it suffers from all the
disadvantages of radar systems as described here. In particular, it
is not capable of giving accurate three-dimensional measurements of
an object on the roadway.
U.S. Pat. No. 5,486,832 to Hulderman employs millimeter wave radar
and optical techniques to eliminate the need for a mechanical
scanning system. A 35-degree arc is illuminated in the azimuth
direction and 6 degrees in elevation. The reflected waves are
separated into sixteen independent, simultaneously overlapping 1.8
degree beams. Each beam, therefore, covers a width of about 3 feet
at 100 feet distance from the vehicle, which is far too large to
form an image of the object in the field of view. As a result, it
is not possible to identify the objects in the field of view. All
that is known is that an object exists. Also, no attempt has been
made to determine whether the object is located on the roadway or
not. Therefore, this invention suffers from the limitations of
other radar systems.
U.S. Pat. No. 5,530,447 to Henderson, et al. shows a system used to
classify targets as threatening or non-threatening, depending on
whether the target is moving relative to the ground. This system is
only for vehicles in an adjacent lane and is primarily meant to
protect against blind-spot type accidents. No estimation is made by
the system of the position of the target vehicle or the threatening
vehicle, only its relative velocity.
U.S. Pat. No. 5,576,972 to Harrison provides a good background of
how neural networks are used to identify various objects. Although
not directly related to intelligent transportation systems or to
accident-avoidance systems such as described herein, these
techniques will be applied to embodiments of the invention
described herein as discussed below.
U.S. Pat. No. 5,585,798 to Yoshioka, et al. uses a combination of a
CCD camera and a laser radar unit. The invention attempts to make a
judgment as to the danger of each of the many obstacles that are
detected. The load on the central processor is monitored by looking
at different obstacles with different frequencies depending on
their danger to the present system. A similar arrangement is
contemplated for embodiments of the invention as disclosed
herein.
U.S. Pat. No. 5,767,953 to McEwan describes a laser tape measure
for measuring distance. It is distinct from laser radars in that
the width of the pulse is measured in sub-nanosecond times, whereas
laser radars are typically in the microsecond range. The use of
this technology in the current invention would permit a much higher
scanning rate than by convention radar systems and thus provide the
opportunity for obtaining an image of the obstructions on the
highway. It is also less likely that multiple vehicles having the
same system would interfere with each other. For example, if an
area 20 feet by 5 feet were scanned with a 0.2 inch pixel size,
this would give about one million pixels. If using laser radar, one
pixel per microsecond is sent out, it would take one second to scan
the entire area during which time the vehicle has traveled 88 feet
at 60 miles an hour. On the other hand, if scanning this array at
100 feet, it would take 200 nanoseconds for the light to travel to
the obstacle and back. Therefore, if a pulse is sent out every
fifth of a microsecond, it will take a fifth of a second to obtain
a million pixels, during which time the vehicle has traveled about
17 feet. If 250,000 pixels are used, the vehicle will only have
traveled about 4 feet.
U.S. Pat. No. 4,352,105 and U.S. Pat. No. 4,298,280 to Harney
describe an infrared radar system and a display system for use by
aircraft. In particular, these patents describe an infrared radar
system that provides high resolution, bad weather penetration,
day-night operation and which can provide simultaneous range,
intensity and high resolution angular information. The technology
uses CO.sub.2 laser and a 10.6 micron heterodyne detection. It is a
compact imaging infrared radar system that can be used with
embodiments of the invention described herein. Harney applies this
technology to aircraft and does not contemplate its application to
collision avoidance or for other uses with land-based vehicles such
as automobiles.
Although, there appears not to be any significant prior art
involving a vehicle communicating safety information to another
vehicle on the roadway, several patents discuss methods of
determining that a collision might take place using infrared and
radar. U.S. Pat. No. 5,249,128 to Markandey et al., for example,
discusses methods of using infrared to determine the distance to a
vehicle in front and U.S. Pat. No. 5,506,584 to Boles describes a
radar-based system. Both systems suffer from a high false alarm
rate and could be substantially improved if a pattern recognition
system such as neural networks were used. Also, neither system
makes use of noise modulation technologies as taught herein.
6. Smart Highways
A paper entitled "Precursor Systems Analyses of Automated Highway
Systems (Executive Summary)" discusses that "an AHS (automated
highway system) can double or triple the efficiency of today's most
congested lanes while significantly increasing safety and trip
quality".
There are one million, sixty-nine thousand, twenty-two miles of
paved non-local roads in the US. Eight hundred twenty-one thousand
and four miles of these are classified as "rural" and the remaining
two hundred forty-eight thousand, eighteen miles are "urban".
The existing interstate freeway system consists of approximately
50,000 miles which is 1% of the total of 3.8 million miles of
roads. Freeways make up 3% of the total urban/suburban arterial
mileage and carry approximately 30% of the total traffic.
In one study, dynamic route guidance systems were targeting at
reducing travel time of the users by 4%. Under the system of this
invention, the travel times would all be known and independent of
congestion once a vehicle had entered the system. Under the current
system, the dynamic delays can change measurably after a vehicle is
committed to a specific route. According to the Federal Highway
Administration Intelligent Transportation Systems (ITS Field
Operational Test), dynamic route guidance systems have not been
successful.
There are several systems presented in the Federal Highway
Administration Intelligent Transportation Systems (ITS Field
Operational Test) for giving traffic information to commuters,
called "Advance Traveler Information System" (ATIS). In none of
these articles does it discuss the variation in travel time during
rush hour for example, from one day to the next. The variability in
this travel time would have to be significant to justify such a
system. A system of this type would be unnecessary in situations
where embodiments of the instant invention has been deployed. The
single most important cause of variability from day to day is
traffic incidents such as accidents, which are eliminated or at
least substantially reduced by the instant invention. One of the
conclusions in a study published in the "Federal Highway
Administration Intelligent Transportation Systems (ITS Field
Operational Test)" entitled "Direct Information Radio Using
Experimental Communication Technologies" was that drivers did not
feel that the system was a significant advance over commercial
radio traffic information. They did think the system was an
improvement over television traffic information and changeable
message signs. The drivers surveyed on average having changed their
route only one time in the eight week test period due to
information they received from the system.
7. Weather and Road Condition Monitoring
A paper by Miyata, Yasuhiro and Otomo, Katsuya, Kato, Haijime,
Imacho, Nobuhiro, Murata, Shigeo, entitled "Development of Road
Icing Prediction System" describes a method of predicting road
icing conditions several hours in advance based on an optical fiber
sensor laid underneath the road and the weather forecast data of
that area.
There is likely a better way of determining ice on the road than
described in this paper. The reflection of an infrared wave off the
road varies significantly depending on whether there is ice on the
road or snow, or the road is wet or dry. A neural network could be
a better solution. The system of this paper measures the road
surface temperature, air temperature and solar radiation. A
combination of active and passive infrared would probably be
sufficient. Perhaps, a specially designed reflective surface could
be used on the road surface in an area where it is not going to be
affected by traffic.
What this paper shows is that if the proper algorithm is used, the
actual road temperature can be predicted without the need to
measure the road surface temperature. This implies that icing
conditions can be predicted and the sensors would not be necessary.
Perhaps, a neural network algorithm that monitors a particular
section of road and compares it to the forecasted data would be all
that is required. In other words, given certain meteorological
data, the neural network ought to be able to determine the
probability of icing. What is needed, therefore, is to pick a
section of roadway and monitor that roadway with a state-owned
vehicle throughout the time period when icing is likely to occur
and determine if icing has occurred and compare that with the
meteorological data using a neural network that is adapted for each
section of road.
8. Communication with Other Vehicles
The RtZF.TM. system of this invention can incorporate
vehicle-to-vehicle communication allowing vehicles to inform other
vehicles of their location, velocity, mass etc.
U.S. Pat. No. 5,506,584 to Boles relates to a system for
communication between vehicles through a transmit and transponder
relationship. The patent mentions that there may be as many as 90
vehicles within one half mile of an interrogation device in a
multi-lane environment, where many of them may be at the same or
nearly the same range. Boles utilizes a transponder device, the
coded responses which are randomized in time, and an interrogation
device which processes the return signals to provide vehicle
identification, speed, location and transponder status information
on vehicles to an operator or for storage in memory. No mention is
made of how a vehicle knows its location or how accurate that
knowledge is and therefore how it can transmit that location to
other vehicles.
U.S. Pat. No. 5,128,669 to Dabbs provides for two-way communication
and addressing messages to specific vehicles. This is unnecessary
and the communications can be general since the amount of
information that is unique to one vehicle is small. A method of
handing bidirectional communication is discussed in U.S. Pat. No.
5,506,584 to Boles. The preferred vehicle-to-vehicle communication
system using pseudonoise techniques is more thoroughly discussed
below.
In embodiments of the invention described herein,
vehicle-to-vehicle communication is used, among other purposes, to
allow the fact that one vehicle knows its position more accurately
than another to use communication to cause the other vehicle to
also improve the accuracy with which it knows its position.
9. Infrastructure-to-Vehicle Communication
The RtZF.TM. system of this invention can also incorporate
communication between a vehicle and infrastructure for a variety of
reasons including obtaining the latest map updates, weather
conditions, road conditions, speed limits, sign contents, accidents
ahead, congestion ahead, construction, general Internet access and
for many other reasons.
The DGPS correction information can be broadcast over the radio
data system (RDS) via FM transmitters for land use. A company
called Differential Correction, Inc. has come up with a technique
to transmit this DGPS information on the RDS channel. This
technique has been used in Europe since 1994 and, in particular,
Sweden has launched a nationwide DPGS service via the RDS (see,
Sjoberg, Lars, "A `1 Meter` Satellite Based Navigation Solutions
for the Mobile Environment That Already Are Available Throughout
Europe"). This system has the potential of providing accuracies on
the premium service of between about 1 and 2 meters. A 1 meter
accuracy, coupled with the carrier phase system to be described
below, provides an accuracy substantially better than about 1 meter
as preferred in the Road to Zero Fatalities.TM. (RtZF.TM.) system
of this invention.
In addition to the FM RDS system, the following other systems can
be used to broadcast DGPS correction data: cellular mobile phones,
satellite mobile phones, satellite Internet, WiFi, WiMAX, MCA
(multi-channel access), wireless tele-terminals, DARCs/RBDS (radio
data systems/radio broadcast data system), type FM sub-carrier,
exclusive wireless, and pagers. In particular, DARC type is used
for vehicle information and communication systems so that its
hardware can be shared. Alternately, the cellular phone system,
coupled with the Internet, could be used for transmitting
corrections (see, Ito, Toru and Nishiguchi, Hiroshi entitled
"Development of DGPS using FM Sub-Carrier For ITS"). Primarily, as
discussed elsewhere, vehicle-to-vehicle communications can be used
to transmit DGPS corrections from one vehicle to another whether
the source is a central DGPS system or one based on PPS or other
system.
One approach for the cellular system is to use the GSM mobile
telephone system, which is the Europe-wide standard. This can be
used for transmitting DGPS and possibly map update information
(see, Hob, A., Ilg, J. and Hampel, A. entitled "Integration
Potential Of Traffic Telematics).
In Choi, Jong and Kim, Hoi, "An Interim Report: Building A Wireless
Internet-Based Traveler's Information System As A Replacement Of
Car Navigation Systems", a system of showing congestion at
intersections is broadcast to the vehicle through the Internet. The
use of satellites is discussed as well as VCS system.
This is another example of the use of the Internet to provide
highway users with up-to-date traffic congestion information.
Nowhere in this example, however, is the Internet used to transmit
map information. In fact, once there is an Internet or equivalent
connection to a vehicle, then other information can be transmitted
such as updated map information, weather and visibility, local
conditions ahead, accident information, congestion information,
DGPS corrections, etc. In fact, with a high bandwidth Internet
connection, much of the computations, especially safety related
computations, can best be done on the Internet where the system
reliability would exceed that of a vehicle-based system. The
forecast that "the network is the computer" will begin to become
reality. The crash of a safety related processor due to a software
bug could not be tolerated in a safety related system and would be
less likely to occur if the critical computations occur on the
network. Furthermore, upgrades to vehicle-based software also
become feasible over such a high bandwidth connection.
A paper by Sheu, Dennis, Liaw, Jeff and Oshizawa, Al, entitled "A
Communication System For In-Vehicle Navigation System" provides
another description of the use of the Internet for real traffic
information. However, the author (unnecessarily) complicates
matters by using push technology which isn't absolutely necessary
and with the belief that the Internet connection to a particular
vehicle to allow all vehicles to communicate, would have to be
stopped which, of course, is not the case. For example, consider
the @home network where everyone on the network is connected all
the time.
A paper by Rick Schuman entitled "Progress Towards Implementing
Interoperable DSRC Systems In North America" describes the
standards for dedicated short-range communications (DSRC). DSRC
could be used for inter-vehicle communications, however, its range
according to the ITS proposal to the Federal Government would be
limited to about 90 meters although there have been recent
proposals to extend this to about 1000 meters. Also, there may be a
problem with interference from toll collection systems, etc.
According to this reference, however, "it is likely that any
widespread deployment of intersection collision avoidance or
automated highways would utilize DSRC". Ultra wide band
communication systems, on the other hand, are a viable alternative
to DSRC as explained below. The DSRC physical layer uses microwaves
in the 902 to 928 megahertz band. However, ITS America submitted a
petition to the FCC seeking to use the 5.85 to 5.925 gigahertz band
for DSRC applications.
A version of CDPD, which is a commercially available mobile,
wireless data network operated in the packet-switching mode,
extends Internet protocol capabilities to cellular channels. This
is reported on in a paper entitled "Intelligent Transportation
Systems (ITS) Opportunity".
According to a paper by Kelly, Robert, Povich, Doublas and Poole,
Katherine entitled "Petition of Intelligent Transportation Society
of America for Amendment Of The Commission's Rules to Add
Intelligent Transportation Services (ITS) As A New Mobile Service
With Co-Primary Status In The 5.850 to 5.925 GHz", from 1989 to
1993 police received an annual average of over 6.25 million vehicle
accident reports. During this same period, the total comprehensive
cost to the nation of motor vehicle accidents exceeded the annual
average of 400 billion dollars. In 1987 alone, Americans lost over
2 billion hours (approximately 22,800 years) sitting in traffic
jams. Each driver in Washington D.C. wastes an average of 70 hours
per year idling in traffic. From 1986 to 1996, car travel has
increased almost 40% which amounts to about a 3.4% increase per
year.
Further, from Kelly et al., the FCC has allocated in Docket 94 124,
46.7 to 46.9 GHz and 76 to 77 GHz bands for unlicensed vehicular
collision avoidance radar. The petition for DSRC calls for a range
of up to about 50 meters. This would not be sufficient for the
RtZF.TM. system. For example, in the case of a car passing another
car at 150 kilometers per hour. Fifty meters amounts to about one
second, which would be insufficient time for the passing vehicle to
complete the passing and return to the safe lane. Something more in
the order of about 500 meters would be more appropriate. This,
however, may interfere with other uses of DSRC such as automatic
toll taking, etc., thus DSRC may not be the optimum communication
system for communication between vehicles. DSRC is expected to
operate at a data rate of approximately 600 kbps. DSRC is expected
to use channels that are six megahertz wide. It might be possible
to allocate one or more of the six megahertz channels to the
RtZF.TM. system.
On DSRC Executive Roundtable--Meeting Summary, Appendix 1--Proposed
Changes to FCC Regulations covering the proposed changes to the FCC
regulations, it is stated that " . . . DSRCS systems utilize
non-voice radio techniques to transfer data over short distances
between roadside and mobile units, between mobile units and between
portable and mobile units to perform operations related to the
improvement of traffic flow, traffic safety and other intelligent
transportation service applications . . . ", etc.
A state or the Federal Government may require in the future that
all vehicles have passive transponders such as RFID tags. This
could be part of the registration system for the vehicle and, in
fact, could even be part of the license plate. This is somewhat
discussed in a paper by Shladover, Steven entitled "Cooperative
Advanced Vehicle Control and Safety Systems (AVCSS)". AVCSS sensors
will make it easy to detect the presence, location and identity of
all other vehicles in their vicinity. Passive radio frequency
transponders are discussed. The use of differential GPS with
accuracies as good as about two (2) centimeters, coupled with an
inertial guidance system, is discussed, as is the ability of
vehicles to communicate their locations to other vehicles. It
discusses the use of accurate maps, but not of lateral vehicle
control using these maps. It is obvious from reading this paper
that the author did not contemplate the safety system aspects of
using accurate maps and accurate GPS. In fact, the author stresses
the importance of cooperation between various government levels and
agencies and the private sector in order to make AVCSS feasible.
"Automotive suppliers cannot sell infrastructure-dependent systems
to their customers until the very large majority of the
infrastructure is suitable equipped."
10. The RtZF.TM. System--Intelligent Transportation Infrastructure
Benefits
A paper entitled "Intelligent Transportation Infrastructure
Benefits: Expected and Experienced", 1996 US Department of
Transportation, provides a summary of costs and benefits associated
with very modest ITS implementations. Although a complete cost
benefit analysis has not been conducted on the instant invention,
it is evident from reading this paper that the benefits to cost
ratio will be a very large number.
According to this paper, the congestion in the United States is
increasing at about 9% per year. In 50 metropolitan areas, the cost
in 1992 was estimated at 48 billion dollars and in Washington,
D.C., it represented an annual cost of $822 per person, or $1,580
per registered vehicle. In 1993, there were 40,115 people killed
and 3 million injured in traffic accidents. Sixty-one percent (61%)
of all fatal accidents occurred in rural areas. This reference
lists the 29 user services that make up the ITS program. It is
interesting that the instant invention provides 24 of the 29 listed
user services. A listing of the services and their proposed
implementation with the RtZF.TM. system is found in U.S. patent
application Ser. No. 10/822,445, now U.S. Pat. No. 7,085,637, and
is incorporated by reference herein.
The above references, among other things, demonstrate that there
are numerous methods and future enhancements planned that will
provide centimeter level accuracy to an RtZF.TM. equipped vehicle.
There are many alternative paths that can be taken but which ever
one is chosen the result is clear that such accuracies are within
the start of the art today.
In the particular area of speed control, U.S. Pat. No. 5,530,651 to
Uemura, et al. describes a combination of an ultrasonic and laser
radar optical detection system which has the ability to detect
soiled lenses, rain, snow, etc. The vehicle control system then
automatically limits the speed, for example, that the vehicle can
travel in adverse weather conditions. The speed of the vehicle is
also reduced when the visibility ahead is reduced due to a blind,
curved corner. The permitted speed is thus controlled based on
weather conditions and road geometry. There is no information in
the vehicle system as to the legal speed limit as provided for in
embodiments of the instant invention.
11. Limitations of the Prior Art
Previous inventions have attempted to solve the collision avoidance
problem for each vehicle independently of the other vehicles on the
roadway. Systems that predict vehicle trajectories generally fail
because two vehicles can be on a collision course and within the
last 0.1 second, a slight change of direction avoids the collision.
This is a common occurrence that depends on the actions of the
individual drivers and no collision avoidance system now in
existence is believed to be able to differentiate this case from an
actual collision. In the present invention described below, every
equipped vehicle will be confined to a corridor and to a position
within that corridor where the corridor depends on sub-meter
accurate digital maps. Only if that vehicle deviates from the
corridor will an alarm sound or the vehicle control system take
over control of the vehicle sufficiently to prevent the vehicle
from leaving its corridor if an accident would result from the
departure from that corridor.
Additionally, no prior art system is believed to have successfully
used the GPS navigational system, or an augmented DGPS to locate a
vehicle on a roadway with sufficient accuracy that that information
can be used to prevent the equipped vehicle from leaving the
roadway or striking another similarly equipped vehicle.
Prior art systems in addition to being poor at locating potential
hazards on the roadway, have not been able to ascertain whether
they are in fact on the roadway or off on the side, whether they
are threatening vehicles, static signs, overpasses etc. In fact, no
credible attempt to date has been made to identify or categorize
objects which may impact the subject vehicle.
The RtZF.TM. system in accordance with this invention also
contemplates a different kind of interrogating system. It is
optionally based on scanning infrared laser radar, terahertz radar
with or without range gating. This system, when used in conjunction
with accurate maps, will permit a precise imaging of an object on
the road in front of the vehicle, for example, permitting it to be
identified (using neural networks) and its location, velocity and
the probability of a collision to be determined.
In particular, the system of this invention is particularly
effective in eliminating accidents at intersections caused by
drivers running stop signs, red stoplights and turning into
oncoming traffic. There are approximately one million such
accidents and they are the largest killer of older drivers who
frequently get confused at intersections.
12. Definitions
"Pattern recognition" as used herein will generally mean any system
which processes a signal that is generated by an object (e.g.,
representative of a pattern of returned or received impulses, waves
or other physical property specific to and/or characteristic of
and/or representative of that object) or is modified by interacting
with an object, in order to determine to which one of a set of
classes that the object belongs. Such a system might determine only
that the object is or is not a member of one specified class, or it
might attempt to assign the object to one of a larger set of
specified classes, or find that it is not a member of any of the
classes in the set. The signals processed are generally a series of
electrical signals coming from transducers that are sensitive to
acoustic (ultrasonic) or electromagnetic radiation (e.g., visible
light, infrared radiation, capacitance or electric and/or magnetic
fields), although other sources of information are frequently
included. Pattern recognition systems generally involve the
creation of a set of rules that permit the pattern to be
recognized. These rules can be created by fuzzy logic systems,
statistical correlations, or through sensor fusion methodologies as
well as by trained pattern recognition systems such as neural
networks, combination neural networks, cellular neural networks or
support vector machines.
"Neural network" as used herein, unless stated otherwise, will
generally mean a single neural network, a combination neural
network, a cellular neural network, a support vector machine or any
combinations thereof. For the purposes herein, a "neural network"
is defined to include all such learning systems including cellular
neural networks, support vector machines and other kernel-based
learning systems and methods, cellular automata and all other
pattern recognition methods and systems that learn. A "combination
neural network" as used herein will generally apply to any
combination of two or more neural networks as most broadly defined
that are either connected together or that analyze all or a portion
of the input data.
A "combination neural network" as used herein will generally apply
to any combination of two or more neural networks that are either
connected together or that analyze all or a portion of the input
data. A combination neural network can be used to divide up tasks
in solving a particular object sensing and identification problem.
For example, one neural network can be used to identify an object
occupying a space at the side of an automobile and a second neural
network can be used to determine the position of the object or its
location with respect to the vehicle, for example, in the blind
spot. In another case, one neural network can be used merely to
determine whether the data is similar to data upon which a main
neural network has been trained or whether there is something
significantly different about this data and therefore that the data
should not be analyzed. Combination neural networks can sometimes
be implemented as cellular neural networks.
What has been described above is generally referred to as modular
neural networks with and without feedback. Actually, the feedback
does not have to be from the output to the input of the same neural
network. The feedback from a downstream neural network could be
input to an upstream neural network, for example.
The neural networks can be combined in other ways, for example in a
voting situation. Sometimes the data upon which the system is
trained is sufficiently complex or imprecise that different views
of the data will give different results. For example, a subset of
transducers may be used to train one neural network and another
subset to train a second neural network etc. The decision can then
be based on a voting of the parallel neural networks, sometimes
known as an ensemble neural network. In the past, neural networks
have usually only been used in the form of a single neural network
algorithm for identifying the occupancy state of the space near an
automobile. At least one of the inventions disclosed herein is
primarily advancing the state of the art and using combination
neural networks wherein two or more neural networks are combined to
arrive at a decision.
The applications for this technology are numerous as described in
the patents and patent applications listed above. However, the main
focus of some of the instant inventions is the process and
resulting apparatus of adapting the system in the patents and
patent applications referenced above and using combination neural
networks for the detection of the presence of an object such as
another vehicle in the environment of the subject vehicle where an
accident may occur and the motion of the vehicle needs to be
controlled so as to prevent such an accident.
A trainable or a trained pattern recognition system as used herein
generally means a pattern recognition system that is taught to
recognize various patterns constituted within the signals by
subjecting the system to a variety of examples. The most successful
such system is the neural network used either singly or as a
combination of neural networks. Thus, to generate the pattern
recognition algorithm, test data is first obtained which
constitutes a plurality of sets of returned waves, or wave
patterns, or other information radiated or obtained from an object
(or from the space in which the object will be situated in the
passenger compartment, i.e., the space above the seat) and an
indication of the identify of that object. A number of different
objects are tested to obtain the unique patterns from each object.
As such, the algorithm is generated, and stored in a computer
processor, and which can later be applied to provide the identity
of an object based on the wave pattern being received during use by
a receiver connected to the processor and other information. For
the purposes here, the identity of an object sometimes applies to
not only the object itself but also to its location and/or
orientation and velocity in the vicinity of the vehicle. For
example, a vehicle that is stopped but pointing at the side of the
host vehicle is different from the same vehicle that is approaching
at such a velocity as to impact the host vehicle. Not all pattern
recognition systems are trained systems and not all trained systems
are neural networks. Other pattern recognition systems are based on
fuzzy logic, sensor fusion, Kalman filters, correlation as well as
linear and non-linear regression. Still other pattern recognition
systems are hybrids of more than one system such as neural-fuzzy
systems.
A pattern recognition algorithm will thus generally mean an
algorithm applying or obtained using any type of pattern
recognition system, e.g., a neural network, sensor fusion, fuzzy
logic, etc.
To "identify" as used herein will generally mean to determine that
the object belongs to a particular set or class. The class may be
one containing, for example, all motorcycles, one containing all
trees, or all trees in the path of the host vehicle depending on
the purpose of the system.
To "ascertain the identity of" as used herein with reference to an
object will generally mean to determine the type or nature of the
object (obtain information as to what the object is), i.e., that
the object is an car, a car on a collision course with the host
vehicle, a truck, a tree, a pedestrian, a deer etc.
A "rear seat" of a vehicle as used herein will generally mean any
seat behind the front seat on which a driver sits. Thus, in
minivans or other large vehicles where there are more than two rows
of seats, each row of seats behind the driver is considered a rear
seat and thus there may be more than one "rear seat" in such
vehicles. The space behind the front seat includes any number of
such rear seats as well as any trunk spaces or other rear areas
such as are present in station wagons.
An "optical image" will generally mean any type of image obtained
using electromagnetic radiation including visual, infrared,
terahertz and radar radiation.
In the description herein on anticipatory sensing, the term
"approaching" when used in connection with the mention of an object
or vehicle approaching another will usually mean the relative
motion of the object toward the vehicle having the anticipatory
sensor system. Thus, in a side impact with a tree, the tree will be
considered as approaching the side of the vehicle and impacting the
vehicle. In other words, the coordinate system used in general will
be a coordinate system residing in the target vehicle. The "target"
vehicle is the vehicle that is being impacted. This convention
permits a general description to cover all of the cases such as
where (i) a moving vehicle impacts into the side of a stationary
vehicle, (ii) where both vehicles are moving when they impact, or
(iii) where a vehicle is moving sideways into a stationary vehicle,
tree or wall.
"Vehicle" as used herein includes any container that is movable
either under its own power or using power from another vehicle. It
includes, but is not limited to, automobiles, trucks, railroad
cars, ships, airplanes, trailers, shipping containers, barges, etc.
The word "container" will frequently be used interchangeably with
vehicle however a container will generally mean that part of a
vehicle that separate from and in some cases may exist separately
and away from the source of motive power. Thus a shipping container
may exist in a shipping yard and a trailer may be parked in a
parking lot without the tractor. The passenger compartment or a
trunk of an automobile, on the other hand, are compartments of a
container that generally only exists attaches to the vehicle
chassis that also has an associated engine for moving the vehicle.
Note a container can have one or a plurality of compartments.
"Out-of-position" as used for an occupant will generally mean that
the occupant, either the driver or a passenger, is sufficiently
close to an occupant protection apparatus (airbag) prior to
deployment that he or she is likely to be more seriously injured by
the deployment event itself than by the accident. It may also mean
that the occupant is not positioned appropriately in order to
attain the beneficial, restraining effects of the deployment of the
airbag. As for the occupant being too close to the airbag, this
typically occurs when the occupant's head or chest is closer than
some distance such as about 5 inches from the deployment door of
the airbag module. The actual distance where airbag deployment
should be suppressed depends on the design of the airbag module and
is typically farther for the passenger airbag than for the driver
airbag.
"Transducer" or "transceiver" as used herein will generally mean
the combination of a transmitter and a receiver. In come cases, the
same device will serve both as the transmitter and receiver while
in others two separate devices adjacent to each other will be used.
In some cases, a transmitter is not used and in such cases
transducer will mean only a receiver. Transducers include, for
example, capacitive, inductive, ultrasonic, electromagnetic
(antenna, CCD, CMOS arrays, laser, radar transmitter, terahertz
transmitter and receiver, focal plane array, pin or avalanche
diode, etc.), electric field, weight measuring or sensing devices.
In some cases, a transducer will be a single pixel either acting
alone, in a linear or an array of some other appropriate shape. In
some cases, a transducer may comprise two parts such as the plates
of a capacitor or the antennas of an electric field sensor.
Sometimes, one antenna or plate will communicate with several other
antennas or plates and thus for the purposes herein, a transducer
will be broadly defined to refer, in most cases, to any one of the
plates of a capacitor or antennas of a field sensor and in some
other cases a pair of such plates or antennas will comprise a
transducer as determined by the context in which the term is
used.
"Adaptation" as used here will generally represent the method by
which a particular occupant or vehicle or other object sensing
system is designed and arranged for a particular vehicle model. It
includes such things as the process by which the number, kind and
location of various transducers is determined. For pattern
recognition systems, it includes the process by which the pattern
recognition system is designed and then taught or made to recognize
the desired patterns. In this connection, it will usually include
(1) the method of training when training is used, (2) the makeup of
the databases used, testing and validating the particular system,
or, in the case of a neural network, the particular network
architecture chosen, (3) the process by which environmental
influences are incorporated into the system, and (4) any process
for determining the pre-processing of the data or the post
processing of the results of the pattern recognition system. The
above list is illustrative and not exhaustive. Basically,
adaptation includes all of the steps that are undertaken to adapt
transducers and other sources of information to a particular
vehicle to create the system that accurately identifies and/or
determines the location of an occupant or other object in a vehicle
or in the environment around the vehicle.
A "morphological characteristic" will generally mean any measurable
property of a human such as height, weight, leg or arm length, head
diameter, skin color or pattern, blood vessel pattern, voice
pattern, finger prints, iris patterns, etc.
A "wave sensor" or "wave transducer" is generally any device which
senses either ultrasonic or electromagnetic waves. An
electromagnetic wave sensor, for example, includes devices that
sense any portion of the electromagnetic spectrum from ultraviolet
down to a few hertz. The most commonly used kinds of
electromagnetic wave sensors include CCD and CMOS arrays for
sensing visible and/or infrared waves, millimeter wave and
microwave radar, and capacitive or electric and/or magnetic field
monitoring sensors that rely on the dielectric constant of the
object occupying a space but also rely on the time variation of the
field, expressed by waves as defined below, to determine a change
in state.
A "CCD" will be defined to include all devices, including CMOS
arrays, APS arrays, QWIP arrays or equivalent, artificial retinas
and particularly HDRC arrays, which are capable of converting light
frequencies, including infrared, visible and ultraviolet, into
electrical signals. The particular CCD array used for many of the
applications disclosed herein is implemented on a single chip that
is less than two centimeters on a side. Data from the CCD array is
digitized and sent serially to an electronic circuit (at times
designated 120 herein) containing a microprocessor for analysis of
the digitized data. In order to minimize the amount of data that
needs to be stored, initial processing of the image data takes
place as it is being received from the CCD array, as discussed in
more detail above. In some cases, some image processing can take
place on the chip such as described in the Kage et al. artificial
retina article referenced above.
The "windshield header" as used herein includes the space above the
front windshield including the first few inches of the roof.
An "occupant protection apparatus" is any device, apparatus, system
or component which is actuatable or deployable or includes a
component which is actuatable or deployable for the purpose of
attempting to reduce injury to the occupant in the event of a
crash, rollover or other potential injurious event involving a
vehicle Inertial measurement unit (IMU), inertial navigation system
(INS) and inertial reference unit (IRU) will in general be used be
used interchangeably to mean a device having a plurality of
accelerometers and a plurality of gyroscopes generally within the
same package. Usually such a device will contain 3 accelerometers
and 3 gyroscopes. In some cases a distinction will be made whereby
the INS relates to an IMU or an IRU plus additional sensors and
software such as a GPS, speedometer, odometer or other sensors plus
optimizing software which may be based on a Kalman filter.
A precise positioning system or PPS is a system based on some
information, usually of a physical nature, in the infrastructure
that determines the precise location of a vehicle independently of
a GPS based system or the IMU. Such a system is employed as a
vehicle is traveling and passes a particular location. A PPS can
make use of various technologies including radar, laser radar,
terahertz radar, RFID tags located in the infrastructure, MIR
transmitters and receivers. Such locations are identified on a map
database resident within the vehicle. In one case, for example, the
map database contains data from a terahertz radar continuous scan
of the environment to the side of a vehicle from a device located
on a vehicle and pointed 45 degrees up relative to the horizontal
plane. The map database contains the exact location of the vehicle
that corresponds to the scan. Another vehicle can then determine
its location by comparing its scan data with that stored with the
map database and when there is a match, the vehicle knows its
location. Of course many other technologies can be used to
accomplish a similar result.
Unless stated otherwise, laser radar, lidar and ladar will be
considered equivalent herein. In all cases they represent a
projected laser beam, which can be in the visual part of the
electromagnetic spectrum but generally will be the infrared part of
the electromagnetic spectrum and usually in the near infrared
wavelengths. The projected laser beam can emanate from the optics
as a nearly parallel beam or as a beam that diverges at any desired
angle from less than zero degrees to ten or more of degrees
depending on the application. A particular implementation may use a
laser beam that at one time diverges at an angle less than one
degree and at another time may diverge at several degrees using
adjustable optics. The laser beam can have a diameter as it leaves
the vehicle ranging from less than a millimeter to several
centimeters. The above represent typical or representative ranges
of dimensions but this invention is not limited by these
ranges.
OBJECTS AND SUMMARY OF THE INVENTION
It is an object of the present invention to provide a new and
improved method for obtaining information about objects in the
environment outside of and around a vehicle.
It is another object of the present invention to provide a new and
improved method and system for avoiding collisions between a
vehicle and another object, such as another vehicle or
infrastructure.
In order to achieve these objects and others, a method for
obtaining information about objects in the environment outside of
and around a vehicle includes directing a laser beam from the
vehicle into the environment, receiving from an object in the path
of the laser beam a reflection of the laser beam at a location on
the vehicle, and analyzing the received laser beam reflections to
obtain information about the object from which the laser beam is
being reflected. Analysis of the laser beam reflections preferably
entails range gating the received laser beam reflections to limit
analysis of the received laser beam reflections to only those
received from an object within a defined (distance) range such that
objects at distances within the range are isolated from surrounding
objects. In this manner, data gathering is optimized in that data
about only objects in the distance range is obtained.
To optimize the method, the direction of the laser beam can be
controlled to cover only a desired operating sector. Also, so that
objects that cannot potentially impact the vehicle are not
considered, thereby reducing wasted processing time for the
processor and false alerts, a digital map may be provided including
information relating to roads on which the vehicle can travel or is
traveling. In this manner, objects which cannot impact the vehicle,
such as those traveling on the same road but in an opposite
direction and when a concrete barrier separates the lanes, are not
considered potentially dangerous. A field into which the laser beam
will be directed is defined based on the map and the laser beam is
directed primarily only into the defined field.
To cover possible situations with curved roads causing the vehicle
to curve, two laser beams can be directed into the environment. The
laser beams can have different scanning speeds.
Analysis of the laser beam reflections may also entail analyzing
the received laser beam reflections to detect the presence of
objects potentially affecting operation of the vehicle, e.g., which
would require the vehicle to alter its travel path to avoid a
collision with the vehicle. Range gating is performed once the
presence of each object is detected and the range is determined to
encompass any objects whose presence has been detected. The range
can be narrowed such that laser beam reflections from only the
object whose presence is detected and other objects in the same
range are analyzed and processed to obtain identification or
identity information about them. Pattern recognition algorithms can
be used to process the received laser beam reflections, e.g., to
ascertain the identity of or identity the objects. If an object is
identified and the potential for a collision between the vehicle
and that object is determined to be present, the driver can be
alerted about the potential collisions, e.g., visually and/or
audibly, and/or a vehicle control system can be activated to alter
the vehicle's travel path to avoid the collision.
A method for avoiding collisions between a vehicle and another
object includes mounting a laser beam projector on the vehicle,
directing a laser beam from the projector outward from the vehicle,
determining whether an object is present in the path of the laser
beam based on reception of reflections of the laser beam caused by
the presence of the object in the path of the laser beam, and when
an object is determined to be present, setting a distance range
including a distance between the vehicle and the object, processing
only received reflections of the laser beam emanating from objects
in the set distance range to determine whether each object may
impact the vehicle, and if a determination is made that the object
may impact the vehicle, effecting a countermeasure with a view
toward preventing the collision. The same enhancements to the
method described above can be applied here as well, e.g., the use
of a digital map to limit the number of objects considered as
potentially dangerous and the countermeasures effected to avoid
collisions.
A system for avoiding collisions between a vehicle and another
object includes a laser beam projector arranged on the vehicle to
directing a laser beam outward from the vehicle, a receiving unit
for receiving reflections of the laser beam which reflect off of
objects in the path of the laser beam, and a control unit, module
or processor arranged to process any received reflections to
determine whether an object is present in the path of the laser
beam. When an object is determined to be present, the processor
sets a distance range including a distance between the vehicle and
the object, processes only received reflections of the laser beam
emanating from objects in the set distance range to determine
whether each object may impact the vehicle, and if a determination
is made that the object may impact the vehicle, causes a
countermeasure to be effected with a view toward preventing the
collision. Optionally, the processor includes a pattern recognition
algorithm which ascertains the identity of or identifies each
object in the set distance range and assesses the potential for and
consequences of a collision between the vehicle and the object
based on the identity or identification of the object. The
countermeasures can be activation of a driver notification system
to alert the driver of the impending collision or activation of a
vehicle control system to vary the travel of the vehicle to avoid
the impending collision.
Other objects and advantages of disclosed inventions include: 1. To
provide a system based partially on the global positioning system
(GPS) or equivalent that permits an onboard electronic system to
determine the position of a vehicle with an accuracy of 1 meter or
less. 2. To provide a system which permits an onboard electronic
system to determine the position of the edges and/or lane
boundaries of a roadway with an accuracy of 1 meter or less in the
vicinity of the vehicle. 3. To provide a system which permits an
onboard vehicle electronic system to determine the position of the
edges and/or lane boundaries of a roadway relative to the vehicle
with an accuracy of less than about 10 centimeters, one sigma. 4.
To provide a system that substantially reduces the incidence of
single vehicle accidents caused by the vehicle inappropriately
leaving the roadway at high speed. 5. To provide a system which
does not require modification to a roadway which permits high speed
controlled travel of vehicles on the roadway thereby increasing the
vehicle flow rate on congested roads. 6. To provide a collision
avoidance system comprising a sensing system responsive to the
presence of at least one other vehicle in the vicinity of the
equipped vehicle and means to determine the location of the other
vehicle relative to the lane boundaries of the roadway and thereby
determine if the other vehicle has strayed from its proper position
on the highway thereby increasing the risk of a collision, and
taking appropriate action to reduce that risk. 7. To provide a
means whereby vehicles near each other can communicate their
position and/or their velocity to each other and thereby reduce the
risk of a collision. 8. To provide a means for accurate maps of a
roadway to be transmitted to a vehicle on the roadway. 9. To
provide a means for weather, road condition and/or similar
information can be communicated to a vehicle traveling on a roadway
plus means within the vehicle for using that information to reduce
the risk of an accident. 10. To provide a means and apparatus for a
vehicle to precisely know its location at certain positions on a
road by passing through or over an infrastructure based local
subsystem thereby permitting the vehicle electronic systems to self
correct for the satellite errors making the vehicle for a brief
time a DGPS station and facilitate carrier phase DGPS for increased
location accuracy. Such a subsystem may be a PPS including one
based on the signature of the environment. 11. To utilize
government operated navigation aid systems such as the WAAS and
LAAS as well as other available or to become available systems to
achieve sub-meter vehicle location accuracies. 12. To utilize the
OpenGIS.TM. map database structure so as to promote open systems
for accurate maps for the RtZF.TM. system. 13. To eliminate
intersection collisions caused by a driver running a red light or
stop sign. 14. To eliminate intersection collisions caused by a
driver executing a turn into oncoming traffic. 15. To provide a
method of controlling the speed of a vehicle based on map
information or information transmitted to the vehicle from the
infrastructure. Such speed control may be based on information as
to the normal legal speed limit or a variable speed limit set by
weather or other conditions.
Other improvements will now be obvious to those skilled in the art.
The above features are meant to be illustrative and not
definitive.
The preferred embodiments of the inventions are shown in the
drawings and described in the detailed description below. Unless
specifically noted, it is applicant's intention that the words and
phrases in the specification and claims be given the ordinary and
accustomed meaning to those of ordinary skill in the applicable
art(s). If applicant intends any other meaning, he will
specifically state he is applying a special meaning to a word or
phrase.
Likewise, applicant's use of the word "function" in the detailed
description is not intended to indicate that he seeks to invoke the
special provisions of 35 U.S.C. .sctn.112, 6 to define his
invention. To the contrary, if applicant wishes to invoke the
provision of 35 U.S.C. .sctn.112, 6, to define his invention, he
will specifically set forth in the claims the phrases "means for"
or "step for" and a function, without also reciting in that phrase
any structure, material or act in support of the function.
Moreover, even if applicant invokes the provisions of 35 U.S.C.
.sctn.112, 6, to define his invention, it is applicant's intention
that his inventions not be limited to the specific structure,
material or acts that are described in preferred embodiments.
Rather, if applicant claims his invention by specifically invoking
the provisions of 35 U.S.C. .sctn.112, 6, it is nonetheless his
intention to cover and include any and all structures, materials or
acts that perform the claimed function, along with any and all
known or later developed equivalent structures, materials or acts
for performing the claimed function.
For example, the present inventions make use of GPS satellite
location technology, including the use of MIR or RFID triads or
radar and reflectors, to derive kinematic vehicle location and
motion trajectory parameters for use in a vehicle collision
avoidance system and method. The inventions described herein are
not to be limited to the specific GPS devices or PPS devices
disclosed in the preferred embodiments, but rather, are intended to
be used with any and all such applicable satellite and
infrastructure location devices, systems and methods, as long as
such devices, systems and methods generate input signals that can
be analyzed by a computer to accurately quantify vehicle location
and kinematic motion parameters in real time. Thus, the GPS and PPS
devices and methods shown and referenced generally throughout this
disclosure, unless specifically noted, are intended to represent
any and all devices appropriate to determine such location and
kinematic motion parameters.
Likewise, for example, the present inventions generate surveillance
image information for analysis by scanning using any applicable
image or video scanning system or method. The inventions described
herein are not to be limited to the specific scanning or imaging
devices or to a particular electromagnetic frequency or frequency
range or part of the electromagnetic spectrum disclosed in the
preferred embodiments, but rather, are intended to be used with any
and all applicable electronic scanning devices, as long as the
device can generate an output signal that can be analyzed by a
computer to detect and categorize objects. Thus, the scanners or
image acquisition devices are shown and referenced generally
throughout this disclosure, and unless specifically noted, are
intended to represent any and all devices appropriate to scan or
image a given area. Accordingly, the words "scan" or "image" as
used in this specification should be interpreted broadly and
generically.
Further, there are disclosed several processors or controllers,
that perform various control operations. The specific form of
processor is not important to the invention. In its preferred form,
applicant divides the computing and analysis operations into
several cooperating computers or microprocessors. However, with
appropriate programming well known to those of ordinary skill in
the art, the inventions can be implemented using a single, high
power computer. Thus, it is not applicant's intention to limit his
invention to any particular form or location of processor or
computer. For example, it is contemplated that in some cases the
processor may reside on a network connected to the vehicle such as
one connected to the Internet.
Further examples exist throughout the disclosure, and it is not
applicant's intention to exclude from the scope of his invention
the use of structures, materials, or acts that are not expressly
identified in the specification, but nonetheless are capable of
performing a claimed function.
The above and other objects are achieved in the present invention
which provides motor vehicle collision avoidance, warning and
control systems and methods using GPS satellite location systems
augmented with Precise Positioning Systems to provide centimeter
location accuracy, and to derive vehicle attitude and position
coordinates and vehicle kinematic tracking information. GPS
location and computing systems being integrated with vehicle video
scanning, radar, laser radar, terahertz radar and onboard
speedometer and/or accelerometers and gyroscopes to provide
accurate vehicle location information together with information
concerning hazards and/or objects that represent impending
collision situations for each vehicle. Advanced image processing
techniques are used to quantify video information signals and to
derive vehicle warning and control signals based upon detected
hazards.
Outputs from multiple sensors as described above are used in
onboard vehicle neural network and neural-fuzzy system computing
algorithms to derive optimum vehicle warning and control signals
designed to avoid vehicle collisions with other vehicles or with
other objects or hazards that may be present on given roadways. In
a preferred embodiment, neural fuzzy control algorithms are used to
develop coordinated braking, acceleration and steering control
signals to control individual vehicles, or the individual wheels of
such vehicles, in an optimal manner to avoid or minimize the
effects of potential collisions. Video, radar, laser radar,
terahertz radar and GPS position and trajectory information are
made available to each individual vehicle describing the movement
of that vehicle and other vehicles in the immediate vicinity of
that vehicle.
In addition, hazards or other obstacles that may represent a
potential danger to a given vehicle are also included in the neural
fuzzy calculations. Objects, obstacles and/or other vehicles
located anywhere to the front, rear or sides of a given vehicle are
considered in the fuzzy logic control algorithms in the derivation
of optimal control and warning signals.
The above and other objects and advantages of the present invention
are achieved by the preferred embodiments that are summarized and
described in detail below.
BRIEF DESCRIPTION OF THE DRAWINGS
The various hardware and software elements used to carry out the
invention described herein are illustrated in the form of system
diagrams, block diagrams, flow charts, and depictions of neural
network algorithms and structures. The preferred embodiment is
illustrated in the following figures:
FIG. 1 illustrates the GPS satellite system with the 24 satellites
revolving around the earth.
FIG. 2 illustrates four GPS satellites transmitting position
information to a vehicle and to a base station which in turn
transmits the differential correction signal to the vehicle.
FIG. 3 illustrates a WADGPS system with four GPS satellites
transmitting position information to a vehicle and to a base
station which in turn transmits the differential correction signal
to the vehicle.
FIG. 4 is a logic diagram showing the combination of the GPS system
and an inertial navigation system.
FIG. 5 is a block diagram of the overall vehicle accident
avoidance, warning, and control system and method of the present
invention illustrating system sensors, radio transceivers,
computers, displays, input/output devices and other key
elements.
FIG. 5A is a block diagram of a representative accident avoidance,
warning and control system.
FIG. 6 is a block diagram of an image analysis computer of the type
that can be used in the accident avoidance system and method of
this invention.
FIG. 7 illustrates a vehicle traveling on a roadway in a defined
corridor.
FIG. 8 illustrated two adjacent vehicles traveling on a roadway and
communicating with each other.
FIG. 9 is a schematic diagram illustrating a neural network of the
type useful in the image analysis computer of FIG. 5.
FIG. 10 is a schematic diagram illustrating the structure of a node
processing element in the neural network of FIG. 9.
FIG. 11 illustrates the use of a Precise Positioning System
employing three micropower impulse radar transmitters, two or three
radar reflectors or three RFID tags in a configuration to allow a
vehicle to accurately determine its position.
FIG. 12a is a flow chart of the method in accordance with the
invention for preventing run off the road accidents.
FIG. 12b is a flow chart of the method in accordance with the
invention for preventing center (yellow) line crossing
accidents.
FIG. 12c is a flow chart of the method in accordance with the
invention for preventing stoplight running accidents.
FIG. 13 illustrates an intersection with stop signs on the lesser
road where there is a potential for a front to side impact and a
rear end impact.
FIG. 14 illustrates a blind intersection with stoplights where
there is a potential for a front side to front side impact.
FIG. 15 illustrates an intersection where there is a potential for
a front-to-front impact as a vehicle turns into oncoming
traffic.
FIG. 16A is a side view of a vehicle equipped with a road-mapping
arrangement in accordance with the invention.
FIG. 16B is a front perspective view of a vehicle equipped with the
road-mapping arrangement in accordance with the invention.
FIG. 17 is a schematic perspective view of a data acquisition
module in accordance with the invention.
FIG. 17A is a schematic view of the data acquisition module in
accordance with the invention.
FIG. 18 shows the view of a road from the video cameras in both of
the data acquisition modules.
FIG. 19 shows a variety of roads and vehicles operating on those
roads that are in communication with a vehicle that is passing
through a Precise Positioning Station.
FIG. 20 is a schematic of the manner in which communications
between a vehicle and a transmitter are conducted according to some
embodiments of the invention.
FIGS. 21A and 21B illustrate a preferred embodiment of a laser
radar system mounted at the four corners of a vehicle above the
headlights and tail lights.
FIGS. 22A and 22B illustrate the system of FIGS. 21A and 21B for
vehicles on a roadway.
FIGS. 23A and 23B illustrate an alternative mounting location for
laser radar units.
FIG. 24 is a schematic illustration of a typical laser radar device
showing the scanning or pointing system with simplified optics.
FIG. 25 is a schematic showing a method for avoiding collisions in
accordance with the invention.
DETAILED DISCUSSION OF THE PREFERRED EMBODIMENTS
1. Vehicle Collision Warning and Control
According to U.S. Pat. No. 5,506,584, the stated goals of the US
DOT IVHS system are: improving the safety of surface transportation
increasing the capacity and operational efficiency of the surface
transportation system enhancing personal mobility and the
convenience and comfort of the surface transportation system
reducing the environmental and energy impacts of the surface
transportation system
The RtZF.TM. system in accordance with the present invention
satisfies all of these goals at a small fraction of the cost of
prior art systems. The safety benefits have been discussed above.
The capacity increase is achieved by confining vehicles to
corridors where they are then permitted to travel at higher speeds.
This can be achieved immediately where carrier phase DGPS is
available or with the implementation of the highway-located precise
location systems as shown in FIG. 11. An improvement is to add the
capability for the speed of the vehicles to be set by the highway
or highway control system. This is a simple additional few bytes of
information that can be transmitted along with the road edge
location map, thus, at very little initial cost. To account for the
tolerances in vehicle speed control systems, the scanning laser
radar, or other technology system, which monitors for the presence
of vehicles without RtZF.TM. is also usable as an adaptive cruise
control system. Thus, if a faster moving vehicle approaches a
slower moving vehicle, it will automatically slow down to keep a
safe separation distance from the leading, slower moving vehicle.
Although the system is not planned for platooning, that will be the
automatic result in some cases. The maximum packing of vehicles is
automatically obtained and thus the maximum vehicle flow rate is
also achieved with a very simple system.
For the Intelligent Highway System (ITS) application, some
provision is required to prevent unequipped vehicles from entering
the restricted lanes. In most cases, a barrier will be required
since if an errant vehicle did enter the controlled lane, a serious
accident could result. Vehicles would be checked while traveling
down the road or at a tollbooth, or similar station, that the
RtZF.TM. system was in operation without faults and with the latest
updated map for the region. Only those vehicles with the RtZF.TM.
system in good working order would be permitted to enter. The speed
on the restricted lanes would be set according to the weather
conditions and fed to the vehicle information system automatically,
as discussed above. Automatic tolling based on the time of day or
percentage of highway lane capacity in use can also be easily
implemented.
For ITS use, there needs to be a provision whereby a driver can
signal an emergency, for example, by putting on the hazard lights.
This would permit the vehicle to leave the roadway and enter the
shoulder when the vehicle velocity is below a certain level. Once
the driver provides such a signal, the roadway information system,
or the network of vehicle-based control systems, would then reduce
the speed of all vehicles in the vicinity until the emergency has
passed. This roadway information system need not be actually
associated with the particular roadway and also need not require
any roadway infrastructure. It is a term used here to represent the
collective system as operated by the network of nearby vehicles and
the inter-vehicle communication system. Eventually, the occurrence
of such emergency situations will be eliminated by vehicle-based
failure prediction systems such as described in U.S. Pat. No.
5,809,437.
Emergency situations will develop on intelligent highways. It is
difficult to access the frequency or the results of such
emergencies. The industry has learned from airbags that if a system
is developed which saves many lives but causes a few deaths, the
deaths will not be tolerated. The ITS system, therefore, must
operate with a very high reliability, that is approaching "zero
fatalities".TM.. Since the brains of the system will reside in each
vehicle, which is under the control of individual owners, there
will be malfunctions and the system must be able to adapt without
causing accidents. An alternative is for the brains to reside on
the network providing that the network connection is reliable.
The spacing of the vehicles is the first line of defense. Secondly,
each vehicle with a RtZF.TM. system has the ability to
automatically communicate to all adjacent vehicles and thus
immediately issue a warning when an emergency event is occurring.
Finally, with the addition of a total vehicle diagnostic system,
such as disclosed in U.S. Pat. No. 5,809,437 (Breed), "On Board
Vehicle Diagnostic System", potential emergencies can be
anticipated and thus eliminated with high reliability.
Although the application for ITS envisions a special highway lane
and high speed travel, the potential exists in the invention to
provide a lower measure of automatic guidance where the operator
can turn control of the vehicle over to the RtZF.TM. system for as
long as the infrastructure is available. In this case, the vehicle
would operate in normal lanes but would retain its position in the
lane and avoid collisions until a decision requiring operator
assistance is required. At that time, the operator would be
notified and if he or she did not assume control of the vehicle, an
orderly stopping of the vehicle, e.g., on the side of the road,
would occur.
For all cases where vehicle steering control is assumed by the
RtZF.TM. system, an algorithm for controlling the steering should
be developed using neural networks or neural fuzzy systems. This is
especially true for the emergency cases discussed herein where it
is well known that operators frequently take the wrong actions and
at the least, they are slow to react. Algorithms developed by other
non-pattern recognition techniques do not, in general, have the
requisite generality or complexity and are also likely to make the
wrong decisions (although the use of such systems is not precluded
in the invention). When the throttle and breaking functions are
also handled by the system, an algorithm based on neural networks
or neural fuzzy systems is even more important.
For the ITS, the driver will enter his or her destination so that
the vehicle knows ahead of time where to exit. Alternately, if the
driver wishes to exit, he merely turns on his turn signal, which
tells the system and other vehicles that he or she is about to exit
the controlled lane.
Neural networks have been mentioned above and since they can play
an important role in various aspects of the invention, a brief
discussion is now presented here. FIG. 9 is a schematic diagram
illustrating a neural network of the type useful in image analysis.
Data representing features from the images from the CMOS cameras 60
are input to the neural network circuit 63, and the neural network
circuit 63 is then trained on this data (see FIG. 6). More
specifically, the neural network circuit 63 adds up the feature
data from the CMOS cameras 60 with each data point multiplied by an
associated weight according to the conventional neural network
process to determine the correlation function.
In this embodiment, 141 data points are appropriately
interconnected at 25 connecting points of layer 1, and each data
point is mutually correlated through the neural network training
and weight determination process. In some implementations, each of
the connecting points of the layer 1 has an appropriate threshold
value, and if the sum of measured data exceeds the threshold value,
each of the connecting points will output a signal to the
connecting points of layer 2. In other cases, an output value or
signal will always be outputted to layer 2 without
thresholding.
The connecting points of the layer 2 comprises 20 points, and the
25 connecting points of the layer 1 are appropriately
interconnected as the connecting points of the layer 2. Similarly,
each data value is mutually correlated through the training process
and weight determination as described above and in the above
referenced neural network texts. Each of the 20 connecting points
of the layer 2 can also have an appropriate threshold value, if
thresholding is used, and if the sum of measured data exceeds the
threshold value, each of the connecting points will output a signal
to the connecting points of layer 3.
The connecting points of the layer 3 comprises 3 points in this
example, and the connecting points of the layer 2 are
interconnected at the connecting points of the layer 3 so that each
data is mutually correlated as described above.
The value of each connecting point is determined by multiplying
weight coefficients and summing up the results in sequence, and the
aforementioned training process is to determine a weight
coefficient Wj so that the value (ai) is a previously determined
output. ai=.SIGMA.WjXj(j=1 to N)+W.sub.0 wherein Wj is the weight
coefficient, Xj is the data N is the number of samples and W.sub.0
is bias weight associated with each node.
Based on this result of the training, the neural network circuit 63
generates the weights and the bias weights for the coefficients of
the correlation function or the algorithm.
At the time the neural network circuit 63 has learned from a
suitable number of patterns of the training data, the result of the
training is tested by the test data. In the case where the rate of
correct answers of the object identification unit based on this
test data is unsatisfactory, the neural network circuit 63 is
further trained and the test is repeated. Typically, about 200,000
feature patterns are used to train the neural network 63 and
determine all of the weights. A similar number is then used for the
validation of the developed network. In this simple example chosen,
only three outputs are illustrated. These can represent another
vehicle, a truck and a pole or tree. This might be suitable for an
early blind spot detector design. The number of outputs depends on
the number of classes of objects that are desired. However, too
many outputs can result in an overly complex neural network and
then other techniques such as modular neural networks can be used
to simplify the process. When a human looks at a tree, for example,
he or she might think "what kind of tree is that?" but not "what
kind of tiger is that". The human mind operates with modular or
combination neural networks where the object to be identified is
first determined to belong to a general class and then to a
subclass etc. Object recognition neural networks can frequently
make use of this principle with a significant simplification
resulting.
In the above example, the image was first subjected to a feature
extraction process and the feature data was input to the neural
network. In other cases, especially as processing power continues
to advance, the entire image is input to the neural network for
processing. This generally requires a larger neural network.
Alternate approaches use data representing the difference between
two frames and the input data to the neural network. This is
especially useful when a moving object of interest is in an image
containing stationary scenery that is of no interest. This
technique can be used even when everything is moving by using the
relative velocity as a filter to remove unwanted pixel data. Any
variations are possible and will now be obvious to those skilled in
the art. Alternately, this image can be filtered based on range,
which will also significantly reduce the number of pixels to be
analyzed.
In another implementation, the scenes are differenced based on
illumination. If infrared illumination is used, for example, the
illumination can be turned on and off and images taken and then
differenced. If the illumination is known only to illuminate an
object of interest then such an object can be extracted from the
background by this technique. A particularly useful method is to
turn the illumination on and off for alternate scan lines in the
image. Adjacent scan lines can then be differenced and the
resulting image sent to the neural network system for
identification.
The neural network can be implemented as an algorithm on a
general-purpose microprocessor or on a dedicated parallel
processing DSP, neural network ASIC or other dedicated parallel or
serial processor. The processing speed is generally considerably
faster when parallel processors are used and this can also permit
the input of the entire image for analysis rather than using
feature data. A combination of feature and pixel data can also be
used.
Neural networks have certain known potential problem areas that
various researchers have attempted to eliminate. For example, if
data representing an object that is totally different from those
objects present in the training data is input to the neural
network, an unexpected result can occur which, in some cases, can
cause a system failure. To solve this and other neural network
problems, researchers have resorted to adding in some other
computational intelligence principles such as fuzzy logic resulting
in a neural-fuzzy system, for example. As the RtZF.TM. system
evolves, such refinements will be implemented to improve the
accuracy of the system. Thus, although pure neural networks are
currently being applied to the problem, hybrid neural networks such
as modular, combination, ensemble and fuzzy neural networks will
undoubtedly evolve.
A typical neural network processing element known to those skilled
in the art is shown in FIG. 10 where input vectors, (X1, X2 . . .
Xn) are connected via weighing elements 120 (W1, W2 . . . Wn) to a
summing node 130. The output of node 130 is passed through a
nonlinear processing element 140, typically a sigmoid function, to
produce an output signal, Y. Offset or bias inputs 125 can be added
to the inputs through weighting circuit 128. The output signal from
summing node 130 is passed through the nonlinear element 140 which
has the effect of compressing or limiting the magnitude of the
output Y.
Neural networks used in the accident avoidance system of this
invention are trained to recognize roadway hazards including
automobiles, trucks, animals and pedestrians. Training involves
providing known inputs to the network resulting in desired output
responses. The weights are automatically adjusted based on error
signal measurements until the desired outputs are generated.
Various learning algorithms may be applied with the back
propagation algorithm with the Delta Bar rule as a particularly
successful method.
2. Accurate Navigation
2.1 GPS
FIG. 1 shows the current GPS satellite system associated with the
earth and including 24 satellites 2, each satellite revolving in a
specific orbital path 4 around the earth. By means of such a GPS
satellite system, the position of any object can be determined with
varying degrees of precision as discussed in detail herein. A
similar system will appear when the European Galileo system is
launched perhaps doubling the number of satellites.
2.2 DGPS, WAAS, LAAS and Pseudolites
FIG. 2 shows an arrangement of four satellites 2 designated
SV.sub.1, SV.sub.2, SV.sub.3 and SV.sub.4 of the GPS satellite
system shown in FIG. 1 transmitting position information to
receiver means of a base station 20, such as an antenna 22, which
in turn transmits a differential correction signal via transmitter
means associated with that base station, such as a second antenna
16, to a vehicle 18.
Additional details relating to FIGS. 1 and 2 can be found in U.S.
Pat. No. 5,606,506 to Kyrtsos, which is incorporated by reference
herein.
FIG. 3 shows an arrangement of four satellites 2 designated
SV.sub.1, SV.sub.2, SV.sub.3 and SV.sub.4 of the GPS satellite
system as in FIG. 2 transmitting position information to receivers
of base stations 20 and 21, such as an antenna 22, which in turn
transmit a differential correction signal via transmitter means
associated with that base stations, such as a second antenna 16, to
a geocentric or low earth orbiting (LEO) satellite 30 which in turn
transmits the differential correction signals to vehicle 18. In
this case, one or more of the base stations 20,21 receives and
performs a mathematical analysis on all of the signals received
from a number of base stations that cover the area under
consideration and forms a mathematical model of the errors in the
GPS signals over the entire area. For the CONUS, for example, a
group of 13 base stations are operated by OmniStar that are
distributed around the country. By considering data from the entire
group of such stations, the errors in the GPS signals for the
entire area can be estimated resulting in a position accuracy of
about 6 10 cm over the entire area. The corrections are then
uploaded to the geocentric or low earth orbiting satellite 30 for
retransmission to vehicles on the roadways. In this way, such
vehicles are able to determine their absolute position to within
about 6 10 centimeters. This is known as Wide Area Deferential GPS
or WADGPS.
It is important to note that future GPS and Galileo satellite
systems plan for the transmission of multiple frequencies for
civilian use. Like a lens, the ionosphere diffracts different
frequencies by different amounts and thus the time of arrival of a
particular frequency will depend on the value of that frequency.
This fact can be used to determine the amount that each frequency
is diffracted and thus the delay or error introduced by the
ionosphere. Thus with more than one frequency being emitted by a
particular satellite, the equivalent of the DGPS corrections can be
determined be each receiver and there in no longer a need for DGPS,
WADGPS, WAAS, LAAS and similar systems.
The WAAS system is another example of WADGPS for use with
airplanes. The U.S. Government estimates that the accuracy of the
WAAS system is about 1 meter in three dimensions. Since the largest
error is in the vertical direction, the horizontal error is much
less.
2.3 Carrier Phase Measurements
If a receiver can receive signals by two paths from a satellite it
can measure the phase difference between the two paths and,
provided that there are not any extra cycles in one of the paths,
the path difference can be determined to less than one centimeter.
The fact that there may be an integer number of extra cycles in one
path and not in the other is what is called the integer ambiguity
problem and a great deal of attention has been paid in the
literature to resolving this ambiguity. Using the Precise
Positioning System (PPS) described in detail below where a vehicle
becomes its own DGPS system, the carrier phase ambiguity problem
also disappears since the number of additional cycles can be
determined as the vehicle travels away from the PPS. In other
words, there are no extra cycles when the vehicle is at the PPS and
as it moves away, it will still know the state of the cycles at the
PPS and can then calculate the increase or decrease in the cycles
at the host vehicle as it moves relatively away from or closer to
the transmitting satellite. There is no ambiguity when the vehicle
is at the PPS station and that is maintained as long as the lock on
a satellite is not lost for more than a few minutes providing that
there is an accurate clock within the vehicle.
There are other sources of information that can be added to
increase the accuracy of position determination. The use of GPS
with four satellites provides the three-dimensional location of the
vehicle plus time. Of the dimensions, the vertical position is the
least accurately known, yet, if the vehicle knows where it is on
the roadway, the vertical dimension is not only the least important
but it is also already accurately known from the roadmap
information plus the inertial guidance system.
2.4 Inertial Navigation System
In the system of the inventions herein, the vehicle will generally
have an inertial measurement unit, inertial reference unit or an
inertial navigation system which for the purposes herein will be
treated as identical. Such a device typically has three
accelerometers and three gyroscopes that are held together in a
single housing. Typically, these 6 devices are MEMS devices and
inherently are very inexpensive. Some companies then proceed to
carefully test each component to determine the repeatable effects
that various environmental factors and aging has on the performance
of each device and then associates with each device, a calibration
or constitute equation that translates the readings of the device
to actual values based on the environmental variable values and
time. This process adds significantly to the cost and in fact may
be the dominant cost. The problem is that age, for example, may
affect a device differently based on how the aging takes place, at
high or low temperatures, for example. Also shock or some other
unexpected event can change the properties of a device. In the
present invention, on the other hand, this complicated and
expensive calibration process is not performed and thus a
calibration equation is not frozen into the device. Since the IMU
will be part of a vehicle system and that system will periodically,
either from the GPS-DGPS type system or from the PPS, know its
exact location, that fact will be used to derive a calibration
equation for each device and since other information such as
temperature etc. will also be known that parameter can also be part
of the equation. The equation can thus be a changing part of the
system that automatically adjusts to actual experience of the
vehicle in the field. Thus, not only is the IMU more accurate than
the prior art but it is considerably less expensive. One method for
handling this change and recalculation of the calibration equations
would be to use an adaptive neural network that has a forgetting
function. Properly designed, this network can allow the calibration
equations to adjust and slowly change over time always providing
the most accurate values regardless of how the devices are changing
in their sensitivity to temperature, for example.
The fact that the IMU resident devices are continuously calibrated
using external measurements renders the IMU an extremely accurate
device comparable with military grade IMUs costing thousands of
dollars. The IMU is far more accurate, for example, than the crash
sensor or chassis control accelerometers and gyroscopes that are
currently being deployed on a vehicle. Thus, when mounting location
considerations permit, the IMU can take over the functions
currently performed by these other devices. This will not only
increase the accuracy of these other functions but reduce the total
cost by eliminating the need for redundant parts and permitting
economies in the electronic circuits and processors to be realized.
The airbag SDM can now be housed with the IMU, for example, and
similarly for the chassis control electronics. If the IMU has the
full complement of three gyros and three accelerometers, then this
additional information can be used to substantially improve the
crash sensing algorithms or the chassis control algorithms. The
sensing and predicting or a rollover event, for example, and the
subsequent control of the throttle, brakes and steering systems as
well as the timely deployment of the side and curtain airbags.
Thus, the use of the IMU for these functions, particularly for the
rollover prediction, mitigation and restraint deployment functions,
are a key teaching of this invention.
As discussed below, many sensors can be used to correct the errors
in the IMU in addition to the GPS and PPS based systems. A gravity
meter can determine the direction of vertically down and can
especially be used when the vehicle is not moving. A magnetic flux
gate compass and/or declinometer values can be included in the map
database and compared by the host vehicle as it passes mapped
areas. Doppler radar or other velocity measurements from the
exterior vehicle monitoring system can provide valuable velocity
information. Vision systems can be used to correct for position if
such data is stored on the map database. If, for example a stored
picture shows a signpost at a particular location that can be
viewed by a resident vision system, then this can also be useful
information for correcting errors in the IMU.
In many cases, especially before the system implementation becomes
mature and the complete infrastructure is in place, there will be
times when a particular vehicle system is not operational. This
could be due to obstructions hiding a clear view of a sufficient
number of GPS satellites, such as when a vehicle enters a tunnel.
It could also be due to a lack of road boundary information, due to
construction or the fact that the road has not been surveyed and
the information recorded and made available to the vehicle, or a
variety of other causes. It is contemplated, therefore, that each
equipped vehicle will contain a warning light or other system that
warns the driver or the vehicle control system when the system is
not operational. If this occurs on one of the specially designated
highway lanes, the vehicle speed will be reduced until the system
again becomes operational.
When the system is non-operational for a short distance, the
vehicle will still accurately know its position if there is, in
addition, one or more laser gyroscopes, micromachined angular rate
sensors or equivalent, and one or more accelerometers that together
are referred to as an Inertial Navigation System (INS, IMU) or
inertial measurement unit (IMU). Generally, such an INS will have
three gyroscopes and three accelerometers and frequently there may
be more than one IMU in a vehicle. Although current versions of the
IMU use MEMS devices, progress is being made on fiber optic-based
gyroscopes. Thus, the present invention is not limited to MEMS
devices but will make use of the best cost effective devices that
are available at a particular time.
As more sensors which are capable of providing information on the
vehicle position, velocity and acceleration are added onto the
vehicle, the system can become sufficiently complicated as to
require a Kalman filter, neural network, or neural-fuzzy, system to
permit the optimum usage of the available information. This becomes
even more important when information from outside the vehicle other
than the GPS related systems becomes more available. For example, a
vehicle may be able to communicate with other vehicles that have
similar systems and learn their estimated location. If the vehicle
can independently measure the position of the other vehicle, for
example through the use of the scanning laser radar system
described below, the differential GPS readings as discussed above,
and thereby determine the relative position of the two or more
vehicles, a further improvement of the position can be determined
for all such vehicles. Adding all such additional information into
the system would probably require a computational method such as
Kalman filters, neural networks or a combination thereof and
perhaps a fuzzy logic system.
One way to imagine the system operation is to consider each car and
roadway edge to behave as if it had a surrounding "force field"
that would prevent it from crashing into another vehicle or an
obstacle along the roadway. A vehicle operator would be prevented
from causing his or her vehicle to leave its assigned corridor.
This is accomplished with a control system that controls the
steering, acceleration and perhaps the vehicle brakes based on its
knowledge of the location of the vehicle, highway boundaries and
other nearby vehicles. In a preferred implementation, the location
of the vehicle is determined by first using the GPS L1 signal to
determine its location within approximately 100 meters. Then, using
DGPS and corrections which are broadcast, whether by FM or
downloaded from geo-synchronous (GEO) or Low Earth Orbiting (LEO)
satellites or obtained from another vehicle or road-based
transmitters, to determine its location within less than about 10
centimeters. Finally, the use of a PPS, discussed below,
periodically permits the vehicle to determine its exact location
and thereby determine the GPS corrections, eliminate the carrier
cycle ambiguity and correct the errors in the INS or IMU system. If
this is still not sufficient, then the phase of the carrier
frequency provides the required location information to less than a
few centimeters. Dead reckoning, using vehicle speed, steering
angle and tire rotation information and inertial guidance, can be
used to fill in the gaps. Where satellites are out of view,
pseudolites, or other systems, can be placed along the highway. A
pulsed scanning infrared laser or terahertz radar system, or an
equivalent system, can be used for obstacle detection.
Communication to other vehicles is by short distance radio or by
spread spectrum time domain pulse radar or terahertz.
3. Maps and Mapping
3.1 Maps
All information regarding the road, both temporary and permanent,
should be part of the map database, including speed limits,
presence of guard rails, width of each lane, width of the highway,
width of the shoulder, character of the land beyond the roadway,
existence of poles or trees and other roadside objects, exactly
where the precise position location apparatus is located, etc. The
speed limit associated with particular locations on the maps should
be coded in such a way that the speed limit can depend upon the
time of day and the weather conditions. In other words, the speed
limit is a variable that will change from time to time depending on
conditions. It is contemplated that there will be a display for
various map information which will always be in view for the
passenger and for the driver at least when the vehicle is operating
under automatic control. Additional user information can thus also
be displayed such as traffic conditions, weather conditions,
advertisements, locations of restaurants and gas stations, etc.
A map showing the location of road and lane boundaries can be
easily generated using a specially equipped survey vehicle that has
the most accurate position measurement system available. In some
cases, it might be necessary to set up one or more temporary local
DGPS base stations in order to permit the survey vehicle to know
its position within a few centimeters. The vehicle would drive down
the roadway while operators, using specially designed equipment,
sight the road edges and lanes. This would probably best be done
with laser pointers and cameras. Transducers associated with the
pointing apparatus record the angle of the apparatus and then by
triangulation determine the distance of the road edge or lane
marking from the survey vehicle. Since the vehicle's position would
be accurately known, the boundaries and lane markings can be
accurately determined. It is anticipated that the mapping activity
would take place continuously such that all roads in a particular
state would be periodically remapped in order to record any changes
which were missed by other monitoring systems and to improve the
reliability of the maps by minimizing the chance for human error.
Any roadway changes that were discovered would trigger an
investigation as to why they were not recorded earlier thus adding
feedback to the mapping part of the process.
The above-described method depends on human skill and attention and
thus is likely to result in many errors. A preferred approach is to
carefully photograph the edge of the road and use the laser
pointers to determine the location of the road lines relative to
the pointers and to determine the slope of the roadway through
triangulation. In this case, several laser pointers would be used
emanating from above, below and to the sides of the camera. The
reduction of the data is then done later using equipment that can
automatically pick out the lane markings and the reflected spots
from the laser pointers. One aid to the mapping process is to place
chemicals in the line paint that could be identified by the
computer software when the camera output is digitized. This may
require the illumination of the area being photographed by an
infrared or ultraviolet light, for example.
In some cases where the roadway is straight, the survey vehicle
could travel at moderate speed while obtaining the boundary and
lane location information. In other cases, where the road in
turning rapidly, more readings would be required per mile and the
survey vehicle would need to travel more slowly. In any case, the
required road information can be acquired semi-automatically with
the survey vehicle traveling at a moderate speed. Thus, the mapping
of a particular road would not require significant time or
resources. It is contemplated that a few such survey vehicles could
map all of the interstate highways in the U.S. in less than one
year.
The mapping effort could be supplemented and cross-checked though
the use of accurate detailed digital photogrammetic systems which,
for example, can determine the road altitude with an accuracy to
<50 cm. Efforts are underway to map the earth with 1-meter
accuracy. The generated maps could be used to check the accuracy
and for missing infrastructure or other roadside installations of
the road-determined maps.
Another improvement that can be added to the system based on the
maps is to use a heads-up display for in-vehicle signage. As the
vehicle travels down the road, the contents of roadside signs can
be displayed on a heads up display, providing such a display is
available in the vehicle, or on a specially installed LCD display.
This is based on the inclusion in the map database of the contents
of all highway signs. A further improvement would be to include
signs having varying messages which would require that the message
be transmitted by the sign to the vehicle and received and
processed for in-vehicle display. This could be done either
directly, by satellite, the Internet, cell phone etc.
As the roadway is being mapped, the availability of GPS satellite
view and the presence of multipath reflections from fixed
structures can also be determined. This information can then be
used to determine the advisability of locating a local precise
location system (PPS), or other infrastructure, at a particular
spot on the roadway. Cars can also be used as probes for this
process and for continuous improvement to check the validity of the
maps and report any errors.
Multipath is the situation where more than one signal from a
satellite comes to a receiver with one of the signals resulting
from a reflection off of a building or the ground, for example.
Since multipath is a function of geometry, the system can be
designed to eliminate its effects based on highway surveying and
appropriate antenna design. Multipath from other vehicles can also
be eliminated since the location of the other vehicles will be
known.
3.2 Mapping
An important part of some embodiments of the invention is the
digital map that contains relevant information relating to the road
on which the vehicle is traveling. The digital map usually includes
the location of the edge of the road, the edge of the shoulder, the
elevation and surface shape of the road, the character of the land
beyond the road, trees, poles, guard rails, signs, lane markers,
speed limits, etc. as discussed in more detail elsewhere herein.
Additionally, it can contain the signature as discussed above. This
data or information is acquired in a unique manner for use in the
invention and the method for acquiring the information and its
conversion to a map database that can be accessed by the vehicle
system is part of this invention. The acquisition of the data for
the maps will now be discussed. It must be appreciated though that
the method for acquiring the data and forming the digital map can
also be used in other inventions.
Local area differential GPS can be utilized to obtain maps with an
accuracy of about 2 cm (one sigma). Temporary local differential
stations are available from such companies as Trimble Navigation.
These local differential GPS stations can be placed at an
appropriate spacing for the road to be mapped, typically every 30
kilometers. Once a local differential GPS station is placed, it
requires some time period such as an hour or more for the station
to determine its precise location. Therefore, sufficient stations
are required to cover the area that is to be mapped within, for
example, four hours. This may require as many as 10 or more such
differential stations for efficient mapping.
With reference to FIGS. 16A, 16B, 17 and 17A, a mapping vehicle 200
is used and obtains its location from GPS satellites and its
corrections from the local differential stations. Such a system is
capable of providing the 2 cm accuracy desired for the map
database. Typically, at least two GPS receivers 226 are mounted on
the mapping vehicle 200. Each GPS receiver 226 is contained within
or arranged in connection with a respective data acquisition module
202, which data acquisition modules 202 also contain a GPS antenna
204, an accurate inertial measurement unit (IMU) 206, a
forward-looking video camera 208, a downward and outward looking
linear array camera 210 and a scanning laser radar 212. The
relative position of these components in FIG. 17 is not intended to
limit the invention.
A processor including a printed circuit board 224 is coupled to the
GPS receivers 226, the IMUs 206, the video cameras 208, the linear
cameras 210 and the scanning laser radars 212 (see FIG. 17A). The
processor 224 receives information regarding the position of the
vehicle from the GPS receivers 226, and optionally the IMUs 206,
and the information about the road from both linear cameras 210 or
from both laser radars 212, or from all of the linear cameras 210
and laser radars 212, and forms the road map database. Information
about the road can also come from one or both of the video cameras
208 and be incorporated into the map database.
The map database can be of any desired structure or architecture.
Preferred examples of the database structure are of the type
discussed in U.S. Pat. No. 6,144,338 (Davies) and U.S. Pat. No.
6,247,019 (Davies), incorporated by reference herein.
The data acquisition modules 202 are essentially identical and each
mounts to the vehicle roof on an extension assembly 214 which
extends forward of the front bumper. Extension assembly 214
includes a mounting bracket 216 from the roof of the vehicle 200
forward to each data acquisition module 210, a mounting bracket 218
extending from the front bumper upward to each data acquisition
module 202 and a cross mounting bracket 220 extending between the
data acquisition modules 202 for support. Since all of the data
acquisition equipment is co-located, its precise location is
accurately determined by the IMU, the mounting location on the
vehicle and the differential GPS system.
The forward-looking video cameras 208 provide views of the road as
shown in FIG. 18. These cameras 208 permit the database team to
observe the general environment of the road and to highlight any
anomalies. They also permit the reading of traffic signs and other
informational displays all of which can be incorporated into the
database. The cameras 208 can be ordinary color video cameras,
high-speed video cameras, wide angle or telescopic cameras, black
and white video cameras, infrared cameras, etc. or combinations
thereof. In some cases, special filters are used to accentuate
certain features. For example, it has been found that lane markers
frequently are more readily observable at particular frequencies,
such as infrared. In such cases, filters can be used in front of
the camera lens or elsewhere in the optical path to block unwanted
frequencies and pass desirable frequencies. Polarizing lenses have
also been found to be useful in many cases. Normally, natural
illumination is used in the mapping process, but for some
particular cases, particularly in tunnels, artificial illumination
can also be used in the form of a floodlight or spotlight that can
be at any appropriate frequency of the ultraviolet, visual and
infrared portions of the electromagnetic spectrum or across many
frequencies. Laser scanners can also be used for some particular
cases when it is desirable to illuminate some part of the scene
with a bright spot. In some cases, a scanning laser rangemeter can
be used in conjunction with the forward-looking cameras 204 to
determine the distance to particular objects in the camera
view.
The video camera system can be used by itself with appropriate
software as is currently being done by Lamda Tech International
Inc. of Waukesha, Wis., to obtain the location of salient features
of a road. However, such a method to obtain accurate maps is highly
labor intensive and therefore expensive. The cameras and associated
equipment in the present invention are therefore primarily used to
supplement the linear camera and laser radar data acquisition
systems to be described now.
The mapping vehicle data acquisition modules will typically contain
both a linear camera and a scanning laser radar, however, for some
applications one or the other may be omitted.
The linear camera 210 is a device that typically contains a linear
CCD, CMOS or other light sensitive array of, for example, four
thousand pixels. An appropriate lens provides a field of view to
this camera that typically extends from approximately the center of
the vehicle out to the horizon. This camera records a
one-dimensional picture covering the entire road starting with
approximately the center of the lane and extending out to the
horizon. This linear array camera 210 therefore covers slightly
more than 90 degrees. Typically, this camera operates using natural
illumination and produces effectively a continuous picture of the
road since it obtains a linear picture, or column of pixels, for
typically every one-inch of motion of the vehicle. Thus, a complete
two-dimensional panoramic view of the road traveled by the mapping
vehicle is obtained. Since there are two such linear camera units,
a 180 degree view is obtained. This camera will typically record in
full color thus permitting the map database team to have a complete
view of the road looking perpendicular from the vehicle. The view
is recorded in a substantially vertical plane. This camera will not
be able to read text on traffic signs, thus the need for the
forward-looking cameras 208. Automated software can be used with
the images obtained from these cameras 208, 210 to locate the edge
of the road, lane markers, the character of land around and
including the road and all areas that an errant vehicle may
encounter. The full color view allows the characterization of the
land to be accomplished automatically with minimal human
involvement.
The scanning laser radar 212 is typically designed to cover a 90
degree or less scan thus permitting a rotating mirror to acquire at
least four such scans per revolution. The scanning laser radar 212
can be coordinated or synchronized with the linear camera 210 so
that each covers the same field of view with the exception that the
camera 210 typically will cover more than 90 degrees. Scanning
laser radar 212 can be designed to cover more or less than 90
degrees as desired for a particular installation. The scanning
laser radar 212 can operate in any appropriate frequency from above
ultraviolet to the terahertz. Typically, it will operate in the
eye-safe portion of the infrared spectrum for safety reasons. The
scanning laser radar 212 can operate either as a pulse-modulated or
a tone-modulated laser as is known in the art. If operating in the
tone-modulated regime, the laser light will be typically modulated
with three or more frequencies in order to eliminate distance
ambiguities. Noise or code modulated radar can also be used.
For each scan, the laser radar 212 provides the distance from the
scanner to the ground for up to several thousand points in a
vertical plane extending from approximately the center of the lane
out to near the horizon. This device therefore provides precise
distances and elevations to all parts of the road and its
environment. The precise location of signs that were observed with
the forward-looking cameras 204, for example, can now be easily and
automatically retrieved. The scanning laser radar therefore
provides the highest level of mapping automation.
Scanning laser radars have been used extensively for mapping
purposes from airplanes and in particular from helicopters where
they have been used to map portions of railway lines in the U.S.
Use of the scanning laser radar system for mapping roadways where
the radar is mounted onto a vehicle that is driving the road is
believed to be novel to the current assignee.
Ideally, all of the above-described systems are present on the
mapping vehicle. Although there is considerable redundancy between
the linear camera and the scanning laser radar, the laser radar
operates at one optical frequency and therefore does not permit the
automatic characterization of the roadway and its environment.
As with the forward-looking cameras, it is frequently desirable to
use filters and polarizing lenses for both the scanning laser radar
and the linear camera. In particular, reflections from the sun can
degrade the laser radar system unless appropriate filters are used
to block all frequencies except frequency chosen for the laser
radar.
Laser radars are frequently also referred to as ladars and lidars.
All such devices that permit ranging to be accomplished from a
scanning system, including radar, are considered equivalent for the
purposes of this invention.
3.3 Map Enhancements
Once the road edge and lane locations, and other roadway
information, are transmitted to the operator, it requires very
little additional bandwidth to include other information such as
the location of all businesses that a traveler would be interested
in such as gas stations, restaurants etc. which could be done on a
subscription basis. This concept was partially disclosed in the
'482 patent discussed above and partially implemented in existing
map databases.
Communication of information to the operator could be done either
visually or orally as described in U.S. Pat. No. 5,177,685 or U.S.
patent application Ser. No. 09/645,709 filed Aug. 14, 2000, now
U.S. Pat. No. 7,126,583. Finally, the addition of a route guidance
system as described in other patents becomes even more feasible
since the exact location of a destination can be determined. The
system can be configured so that a vehicle operator could enter a
phone number, for example, or an address and the vehicle would be
automatically and safely driven to that location. Since the system
knows the location of the edge of every roadway, very little, if
any, operator intervention would be required. Even a cell phone
number can be used if the cell phone has the SnapTrack GPS location
system as soon to be provided by Qualcomm.
Very large may databases can now reside on a vehicle as the price
of memory continues to drop. Soon it may be possible to store the
map database of an entire country on the vehicle and to update it
as changes are made. The area that is within, for example, 1000
miles from the vehicle can certainly be stored and as the vehicle
travels from place to place the remainder of the database can be
updated as needed though a connection to the Internet, for
example.
4. Precise Positioning
Another important aid as part of some of the inventions disclosed
herein is to provide markers along the side(s) of roadways which
can be either visual, passive or active transponders, reflectors,
or a variety of other technologies including objects that are
indigenous to or near the roadway, which have the property that as
a vehicle passes the marker it can determine the identity of the
marker and from a database it can determine the exact location of
the marker. The term "marker" is meant in the most general sense.
The signature determined by a continuous scan of the environment,
for example, would be a marker if it is relatively invariant over
time such as, for example, buildings in a city. Basically, there is
a lot of invariant information in the environment surrounding a
vehicle as it travels down a road toward its destination. From time
to time, a view of this invariant landscape or information may be
obstructed but it is unlikely that all of it will be during the
travel of a mile, for example. Thus, a vehicle should be able to
match the signature sensed with the expected one in the map
database and thereby obtain a precise location fix. This signature
can be obtained through the use of radar or laser radar
technologies as reported elsewhere herein. See in particular
Section 5 below and for example, Wang Yanli, Chen Zhe, "Scene
matching navigation based on multisensor image fusion" SPIE Vol.
5286 p. 788 793, 2003.
For the case of specific markers placed on the infrastructure, if
three or more of such markers are placed along a side of the
roadway, a passing vehicle can determine its exact location by
triangulation. Note that even with two such markers using radar
with distance measuring capability, the precise position of a
vehicle can be determined as discussed below in reference to the
Precise Positioning System. In fact, if the vehicle is only able to
observe a single radar reflector and take many readings as the
reflector is passed, it can determine quite accurately its position
based on the minimum distance reading that is obtained during the
vehicle's motion past the reflector. Although it may be impractical
to initially place such markers along all roadways, it would be
reasonable to place them in particularly congested areas or places
where it is known that a view of one or more of the GPS satellites
is blocked. A variation of this concept will be discussed
below.
Although initially it is preferred to use the GPS navigational
satellites as the base technology, the invention is not limited
thereby and contemplates using all methods by which the location of
the vehicle can be accurately determined relative to the earth
surface. The location of the roadway boundaries and the location of
other vehicles relative to the earth surface are also to be
determined and all relevant information used in a control system to
substantially reduce and eventually eliminate vehicle accidents.
Only time and continued system development will determine the mix
of technologies that provide the most cost effective solution. All
forms of information and methods of communication to and between
vehicles are contemplated including direct communication with
stationary and moving satellites, communication with fixed
earth-based stations using infrared, optical, terahertz, radar,
radio and other segments of the electromagnetic spectrum and
inter-vehicle communication. Some additional examples follow:
A pseudo-GPS can be delivered from cell phone stations, in place of
or in addition to satellites. In fact, the precise location of a
cell phone tower need not initially be known. If it monitors the
GPS satellites over a sufficiently long time period, the location
can be determined as the calculated location statistically
converges to the exact location. Thus, every cell phone tower could
become an accurate DGPS base station for very little cost. DGPS
corrections can be communicated to a vehicle via FM radio via a
sub-carrier frequency for example. An infrared or radar transmitter
along the highway can transmit road boundary location information.
A CD-ROM or other portable mass storage can be used at the
beginning of a controlled highway to provide road boundary
information to the vehicle. Finally, it is contemplated that
eventually a satellite will broadcast periodically, perhaps every
five minutes, a table of dates covering the entire CONUS that
provides the latest update date of each map segment. If a
particular vehicle does not have the latest information for a
particular region where it is operating, it will be able to use its
cell phone or other communication system to retrieve such road maps
perhaps through the Internet or from an adjacent vehicle. Emergency
information would also be handled in a similar manner so that if a
tree fell across the highway, for example, all nearby vehicles
would be notified.
One of the possible problems with the RtZF.TM. system described
herein is operation in certain areas of large cities such as lower
Manhattan, N.Y. In such locations, unless there are a plurality of
local pseudolites or precise position location system installations
or the environment signature system is invoked such as with
adaptive associative memories as described above, the signals from
the GPS satellites can be significantly blocked. Also there is
frequently a severe multipath problem in cities. A solution is to
use the LORAN system as a backup for such locations. The accuracy
of LORAN can be comparable to DGPS. Use of multiple roadway-located
Precise Positioning Systems would be a better solution or a
complementary solution. Additionally, some location improvement can
result from application of the SnapTrack system as described in
U.S. Pat. No. 5,874,914 and other patents to Krasner of
SnapTrack.
The use of geo-synchronous satellites as a substitute for earth
bound base stations in a DGPS system, with carrier phase
enhancements for sub-meter accuracies, is also a likely improvement
to the RtZF.TM. system that can have a significant effect in urban
areas.
Another enhancement that would be possible with dedicated
satellites and/or earth bound pseudolites results from the greater
control over the information transmitted than is available from the
present GPS system. Recognizing that this system could save in
excess of 40,000 lives per year in the U.S. alone, the cost of
deploying such special purpose stations can easily be justified.
For example, say there exists a modulated wave that is 10000
kilometers long, another one which is 1000 km long etc. down to 1
cm. It would then be easy to determine the absolute distance from
one point to the other. The integer ambiguity of RTK DGPS would be
eliminated. Other types of modulation are of course possible to
achieve the desired result of simply eliminating the carrier
integer uncertainty that is discussed in many U.S. patents and
other literature. This is not meant to be a recommendation but to
illustrate that once the decision has been made to provide
information to every vehicle that will permit it to always know its
location within 10 cm, many technologies will be there to make it
happen. The cost savings resulting from eliminating fatalities and
serious injuries will easily cover the cost of such technologies
many times over. The provision of additional frequencies can also
enhance the system and render differential corrections unnecessary.
Each frequency from a satellite is diffracted differently by the
ionosphere. The properties of the ionosphere can thus be determined
if multiple frequencies are transmitted. This will partially be
achieved with the launch of the European Galileo GPS satellite
system in combination with the U.S. GPS system.
It is expected, especially initially, that there will be many holes
in the DGPS or GPS and their various implementations that will
leave the vehicle without an accurate means of determining its
location. The inertial navigation system described above will help
in filling these holes but its accuracy is limited to a time period
significantly less than an hour and a distance of less than 50
miles before it needs correcting. That may not be sufficient to
cover the period between DGPS availability. It is therefore
contemplated that the RtZF.TM. system will also make use of low
cost systems located along the roadways that permit a vehicle to
accurately determine its location. One example of such a system
would be to use a group of three Micropower Impulse Radar (MIR)
units such as developed by Lawrence Livermore Laboratory.
A MIR operates on very low power and periodically transmits a very
short spread spectrum radar pulse. The estimated cost of a MIR is
less than $10 even in small quantities. If three such MIR
transmitters, 151, 152 and 153, as shown in FIG. 11, are placed
along the highway and triggered simultaneously or with a known
delay, and if a vehicle has an appropriate receiver system, the
time of arrival of the pulses can be determined and thus the
location of the vehicle relative to the transmitters determined.
The exact location of the point where all three pulses arrive
simultaneously would be the point that is equidistant from the
three transmitters 151, 152, 153 and would be located on the map
information. Only three devices are required since only two
dimensions need to be determined and it is assumed that the vehicle
in on the road and thus the vertical position is known, otherwise
four MIRs would be required. Thus it would not even be necessary to
have the signals contain identification information since the
vehicle would not be so far off in its position determination
system to confuse different locations. By this method, the vehicle
would know exactly where it was whenever it approached and passed
such a triple-MIR installation. The MIR triad PPS or equivalent
could also have a GPS receiver and thereby determine its exact
location over time as described above for cell phone towers. After
the location has been determined, the GPS receiver can be removed.
In this case, the MIR triad PPS or equivalent could be placed at
will and they could transmit their exact location to the passing
vehicles. An alternate method would be to leave the GPS receiver
with the PPS time of arrival of the GPS data from each satellite so
that the passing vehicles that do not go sufficiently close to the
PPS can still get an exact location fix. A similar system using
RFID tags is discussed below.
Several such readings and position determinations can be made with
one approach to the MIR installation, the vehicle need not wait
until they all arrive simultaneously. Also the system can be
designed so that the signals never arrive at the same time and
still provide the same accuracy as long as there is a sufficiently
accurate clock on board the vehicle. One way at looking at FIG. 11
is that transmitters 151 and 152 fix the lateral position of the
vehicle while transmitters 151 and 153 fix the location of the
vehicle longitudinally. The three transmitters 151,152,153 need not
be along the edges on one lane but could span multiple lanes and
they need not be at ground level but could be placed sufficiently
in the air so that passing trucks would not block the path of the
radiation from an automobile. Particularly in congested areas, it
might be desirable to code the pulses and to provide more than
three transmitters to further protect against signal blockage or
multipath.
The power requirements for the MIR transmitters are sufficiently
low that a simple photoelectric cell array can provide sufficient
power for most if not all CONUS locations. With this exact location
information, the vehicle can become its own DGPS station and can
determine the corrections necessary for the GPS. It can also
determine the integer ambiguity problem and thereby know the exact
number of wave lengths between the vehicle and the satellites or
between the vehicle and the MIR station. These calculations can be
done on vehicle if there is a connection to a network, for example.
This would be particularly efficient as the network, once it had
made the calculations for one vehicle, would have a good idea of
the result for another nearby vehicle and for other vehicles
passing the same spot at a different time.
MIR is one of several technologies that can be used to provide
precise location determination. Others include the use of an RFID
tag that is designed in cooperation with its interrogator to
provide a distance to the tag measurement. Such as RFID can be
either an active device with an internal battery or solar charger
or a passive device obtaining its power from an RF interrogation
signal to charge a capacitor or a SAW-based tag operating without
power. An alternate and preferred system uses radar or other
reflectors where the time of flight can be measured, as disclosed
in more detail elsewhere herein.
Once a vehicle passes a Precise Positioning Station (PPS) such as
the MIR triad described above, the vehicle can communicate this
information to surrounding vehicles. If the separation distance
between two communicating vehicles can also be determined by the
time-of-flight or equivalent method, then the vehicle that has just
passed the triad can, in effect, become a satellite equivalent or
moving pseudolite. That is, the vehicle sends (such as by
reflection so as not to introduce a time delay) its GPS data from
the satellite and the receiving vehicle then gets the same message
from two sources and the time difference is the time of flight.
Finally, if many vehicles are communicating their positions to many
other vehicles along with an accuracy of position assessment, each
vehicle can use this information along with the calculated
separation distances to improve the accuracy of its position
determination. In this manner, as the number of such vehicles
increases, the accuracy of the entire system increases until an
extremely accurate positioning system for all vehicles results.
Such a system, since it combines many sources of position
information, is tolerant of the failure of any one or even several
such sources. Thus, the RtZF.TM. system becomes analogous to the
Internet in that it can't be shut down and the goal of perfection
is approached. Some of the problems associated with this concept
will be discussed in more detail below.
Precise Positioning was described in detail above and relates to
methods of locating a vehicle independently of GPS within sub meter
accuracy. This can be done using an MIR triads; barcodes painted on
the roadway; radar, laser radar or terahertz radar and
infrastructure mounted reflectors; RFID markets; or through the use
of matching a signature obtained from the environment with a stored
signature using, for example, Adaptive Associative Memories (AAS)
based on Cellular Neural Networks (CNN).
AAS is a type of neural network that is distinguished in that it
can do precise identification from poor and sparse data in contrast
to ordinary back propagation neural networks discussed elsewhere
herein that generalize and always give an approximate answer.
Applications for AAS include: (1) Occupant recognition (face, iris,
voice print, fingerprints etc.), and (2) Vehicle location
recognition for the RtZF.TM. Precise Positioning System, which is
the focus here. In contrast to other PPS systems described above,
AAS permits the precise location of a vehicle on a roadway within
centimeters without the use of additions to the infrastructure. A
radar, laser scanner, or terahertz radar continuously is projected
from the vehicle toward the environment, such as the roadway to the
side of the vehicle, and from the returned reflected waves it
obtains a signature of the passing environment and compares it with
a recorded signature using ASM. This signature, for example, can be
the distance from the vehicle to the infrastructure which has been
normalized for the purpose of signature matching with some method
such as the average or some other datum. Thus it is the relative
distance signature that can be compared with a stored signature
thus removing the position of the vehicle on the roadway as a
variable. When a match is found the distance to a precise object
can be determined placing the vehicle precisely on the road in both
the longitudinal and lateral dimensions. As discussed above, this
can make the vehicle a DGPS station for correction of the GPS
errors but it also can be used as the primary location system
without GPS.
Other methods can be used to precisely locate a vehicle using the
infrastructure and only one preferred method has been described
herein. For example, the vertical motion signature of the vehicle
can in some cases be used. This could involve determining this
signature from the natural road or a pattern of disturbances
similar to a rubble strip can be placed in the roadway and sensed
by an accelerometer, microphone or other sensor. Even the signature
of the magnetic or reflective properties of the roadway or the
environment at the side of road can be candidates with the
appropriate sensors. Basically, any system that provides a
signature indication location that is derived from the
infrastructure with appropriate sensors would qualify.
Another method, for example, is to match camera images where again
an AAM can be used. Since the vehicle knows approximately where it
is, the recorded signature used in the AAM will change as the
vehicle moves and thus only a small amount of data need be used at
a particular time. The AAM system is fast and relatively simple.
Typically twenty data points will be used to determine the match,
for example. What follows is a general description of AAM
Associative (context-addressable) memory is frequently dedicated to
data search and/or restoration from available fragments.
Associative retrieval requires minimal information on sought
objects, so such a machine might be used for most complicated tasks
of data identification for partially destroyed or corrupted images.
It can be applied to authenticity attribution, document
falsification detection, message fragment identification in the
Internet etc. as well as signature matching with the environment
for PPS.
Neural associative memory works due to multi-stability of strong
feedback systems. Common models like Hopfield networks and
bi-directional associative memory provide memorization by means of
computation network weights. It does not corrupt previously stored
images. Unfortunately, these networks cannot be widely used because
of their low capacity and inefficient physical memory usage. A
number M of vectors memorized does not exceed 14% of the number of
neurons in the network N. Since a network contains N.sup.2
connections, it needs storage of at least 25M.sup.2 real weight
values.
Cellular architecture can exhaustively solve the problem of
physical memory usage. Cellular memories have band-like synaptic
matrix. The volume (number of elements) grows linearly with respect
to neuron number. This is why cellular neural networks (CNNs) can
be useful for very large data processing problems. Pioneering
models of associative memories via CNNs were proposed in some
earlier works. However, more detailed studies showed some
fundamental limitations. Indeed, it has now been shown that the
number of images stored is restricted by a cell size. Hence, it
does not depend on the number of neurons. A more efficient way of
redundancy reduction has also been found due to connection
selection after training. This results in the use of only a small
part of physical memory without corruption of memorized data. The
network after weight selection looks like the cellular one; so by
combining cellular training algorithms and weight selection, a
novel network paradigm has resulted. It is an adaptive neural
paradigm with great memorizing capacity.
At present, some breakthrough associative memories have been
implemented in a software package available from the current
assignee. The results can be applied for processing of large
databases, real-time information retrieval, PPS etc. Other
applications for this technology include face, iris, fingerprint,
voiceprint, character, signature, etc. recognition.
FIG. 11 shows the implementation of the invention using the Precise
Positioning System (PPS) 151, 152, 153, in which a pair of vehicles
18, 26 are traveling on a roadway each in a defined corridor
delineated by lines 14 and each is equipped with a system in
accordance with the invention and in particular, each is equipped
with PPS receivers. Four versions of the PPS system will now be
described. This invention is not limited to these examples but they
will serve to illustrate the principals involved.
Vehicle 18 contains two receivers 160,161 for the micropower
impulse radar (MIR) implementation of the invention. MIR
transmitter devices are placed at locations 151,152 and 153
respectively. They are linked together with a control wire, not
shown, or by a wireless connection such that each device transmits
a short radar pulse at a precise timing relative to the others.
These pulses can be sent simultaneously or at a precise known
delay. Vehicle 18 knows from its map database the existence and
location of the three MIR transmitters. The transmitters 151,152
and 153 can either transmit a coded pulse or non-coded pulse. In
the case of the coded pulse, the vehicle PPS system will be able to
verify that the three transmitters 151, 512, 153 are in fact the
ones that appear on the map database. Since the vehicle will know
reasonably accurately its location and it is unlikely that other
PPS transmitters will be nearby or within range, the coded pulse
may not be necessary. Two receivers 160 and 161 are illustrated on
vehicle 18. For the MIR implementation, only a single receiver is
necessary since the position of the vehicle will be uniquely
determined by the time of arrival of the three MIR pulses. A second
receiver can be used for redundancy and also to permit the vehicle
to determine the angular position of the MIR transmitters as a
further check on the system accuracy. This can be done since the
relative time of arrival of a pulse from one of the transmitters
151, 152, 153 can be used to determine the distance to each
transmitter and by geometry, its angular position relative to the
vehicle 18. If the pulses are coded, then the direction to the MIR
transmitters 151,152,153 will also be determinable.
The micropower impulse radar units require battery power or another
power mechanism to operate. Since they may be joined together with
a wire in order to positively control the timing of the three
pulses, a single battery can be used to power all three units. This
battery can also be coupled with a solar panel to permit
maintenance free operation of the system. Since the MIR
transmitters use very small amounts of power, they can operate for
many years on a single battery.
Although the MIR systems are relatively inexpensive, on the order
of ten dollars each, the installation cost of the system will be
significantly higher than the RFID and radar reflector solutions
discussed next. The MIR system is also significantly more complex
than the RFID system; however, its accuracy can be checked by each
vehicle that uses the system. Tying the MIR system to a GPS
receiver and using the accurate clock on the GPS satellites as the
trigger for the sending of the radar pulses can add additional
advantages and complexity. This will permit vehicles passing by to
additionally accurately set their clocks to be in synchronization
with the GPS clocks. Since the MIR system will know its precise
location, all errors in the GPS signals can be automatically
corrected and in that case, the MIR system becomes a differential
GPS base station. For most implementations, this added complexity
is not necessary since the vehicle themselves will be receiving GPS
signals and they will also know precisely their location from the
triad of MIR transmitters 151, 152, 153.
A considerably simpler alternate approach to the MIR system
described above utilizes reflective RFID tags. These tags, when
interrogated by an interrogator type of receiver 160, 161, reflect
a modified RF signal with the modification being the identification
of the tag. Such tags are described in many patents on RFID
technology and can be produced for substantially less than one
dollar each. The implementation of the RFID system would involve
the accurate placement of these tags on known objects on or in
connection with infrastructure. These objects could be spots on the
highway, posts, signs, sides of buildings, poles, or structures
that are dedicated specifically for this purpose. In fact, any
structure that is rigid and unlikely to change position can be used
for mounting RFID tags. In downtown Manhattan, building sides,
street lights, stoplights, or other existing structures are ideal
locations for such tags. A vehicle 18 approaching a triad of such
RFID tags represented by 151, 152, 153 would transmit an
interrogation pulse from interrogator 160 and/or 161. The pulse
would reflect off of each tag within range and the reflected signal
would be received by the same interrogator(s) 160, 161 or other
devices on the vehicle. Once again, a single interrogator is
sufficient. It is important to note that the range to RFID tags is
severely limited unless a source of power is provided. It is very
difficult to provide enough power from RF radiation from the
interrogator for distances much greater than a few feet. For longer
distances, a power source must be provided which can be a battery,
connection to a power line, solar power, energy harvested from the
environment via vibration, for example, unless the RFID is based on
SAW technology. For SAW technology reading ranges may extend to
tens of meters. Greater distances can be achieved using reflectors
or reflecting antennas.
Electronic circuitry, not shown, associated with the interrogator
160 and/or 161 would determine the precise distance from the
vehicle to the RFID tag 151, 152, 153 based on the round trip time
of flight. This will provide the precise distance to the three RFID
tags 151, 152, 153. Once again, a second interrogator 161 can also
be used, in which case, it could be a receiver only and would
provide redundancy information to the main interrogator 160 and
also provide a second measure of the distance to each of the RFID
tags. Based on the displacement of the two receivers 160, 161, the
angular location of each RFID tag relative into the vehicle can be
determined providing further redundant information as to the
position of the vehicle relative to the tags.
Radar corner reflectors can be placed on poles or other convenient
places such that a radar beam pointed upwards at an angle, such as
30 to 45 degrees from the vehicle, will cause the beam to
illuminate the reflector and thereby cause a reflection to return
to the vehicle. Through well-known methods, the distance to the
reflector can be accurately measured with pulse radar, modulated
radar and phase measurements or noise radar and correlations
measurements. In such a manner, the host vehicle can determine its
position relative to one or more such reflectors and if the
location of the reflector(s) is known and recorded on the map
database, the vehicle can determine its position to within about 2
centimeters. The more reflectors that are illuminated, the better
the accuracy of vehicle location determination. The reflectors can
be simple corner reflectors or a group of reflectors can be
provided giving a return code to the host vehicle.
Using the PPS system, a vehicle can precisely determine its
location within two centimeters or better relative to the MIR, RFID
tags or radar and reflectors and since the precise location of
these devices has previously been recorded on the map database, the
vehicle will be able to determine its precise location on the
surface of the earth. With this information, the vehicle will
thereafter be able to use the carrier wave phase to maintain its
precise knowledge of its location, as discussed above, until the
locks on the satellites are lost. Similarly, the vehicle 18 can
broadcast this information to vehicle 26, for example, permitting a
vehicle that has not passed through the PPS triad to also greatly
improve the accuracy with which it knows its position. Each vehicle
that has recently passed through a PPS triad now becomes a
differential GPS station for as long as the satellite locks are
maintained. Therefore, through inter-vehicle communications, all
vehicles in the vicinity can also significantly improve their
knowledge of their position accuracy resulting in a system which is
extremely redundant and therefore highly reliable and consistent
with the "Road to Zero Fatalities".TM. process. Once this system is
operational, it is expected that the U.S. government and other
governments will launch additional GPS type satellites, each with
more civilian readable frequencies, or other similar satellite
systems, further strengthening the system and adding further
redundancy eventually resulting in a highly interconnected system
that approaches 100% reliability and, like the Internet, cannot be
shut down.
As the system evolves, the problems associated with urban canyons,
tunnels, and other obstructions to satellite view will be solved by
the placement of large numbers of PPS stations, or other devices
providing similar location information.
The final PPS system uses reflected energy from the environment to
create a signature that can be matched with a recorded signature
using a technology such as adaptive associative memories (AAM), or
equivalent including correlation. Since the AAM was discussed
above, the correlation system will be discussed here. As the
mapping vehicle traverses the roadway, it can record the distance
to various roadside objects as a continuous signal having peaks and
valleys. In fact, several such signatures can economically be
recorded such that regardless of where on the roadway a subsequent
vehicle appears, it will record a similar signature. The signature
can be enhanced if dual frequency terahertz is used since the
reflectance from an object can vary significantly from one
terahertz frequency to another depending on the composition of the
object. Thus for one frequency, a metal and a wood object may both
be highly reflective while at another frequency, there can be a
significant difference. Significantly more information is available
when more than one frequency is used.
Using the correlation system, a vehicle will continuously be
comparing its received signature at a particular location to the
previously recorded signature and shifting the two relative to each
other until the best match occurs. Since this will be done
continuously and since we will know the velocity of the vehicle, it
should never deviate significantly from the recorded position and
thus the vehicle will always have a non-GPS method of determining
its exact location. There are certain areas where the signature
matching may be problematic such as going by a wheat field or the
ocean. Fortunately, such wide open spaces are precisely where the
GPS satellite system should work best and similarly the places
where the signature method works best is where the GPS has
problems. Thus, the systems are complementary. In most places, both
systems will work well providing a high degree of redundancy.
Many mathematical methods exist for determining the best shift of
the two signatures (the previously recorded one and a new one) and
therefore the various correlation methods will not be presented
here.
Although the system has been illustrated for use with automobiles,
the same system would apply for all vehicles including trucks,
trains an even airplanes taxiing on runways. It also would be
useful for use with cellular phones and other devices carried by
humans. The combination of the PPS system and cellular phones
permits the precise location of a cellular phone to be determined
within centimeters by an emergency operator receiving a 911 call,
for example. Such RFID tags can be inexpensively placed both inside
and outside of buildings, for example.
The range of RFID tags is somewhat limited to approximately 10
meters for current technology. If there are obstructions preventing
a clear view of the RFID tag by the interrogator, the distance
becomes less. For some applications where it is desirable to use
larger distances, battery power can be provided to the RFID tags.
In this case, the interrogator would send a pulse to the tag that
would turn on the tag and at a precise, subsequent time the tag
would transmit an identification message. In some cases, the
interrogator itself can provide the power to drive the RFID
circuitry, in which case the tag would again operate in the
transponder mode as opposed to the reflective mode.
The RFID tags discussed herein can be either the electronic circuit
or SAW designs.
From the above discussion, those skilled in the art will understand
that other devices can be interrogated by a vehicle traveling down
the road. Such devices might include various radar types or designs
of reflectors, mirrors, other forms of transponders, or other forms
of energy reflectors. All such devices are contemplated by this
invention and the invention is not limited to be specific examples
described. In particular although various frequencies including
radar, terahertz and infrared have been discussed, this invention
is not limited to those portions of the electromagnetic spectrum.
In particular, the X-ray band of frequencies may have some
particular advantages for some external and interior imaging
applications.
Any communication device can be coupled with an interrogator that
utilizes the MIR, radar or RFID PPS system described above. Many
devices are now being developed that make use of the Bluetooth
communication specification. All such Bluetooth enabled devices can
additionally be outfitted with a PPS system permitting the location
of the Bluetooth device to be positively determined. This enabling
technology will permit a base station to communicate with a
Bluetooth-enabled device whose location is unknown and have the
device transmit back its precise location on the surface of the
earth. As long as the Bluetooth-enabled device is within the range
of the base station, its location can be precisely determined.
Thus, the location of mobile equipment in a factory, packages
within the airplane cargo section, laptop computers, cell phones,
PDAs, and eventually even personal glasses or car keys or any
device upon which a Bluetooth-enabled device can be attached can be
determined. Actually, this invention is not limited to Bluetooth
devices but encompasses any device that can communicate with any
other devices.
Once the location of an object can be determined, many other
services can be provided. These include finding the device, or the
ability to provide information to that device or to the person
accompanying that device such as the location of the nearest
bathroom, restaurant, or the ability to provide guided tours or
other directions to people traveling to other cities, for
example.
A particularly important enhancement to the above-described system
uses precise positioning technology independent of GPS. The precise
positioning system, also known as the calibration system, generally
permits a vehicle to precisely locate itself independently of the
IMU or DGPS systems.
One example of this technology involves the use of a radar and
reflector system wherein radar transceivers are placed on the
vehicle that send radar waves to reflectors that are mounted at the
side of road. The location of reflectors either is already
precisely known or is determined by the mapping system during data
acquisition process. The radar transceivers transmit a pulse, code
or frequency or noise modulated radar signal to the road-mounted
reflectors, typically corner reflectors, which reflect a signal
back to the radar transceiver. This permits the radar system to
determine the precise distance from the transceiver to the
reflector by either time-of-flight or phase methods. Note that
although radar will be used below in the illustrations, terahertz
can also be used and thus the word "radar" will be used to cover
both parts of the electromagnetic spectrum.
In one possible implementation, each vehicle is equipped with two
radar devices operating in the 24 77 GHz portion of the spectrum.
Each radar unit will be positioned on the vehicle and aimed
outward, slightly forward and up toward the sides of the roadway.
Poles would be positioned along the roadway at appropriate
intervals and would have multiple corner cube radar reflectors
mounted thereon to thereto, possibly in a vertical alignment. The
lowest reflector on the pole would be positioned so that the
vehicle radar will illuminate the reflector when the vehicle is in
the lane closest to the pole. The highest reflector on the pole
would be positioned so that the vehicle radar will illuminate the
reflector when the vehicle is in the lane most remote from the
pole. The frequency of the positioning of the poles will be
determined by such considerations as the availability of light
poles or other structures currently in place, the probability of
losing access to GPS satellites, the density of vehicle traffic,
the accuracy of the IMU and other similar considerations.
Initially, rough calculations have found that a spacing of about
1/4 mile would likely be acceptable.
If the precise location of the reflectors has been previously
determined and is provided on a road map database, then the vehicle
can use this information to determine its precise location on the
road. In a more typical case, the radar reflectors are installed
and the mapping vehicle knows its location precisely from the
differential GPS signals and the IMU, which for the mapping vehicle
is typically of considerably higher accuracy than will be present
in the vehicles that will later use the system. As a result, the
mapping vehicle can also map a tunnel, for example, and establish
the locations of radar reflectors that will later be used by
non-mapping vehicles to determine their precise location when the
GPS and differential GPS signals are not available. Similarly, such
radar reflectors can be located for an appropriate distance outside
of the tunnel to permit an accurate location determination to be
made by a vehicle until it acquires the GPS and differential GPS
signals. Such a system can also be used in urban canyons and at all
locations where the GPS signals can be blocked or are otherwise not
available. Since the cost of radar reflectors is very low, it is
expected that eventually they will be widely distributed on roads
in the U.S.
The use of radar and reflectors for precise positioning is only one
of many systems being considered for this purpose. Others include
markings on roadway, RFID tags, laser systems, laser radar and
reflectors, magnetic tags embedded in the roadway, magnetic tape,
etc. The radar and reflector technology has advantages over some
systems in that it is not seriously degraded by bad weather
conditions, is not affected if covered with snow, does not pose a
serious maintenance problem, and other cost and durability
features. Any movement in the positioning of the reflectors can be
diagnosed from vehicle PPS-mounted systems.
The radar transceivers used are typically mounted on either side of
vehicle and pointed upward at between 30 and 60 degrees. They are
typically aimed so that they project across the top of the vehicle
so that several feet of vertical height can be achieved prior to
passing over adjacent lanes where the signal could be blocked by a
truck, for example. Other mounting and aiming systems can be
used.
The radar reflectors are typically mounted onto a pole, building,
overpass, or other convenient structure. They can provide a return
code by the placement of several such reflectors such that the
reflected pulse contains information that identifies this reflector
as a particular reflector on the map database. This can be
accomplished in numerous ways including the use of a collection of
radar reflectors in a spaced-apart geometric configuration on a
radius from the vehicle. The presence or absence of a reflector can
provide a returned binary code, for example.
The operation of the system is as follows. A vehicle traveling down
a roadway in the vicinity of the reflector poles would transmit
radar pulses at a frequency of perhaps once per microsecond. These
radar pulses would be encoded, perhaps with noise or code
modulation, so that each vehicle knows exactly what radar returns
are from its transmissions. As the vehicle approaches a reflector
pole, it will begin to receive reflections based on the speed of
the vehicle. By observing a series of reflections, the vehicle
software can select either the maximum amplitude reflection or the
average or some other scheme to determine the proper reflection to
consider. The radar pulse will also be modulated to permit a
distance to the reflector calculation to be made based on the phase
of the returned signal or through correlation. Thus, as a vehicle
travels down the road and passes a pair of reflector poles, it will
be able to determine its longitudinal position on the roadway based
on the pointing angle of the radar devices and the selected maximum
return as described above. It will also be able to determine its
lateral position on the roadway based on the measured distance from
the radar to the reflector.
Each reflector pole will have multiple reflectors determined by
intersections of the radar beam from the vehicle traveling in the
closest and furthest lanes. The spacing of reflectors on the pole
would be determined by the pixel diameter of the radar beam. For
example, a typical situation may require radar reflectors beginning
at 4 m from the ground and ending at 12 m with a reflector every
one-meter. For the initial demonstrations, it is expected that
existing structures will be used. The corner cube radar reflectors
are very inexpensive so therefore the infrastructure investment
will be small as long as existing structures can be used. In the
downtown areas of cities, buildings etc. can also be used as
reflector locations.
To summarize this aspect of the invention, an inexpensive
infrastructure installation concept is provided which will permit a
vehicle to send a radar pulse and receive a reflection wherein the
reflection is identifiable as the reflection from the vehicle's own
radar and contains information to permit an accurate distance
measurement. The vehicle can thus locate itself accurately
longitudinally and laterally along the road. A variation of the PPS
system using a signature from a continuously reflected laser or
radar has been discussed above and will not be repeated here.
FIG. 19 shows a variety of roads and vehicles operating on those
roads that are in communication with a vehicle that is passing
through a Precise Positioning Station (PPS). The communication
system used is based on noise modulated spread spectrum
technologies such as described in the above-listed papers by Lukin
et al.
FIG. 20 shows a schematic of the operation of a communication
and/or information system and method in accordance with the
invention. Transmitters are provided, for example at fixed
locations and/or in vehicles or other moving objects, and data
about each transmitter, such as its location and an identification
marker, is generated at 240. The location of the transmitter is
preferably its GPS coordinates as determined, for example, by a
GPS-based position determining system (although other position
determining systems can alternatively or additionally be used). The
data may include, when the transmitter is a moving vehicle, the
velocity, the direction of travel, the estimated travel path and
the destination of the vehicle. The data is encoded at 242 using
coding techniques such as those described above, e.g., phase
modulation of distance or time between code transmissions, phase or
amplitude modulation of the code sequences themselves, changes of
the polarity of the entire code sequence or the individual code
segments, or bandwidth modulation of the code sequence. The coded
data is transmitted at 244 using, e.g., noise or pseudo-noise
radar.
Instead of data about each transmitter being generated at 240,
general data for transmission could also be generated such as road
condition information or traffic information.
A vehicle 246 includes an antenna 248 coupled to a control module,
control unit, processor or computer 250. The antenna 248 receives
transmissions (waves) including transmissions 252 when in range of
the transmitters. The processor 250 analyzes the transmissions 252.
Such analysis may include a determination whether any transmissions
are from transmitters within a pre-determined area relative to the
vehicle, transmitters situated within a pre-determined distance
from the vehicle, from transmitters traveling in a direction toward
the vehicle's current position, transmitters traveling in a
direction toward the vehicle's projected position based on its
current position and velocity, the angle between the transmitter
and the vehicle, and any combinations of such determinations. Other
analyses could be whether any are from particular transmitters
which might be dedicated to the transmission of road conditions
data, traffic data, map data and the like. Once the processor 250
ascertains a particular transmission of interest (for operation of
the vehicle, or for any other pre-determined purpose), it extracts
the information coded in the transmission, but does not extract
information coded in transmission from transmitters which are not
of interest, e.g., those from a location outside of the
pre-determined area. It knows the code because the code is provided
by the transmission, i.e., the initial part of the transmission
252a contains data on the location of the transmitter and the code
is based on the location of the transmitter. As such, once the
initial part of the transmission is received and the location of
the transmitter extracted, the code for the remainder of the
transmission 252b can be obtained.
In this manner, the extraction of information from radio frequency
wave transmission is limited based on a threshold determination (a
filter of sorts) as to whether the transmission is of potential
interest, e.g., to the operation of the vehicle based on its
position. To enable this threshold determination from the analysis
of the waves or filtering of information, the initial part of the
transmission 252a can be provided with positional information about
the transmitter and information necessitated by the information
transferring arrangement (communication protocol data) and the
remainder of the transmission 252b provided with additional
information of potential interest for operation of the vehicle. The
information contained in initial part of each transmission (or set
of waves) is extracted to determine whether the information in the
final part of the transmission is of interest. If not, the
information in the final part of the transmission is not extracted.
This reduces processing time and avoids the unnecessary extraction
of mostly if not totally irrelevant information. An information
filter is therefore provided.
Further, the antenna 248 serves as a transmitter for transmitting
signals generated by the processor 250. The processor 248 is
constructed or programmed to generate transmissions or noise
signals based on its location, determined by a position determining
device 254 in any known manner including those disclosed herein,
and encode information about the vehicle in the signals. The
information may be an identification marker, the type of vehicle,
its direction, its velocity, its proposed course, its occupancy,
etc. The processor 248 can encode the information in the signals in
a variety of methods as disclosed above in the same manner that the
data about the transmitter is encoded. Thus, the processor 248 not
only interprets the signals and extracts information, it also is
designed to generate appropriate noise or otherwise coded signals
which are then sent from the antenna 248.
Consider the case where the automobile becomes a pseudolite or a
DGPS equivalent station since it has just determined its precise
location from the PPS. Thus the vehicle can broadcast just like a
pseudolite. As the vehicle leaves the PPS station, its knowledge of
its absolute position will degrade with time depending on the
accuracy of its clock and inertial guidance system and perhaps its
view of the satellites or other pseudolites. In some cases, it
might even be possible to eliminate the need for satellites if
sufficient PPS positions exist.
Another point is that the more vehicles that are in the vicinity of
a PPS, the higher the likelihood that one of the vehicles will know
precisely where it is by being at or close to the PPS and thus the
more accurately every vehicle in the vicinity would know its own
location. Thus, the more vehicles on the road, the accuracy with
which every vehicle knows its location increases. When only a
single vehicle is on the road, then it really doesn't need to know
its position nearly as accurately at least with regard to other
vehicles. It may still need to know its accuracy to a comparable
extent with regard to the road edges.
5. Radar and Laser Radar Detection and Identification of Objects
External to the Vehicle
5.1 Sensing of Non-RtZF.TM. Equipped Objects
Vehicles with the RtZF.TM. system in accordance with the invention
must also be able to detect those vehicles that do not have the
system as well as pedestrians, animals, bicyclists, and other
hazards that may cross the path of the equipped vehicle.
Systems based on radar have suffered from the problem of being able
to sufficiently resolve the images which are returned to be able to
identify the other vehicles, bridges, etc. except when they are
close to the host vehicle. One method used for adaptive cruise
control systems is to ignore everything that is not moving. This,
of course, leads to accidents if this were used with the instant
invention. The problem stems from the resolution achievable with
radar unless the antenna is made very large or the object is close.
Since this is impractical for use with automobiles, only minimal
collision avoidance can be obtained using radar.
Optical systems can provide the proper resolution but may require
illumination with a bright light or laser. If the laser is in the
optical range, there is a danger of causing eye damage to
pedestrians or vehicle operators. At a minimum, it will be
distracting and annoying to other vehicle operators. A laser
operating in the infrared part of the electromagnetic spectrum
avoids the eye danger problem, provided the frequency is
sufficiently far from the visible, and, since it will not be seen,
it will not be annoying. If the IR light is sufficiently intense to
provide effective illumination for the host vehicle, it might be a
source of blinding light for the system of another vehicle.
Therefore a method of synchronization may be required. This could
take the form of an Ethernet protocol, for example, where when one
vehicle detects a transmission from another then it backs off and
transmits at a random time later. The receiving electronics would
then only be active when the return signal is expected.
Another problem arises when multiple vehicles are present that
transmit infrared at the same time if there is a desire to obtain
distance information from the scene. In this case, each vehicle
needs to be able to recognize its transmission and not be fooled by
transmissions from another vehicle. This can be accomplished, as
discussed above, through the modulation scheme. Several such
schemes would suffice with a pseudo-noise or code modulation as the
preferred method for the present invention. This can also be
accomplished if each vehicle accurately knows its position and
controls its time of transmission according to an algorithm that
time multiplexes transmissions based on the geographical location
of the vehicle. Thus if multiple vehicles are sensed in a given
geographical area, they each can control their transmissions based
on a common algorithm that uses the GPS coordinates of the vehicle
to set the time slot for transmission so as to minimize
interference between transmissions from different vehicles. Other
multiplexing methods can also be used such as FDMA, CDMA or TDMA,
any of which can be based on the geographical location of the
vehicles.
Infrared and terahertz also have sufficient resolution so that
pattern recognition technologies can be employed to recognize
various objects, such as vehicles, in the reflected image as
discussed above. Infrared has another advantage from the object
recognition perspective. All objects radiate and reflect infrared.
The hot engine or tires of a moving vehicle in particular are
recognizable signals. Thus, if the area around a vehicle is
observed with both passive and active infrared, more information
can be obtained than from radar, for example. Infrared is less
attenuated by fog than optical frequencies, although it is not as
good as radar. Infrared is also attenuated by snow but at the
proper frequencies it has about five times the range of human
sight. Terahertz under some situations has an effective range of as
much as several hundred times that of human sight. Note, as with
radar, infrared and terahertz can be modulated with noise,
pseudonoise, or other distinctive signal to permit the separation
of various reflected signals from different transmitting
vehicles.
An example of such an instrument is made by Sumitomo Electric and
is sufficient for the purpose here. The Sumitomo product has been
demonstrated to detect leaves of a tree at a distance of about 300
meters. The product operates at a 1.5 micron wavelength.
This brings up a philosophical discussion about the trade-offs
between radar with greater range and infrared laser radar, or
lidar, with more limited range but greater resolution. At what
point should driving during bad weather conditions be prohibited?
If the goal of Zero Fatalities.TM. is to be realized, then people
should not be permitted to operate their vehicles during dangerous
weather conditions. This may require closing roads and highways
prior to the start of such conditions. Under such a policy, a
system which accurately returns images of obstacles on the roadway
that are two to five times the visual distance should be adequate.
In such a case, radar would not be necessary.
5.2 Laser and Terahertz Radar Scanning System
Referring to FIG. 25, a digital map (116) can be provided and when
the vehicle's position is determined (118), e.g., by a GPS-based
system, the digital map can be used to define the field (122) that
the laser or terahertz radar scanner (102) will interrogate. Note,
when the term scanner is used herein, it is not meant to imply that
the beam is so narrow as to require a back and forth motion (a
scan) in order to completely illuminate an object of interest. To
the contrary, the inventions herein are not limited to a particular
beam diameter other than that required for eye safety. Also a
scanner may be limited to an angular motion that just covers a
vehicle located 100 meters, for example, from the transmitting
vehicle, which may involve no angular motion of the scanner at all,
or to an angular motion that covers 90 or more degrees of the space
surrounding the transmitting vehicle. Through the use of
high-powered lasers and appropriate optics, an eye safe laser beam
can be created that is 5 cm in diameter, for example, with a
divergence angle less than one degree. Such an infrared spotlight
requires very little angular motion to illuminate a vehicle at 100
meters, for example.
Generally herein, when laser radar, or lidar, is used it will also
mean a system based on terahertz where appropriate. The laser radar
or lidar scanner will return information as to distance to an
object in the scanned field, e.g., laser beam reflections will be
indicative of presence of object in path of laser beam (104) and
from these reflections, information such as the distance between
the vehicle and the object can be obtained. This will cover all
objects that are on or adjacent to the highway. The laser pulse can
be a pixel that is two centimeters or 1 meter in diameter at 50
meters, for example and that pixel diameter can be controlled by
the appropriate optical system that can include adaptive optics and
liquid lenses (such as described in "Liquid lens promises cheap
gadget optics", NewScientist.com news service, Mar. 8, 2004). The
scanner must scan the entire road at such a speed that the motion
of the car can be considered insignificant. Alternately, a separate
aiming system that operates at a much lower speed, but at a speed
to permit compensation for the car angle changes, may be provided.
Such an aiming system is also necessary due to the fact that the
road curves up and down. Therefore two scanning methods, one a
slow, but for large angle motion and the other fast but for small
angles may be required. The large angular system requires a motor
drive while the small angular system can be accomplished through
the use of an acoustic wave system, such as Lithium Niobate
(LiNbO.sub.3), which is used to drive a crystal which has a large
refractive index such as Tellurium dioxide. Other acoustic optical
systems can also be used as scanners.
For these systems, frequently some means is needed to stabilize the
image and to isolate it from vehicle vibrations. Several such
stabilization systems have been used in the past and would be
applicable here including a gyroscopic system that basically
isolates the imaging system from such vibrations and keeps it
properly pointed, a piezoelectric system that performs similarly,
or the process can be accomplished in software where the image is
collected regardless of the vibration but where the image covers a
wider field of view then is necessary and software is used to
select the region of interest.
Alternately, two systems can be used, a radar system for
interrogating large areas and a laser radar for imaging smaller
areas. Either or both systems can be range gated and noise or
pseudonoise modulated.
The laser radar scanner can be set up in conjunction with a range
gate (106) so that once it finds an object, the range can be
narrowed so that only that object and other objects at the same
range, 65 to 75 feet for example, are allowed to pass to the
receiver. In this way, an image of a vehicle can be separated from
the rest of the scene for identification by pattern recognition
software (108). Once the image of the particular object has been
captured, the range gate is broadened, to about 20 to 500 feet for
example, and the process repeated for another object. In this
manner, all objects in the field of interest to the vehicle can be
separated and individually imaged and identified. Alternately, a
scheme based on velocity can be used to separate a part of one
object from the background or from other objects. The field of
interest, of course, is the field where all objects with which the
vehicle can potentially collide reside. Particular known and mapped
features on the highway can be used as aids to the scanning system
so that the pitch and perhaps roll angles of the vehicle can be
taken into account. Once the identity of the object is known, the
potential for a collision between the vehicle and that object
and/or consequences of a potential collision with that object are
assessed, e.g., by a control module, control unit or processor
(112). If collision is deemed likely, countermeasures are effected
(114), e.g., activation of a driver alert system and/or activation
of a vehicle control system to alter the travel of the vehicle (as
discussed elsewhere herein).
Range gates can be achieved as high speed shutters by a number of
devices such as liquid crystals, garnet films, Kerr and Pockel
cells or as preferred herein as described in patents and patent
applications of 3DV Systems Ltd., Yokneam, Israel including U.S.
Pat. No. 6,327,073, U.S. Pat. No. 6,483,094, U.S. 2002/0185590,
WO98/39790, WO97/01111, WO97/01112 and WO97/01113.
Prior to the time that all vehicles are equipped with the RtZF.TM.
system described above, roadways will consist of a mix of vehicles.
In this period, it will not be possible to totally eliminate
accidents. It will be possible to minimize the probability of
having an accident however, if a laser radar or Lidar system
similar to that described in Shaw (U.S. Pat. No. 5,529,138), or
more recently in various patents and patent applications of Ford
Global Technologies such as U.S. Pat. No. 6,690,017, U.S. Pat. No.
6,730,913, U.S. 2003/0034462, U.S. 2003/0155513 and U.S.
2003/0036881, with some significant modifications is used. It is
correctly perceived by Shaw that the dimensions of a radar beam are
too large to permit distinguishing various objects which may be on
the roadway in the path of the instant vehicle. Laser radar
provides the necessary resolution that is not provided by radar.
Laser radar as used in the present invention however would acquire
significantly more data than anticipated by Shaw. Sufficient data
in fact would be attained to permit the acquisition of a
three-dimensional image of all objects in the field of view. The X
and Y dimensions of such objects would, of course, be determined
knowing the angular orientation of the laser radar beam. The
longitudinal or Z dimension can be obtained by such methods as
time-of-flight of the laser beam to a particular point on the
object and reflected back to the detector, by phase methods or by
range gating. All such methods are described elsewhere herein and
in the patents listed above.
At least two methods are available for resolving the longitudinal
dimension for each of the pixels in the image. In one method, a
laser radar pulse having a pulse width of one to ten nanoseconds,
for example, can be transmitted toward the area of interest and as
soon as the reflection is received and the time-of-flight
determined, a new pulse would be sent at a slightly different
angular orientation. The laser, therefore, would be acting as a
scanner covering the field of interest. A single detector could
then be used, if the pixel is sufficiently small, since it would
know which pixel was being illuminated. The distance to the
reflection point could be determined by time-of-flight thus giving
the longitudinal distance to all points in view on the object.
Alternately, the entire area of interest can be illuminated and an
image focused on a CCD or CMOS array. By checking the
time-of-flight to each pixel, one at a time, the distance to that
point on the vehicle would be determined. A variation of this would
be to use a garnet crystal as a pixel shutter and only a single
detector. In this case, the garnet crystal would permit the
illumination to pass through one pixel at a time through to a
detector. A preferred method, however, for this invention is to use
range gating as described elsewhere herein.
Other methods of associating a distance to a particular reflection
point, of course, can now be performed by those skilled in the art
including variations of the above ideas using a pixel mixing device
(such as described in Schwarte, R. "A New Powerful Sensory Tool in
Automotive Safety Systems Based on PMD-Technology", S-TEC GmbH
Proceedings of the AMAA 2000) or variations in pixel illumination
and shutter open time to determine distance through comparison of
range gated received reflected light. In the laser scanning cases,
the total power required from the laser is significantly less than
in the area illumination design. However, the ability to correctly
change the direction of the laser beam in a sufficiently short
period of time complicates the scanning design. The system can work
approximately as follows: The entire area in front of the instant
vehicle, perhaps as much as a full 180 degree arc in the horizontal
plane can be scanned for objects using either radar or laser radar.
Once one or more objects had been located, the scanning range can
be severely limited to basically cover that particular object and
some surrounding space using laser radar. Based on the range to
that object, a range gate can be used to eliminate all background
and perhaps interference from other objects. In this manner, a very
clear picture or image of the object of interest can be obtained as
well as its location and, through the use of a neural network,
combination neural network or optical correlation or other pattern
of recognition system, the identity of the object can be
ascertained as to whether it is a sign, a truck, an animal, a
person, an automobile or other object. The identification of the
object will permit an estimate to be made of the object's mass and
thus the severity of any potential collision.
Once a pending collision is identified, this information can be
made available to the driver and if the driver ceases to heed the
warning, control of the vehicle could be taken from him or her by
the system. The actual usurpation of vehicle control, however, is
unlikely initially since there are many situations on the highway
where the potential for a collision cannot be accurately
ascertained. Consequently, this system can be thought of as an
interim solution until all vehicles have the RtZF.TM. system
described above.
To use the laser radar in a scanning mode requires some mechanism
for changing the direction of the emitted pulses of light. One
acoustic-optic method of using an ultrasonic wave to change the
diffraction angle of a Tellurium dioxide crystal is disclosed
elsewhere herein. This can also be done in a variety of other ways
such as through the use of a spinning multifaceted mirror, such as
is common with laser scanners and printers. This mirror would
control the horizontal scanning, for example, with the vertical
scanning controlled though a stepping motor or the angles of the
different facets of the mirror can be different to slightly alter
the direction of the scan, or by other methods known in the art.
Alternately, one or more piezoelectric materials can be used to
cause the laser radar transmitter to rotate about a pivot point. A
rotating laser system, such as described in Shaw is the least
desirable of the available methods due to the difficulty in
obtaining a good electrical connection between the laser and the
vehicle while the laser is spinning at a very high angular
velocity. Another promising technology is to use MEMS mirrors to
deflect the laser beam in one or two dimensions.
Although the system described above is intended for collision
avoidance or at least the notification of a potential collision,
when the roadway is populated by vehicles having the RtZF.TM.
system and vehicles which do not, its use is still desirable after
all vehicles are properly equipped. It can be used to search for
animals or other objects which may be on or crossing the highway, a
box dropping off of a truck for example, a person crossing the road
who is not paying attention to traffic. Motorcycles, bicycles, and
other non-RtZF equipped vehicles can also be monitored.
One significant problem with all previous collision avoidance
systems which use radar or laser radar systems to predict impacts
with vehicles, is the inability to know whether the vehicle that is
being interrogated is located on the highway or is off the road. In
at least one system of the present invention, the location of the
road at any distance ahead of the vehicle would be known precisely
from the sub-meter accuracy maps, so that the scanning system can
ignore, for example, all vehicles on lanes where there is a
physical barrier separating the lanes from the lane on which the
subject vehicle is traveling. This, of course, is a common
situation on super highways. Similarly, a parked vehicle on the
side of the road would not be confused with a stopped vehicle that
is in the lane of travel of the subject vehicle when the road is
curving. This permits the subject invention to be used for
automatic cruise control. In contrast with radar systems, it does
not require that vehicles in the path of the subject vehicle be
moving, so that high speed impacts into stalled traffic can be
avoided.
If a system with a broader beam to illuminate a larger area on the
road in front of the subject vehicle is used, with the subsequent
focusing of this image onto a CCD or CMOS array, this has an
advantage of permitting a comparison of the passive infrared signal
and the reflection of the laser radar active infrared. Metal
objects, for example appear cold to passive infrared. This permits
another parameter to be used to differentiate metallic objects from
non-metallic objects such as foliage or animals such as deer. The
breadth of the beam can be controlled and thereby a particular
object can be accurately illuminated. With this system, the speed
with which the beam steering is accomplished can be much slower.
Both systems can be combined into the maximum amount of information
to be available to the system.
Through the use of range gating, objects can be relatively isolated
from the environment surrounding it other than for the section of
highway which is at the same distance. For many cases, a properly
trained neural network or other pattern recognition system can use
this data and identify the objects. An alternate approach is to use
the Fourier transform of the scene as input to the neural network
or other pattern recognition system. The advantages of this latter
approach are that the particular location of the vehicle in the
image is not critical for identification. Note that the Fourier
transform can be accomplished optically and optically compared with
stored transforms using a garnet crystal or garnet films, for
example, as disclosed in U.S. Pat. No. 5,473,466.
At such time when the system can take control of the vehicle, it
will be possible to have much higher speed travel. In such cases,
all vehicles on the controlled roadway will need to have the
RtZF.TM. or similar system as described above. Fourier transforms
of the objects of interest can be done optically though the use of
a diffraction system. The Fourier transform of the scene can then
be compared with the library of the Fourier transforms of all
potential objects and, through a system used in military target
recognition, multiple objects can be recognized and the system then
focused onto one object at time to determine the degree of threat
that it poses.
Of particular importance is the use of a high powered laser radar
such as a 30 to 100 watt laser diode in an expanded beam form to
penetrate fog, rain and snow through the use of range gating. If a
several centimeter diameter beam is projected from the vehicle in
the form of pulses of from 1 to 10 nanoseconds long, for example,
and the reflected radiation is blocked except that from the region
of interest, an image can still be captured even though it cannot
be seen by the human eye. This technique significantly expands the
interrogation range of the system and, when coupled with the other
imaging advantages of laser radar, offers a competitive system to
radar and may in fact render the automotive use or radar
unnecessary. One method is to use the techniques described in the
patents to 3DV listed above. In one case, for example, if the
vehicle wishes to interrogate an area 250 feet ahead, a 10
nanosecond square wave signal can be used to control the shutter
which is used both for transmission and reception and where the off
period can be 480 nanoseconds. This can be repeated until
sufficient energy has been accumulated to provide for a good image.
In this connection, a high dynamic range camera may be used such as
that manufactured by IMS chips of Stuttgart, Germany as mentioned
above. Such a camera is now available with a dynamic range of 160
db.
These advantages are also enhanced when the laser radar system
described herein is used along with the other features of the
RtZF.TM. system such as accurate maps and accurate location
determination. The forward looking laser radar system can thus
concentrate its attention to the known position of the roadway
ahead rather than on areas where there can be no hazardous
obstacles or threatening vehicles.
5.3 Blind Spot Detection
The RtZF.TM. system of this invention also can eliminate the need
for blind spot detectors such as discussed in U.S. Pat. No.
5,530,447 to Henderson. Alternately, if a subset of the complete
RtZF.TM. system is implemented, as is expected in the initial
period, the RtZF.TM. system can be made compatible with the blind
spot detector described in the '447 patent.
One preferred implementation for blind spot monitoring as well as
for monitoring other areas near the vehicle is the use of range
gated laser radar using a high power laser diode and appropriate
optics to expand the laser beam to the point where the transmitted
infrared energy per square millimeter is below eye safety limits.
Such a system is described above.
5.4 Anticipatory Sensing--Smart Airbags, Evolution of the
System
A key to anticipating accidents is to be able to recognize and
categorize objects that are about to impact a vehicle as well as
their relative velocity. As set forth herein and in current
assignee's patents and patent applications referenced above, this
can best be done using a pattern recognition system such as a
neural network, combination neural network, optical correlation
system, sensor fusion and related technologies. The data for such a
pattern recognition system can be derived from a camera image but
such an image can be overwhelmed by reflected light from the sun.
In fact, lighting variations in general plague camera-based images
resulting in false classifications or even no classification.
Additionally, camera-based systems are defeated by poor visibility
conditions and, additionally, have interference problems when
multiple vehicles have the same system which may require a
synchronization, taking time away from the critical anticipatory
sensing function.
To solve these problems imaging systems based of millimeter wave
radar, laser radar (lidar) and more recently terahertz radar can be
used. All three systems generally work for anticipatory sensors
since the objects are near the vehicle where even infrared scanning
laser radar in a non-range gated mode has sufficient range in fog.
Millimeter wave radar is expensive and to obtain precise images, a
narrow beam is required resulting in large scanning antennas. Laser
radar systems are less expensive and since the beams are formed
using optic technology, they are smaller and easier to
manipulate.
When computational power is limited, it is desirable to determine
the minimum number of pixels that are required to identify an
approaching object with sufficient accuracy to make the decision to
take evasive action or to deploy a passive restraint such as an
airbag. In one military study for anti-tank missiles, it was found
that a total of 25 pixels are all that is required to identify a
tank on a battlefield. For optical occupant detection within a
vehicle, thousands of pixels are typically used. Experiments
indicate that by limiting the number of horizontal scans to three
to five, with on the order of 100 to 300 pixels per scan that
sufficient information is available to find an object near to the
vehicle and in most cases to identify the object. Once the object
has been located, then the scan can be confined to the position of
the object and the number of pixels available for analysis
substantially increases. There are obviously many algorithms that
can be developed and applied to this problem and it is therefore
left to those skilled in the art. At least one invention herein is
based on the fact that a reasonable number of pixels can be
obtained from the reflections of electromagnetic energy from an
object to render each of the proposed systems practical for
locating, identifying and determining the relative velocity of an
object in the vicinity of a vehicle that poses a threat to impact
the vehicle so that evasive action can be taken or a passive
restraint deployed. See the discussion in section 5.5 below for a
preferred implementation.
The RtZF.TM. system is also capable of enhancing other vehicle
safety systems. In particular, by knowing the location and velocity
of other vehicles, for those cases where an accident cannot be
avoided, the RtZF.TM. system will in general be able to anticipate
a crash and make an assessment of the crash severity using, for
example, neural network technology. Even with a limited
implementation of the RtZF.TM. system, a significant improvement in
smart airbag technology results when used in conjunction with a
collision avoidance system such as described in Shaw (U.S. Pat. No.
5,314,037 and U.S. Pat. No. 5,529,138) and a neural network
anticipatory sensing algorithm such as disclosed in U.S. Pat. No.
6,343,810 to Breed. A further enhancement would be to code a
vehicle-to-vehicle communication signal from RtZF.TM.
system-equipped vehicles with information that includes the size
and approximate weight of the vehicle. Then, if an accident is
inevitable, the severity can also be accurately anticipated and the
smart airbag tailored to the pending event. Information on the size
and mass of a vehicle can also be implemented as an RFID tag and
made part of the license plate.
Recent developments by Mobileye (www.mobileye.com) illustrate a
method of obtaining the distance to an object and thus the relative
velocity. Although this technique has many limitations, it may be
useful in some implementations of one or more of the current
inventions.
A further recent development is reported in U.S. patent application
publication No. 20030154010, as well as other patents and patent
publications assigned to Ford Global Technologies including U.S.
Pat. Nos. 06,452,535, 06,480,144, 06,498,972, 06,650,983,
06,568,754, 06,628,227, 06,650,984, 06,728,617, 06,757,611,
06,775,605, 06,801,843, 06,819,991, 20030060980, 20030060956,
20030100982, 20030154011, 20040019420, 20040093141, 20040107033,
20040111200, and 20040117091. In the disclosures herein, emphasis
has been placed on identifying a potentially threatening object and
once identified, the properties of the object such as its size and
mass can be determined. An inferior system can be developed as
described in U.S. patent application publication No. 20030154010
where only the size is determined. In the inventions described
herein, the size is inherently determined during the process of
imaging the object and identifying it. Also, the Ford patent
publications mention the combined use of a radar or a lidar and a
camera system. The combined use of radar and a camera are of course
anticipated herein and in assignee's patents cross-referenced
above.
Another recent development by the U.S. Air Force uses a high
powered infrared laser operating at wavelengths greater than 1.5
microns and a focal plane array as is reported in
"Three-Dimensional Imaging" in AFRL Technology Horizons, April
2004. Such a system is probably too expensive at this time for
automotive applications. This development illustrates the fact that
it is not necessary to limit the lidar to the near infrared part of
the spectrum and in fact, the further that the wavelength is away
from the visible spectrum, the higher the power permitted to be
transmitted. Also, nothing prevents the use of multiple frequencies
as another method of providing isolation from transmissions from
vehicles in the vicinity. As mentioned above for timing
transmissions, the GPS system can also be used to control the
frequency of transmission thus using frequency as a method to
prevent interference. The use of polarizing filters to transmit
polarized infrared is another method to provide isolation between
different vehicles with the same or similar systems. The
polarization angle can be a function of the GPS location of the
vehicle.
It is the intention of some of the inventions herein to provide a
system that can be used both in daytime and at night. Other systems
are intended solely for night vision such as those disclosed in
U.S. Pat. No. 6,730,913, U.S. Pat. No. 6,690,017 and U.S. Pat. No.
6,725,139. Note that the use of the direction of travel as a method
of determining when to transmit infrared radiation, as disclosed in
these and other Ford Global patents and patent applications, can be
useful but it fails to solve the problem of the transmissions from
two vehicles traveling in the same vicinity and direction from
receiving reflections from each others' transmissions. If the
directional approach is used, then some other method is required
such as coding the pulses, for example.
U.S. Pat. No. 6,730,913 and U.S. Pat. No. 6,774,367 are
representative of a series of patents awarded to Ford Global
Technologies as discussed above. This patent uses range gating as
taught by assignee's earlier patents. The intent is to supplement
the headlights with a night vision system for illuminating objects
on the roadway in the path of the vehicle but are not seen by the
driver and displaying these objects in a heads up display. No
attempt is made to locate the eyes of the driver and therefore the
display cannot place the objects where they would normally be
located in the driver's field of view as taught in the current
assignee's other patents. Experiments have shown that without this
feature, the night vision system is of little value and may even
distract the driver to where his or her ability to operate the
motor vehicle is degraded. Other differences in the '913 and '367
system is an attempt to compensate for falloff in illumination due
to distance, neglecting a similar and potentially more serious
falloff due to scattering due to fog etc. In at least one of the
inventions disclosed herein, no attempt is made to achieve this
compensation in a systematic manner but rather the exposure is
adjusted so that a sufficiently bright image is achieved to permit
object identification regardless of the cause of the attenuation.
Furthermore, in at least one embodiment, a high dynamic range
camera is used which automatically compensates for much of the
attenuation and thus permits the minimum exposure requirements for
achieving an adequate image. In at least one of the inventions
disclosed herein, the system is used both at night and in the
daytime of locating and identifying objects and, in some cases,
initiating an alarm or even taking control of the vehicle to avoid
accidents. None of these objects are disclosed in the '913 or '367
and related patents. Additionally, U.S. 20030155513, also part of
this series of Ford Global patents and applications, describes
increasing the illumination intensity based on distance to the
desired field of view. In at least one of the inventions disclosed
herein, the illumination intensity is limited by eye safety
considerations rather than distance to the object of interest. If
insufficient illumination is not available on one pulse, additional
pulses are provided until sufficient illumination to achieve an
adequate exposure is achieved.
If the laser beam diverges, then the amount of radiation per square
centimeter illuminating a surface will be a function of the
distance of that surface from the transmitter. If that distance can
be measured, then the transmitted power can be increased while
keeping the radiation per square centimeter below the eye safe
limits. Using this technique, the amount of radiated power can be
greatly increased thus enhancing the range of the system in
daylight and in bad weather. A lower power pulse would precede a
high power pulse transmitted in a given direction and the distance
measured to a reflective object would be measured and the
transmitted power adjusted appropriately. If a human begins to
intersect the path of transmission, the distance to the human would
be measured before he or she could put his or her eye into the
transmission path and the power can be reduced to remain within the
safety standards.
It is also important to point out that the inventions disclosed
herein that use lidar (laser radar or ladar) can be used in a
scanning mode when the area to be covered is larger that the beam
diameter or in a pointing mode when the beam diameter is sufficient
to illuminate the target of interest, or a combination thereof.
It can be seen from the above discussion that the RtZF.TM. system
will evolve to solve many safety, vehicle control and ITS problems.
Even such technologies as steering and drive by wire will be
enhanced by the RtZF.TM. system in accordance with invention since
it will automatically adjust for failures in these systems and
prevent accidents.
5.5 A preferred Implementation
FIGS. 21A and 21B illustrate a preferred embodiment of a laser
radar system having components mounted at the four corners of a
vehicle above the headlights and tail lights. Laser radar units or
assemblies 260 and 261 have a scan angle of approximately 150
degrees; however, for some applications a larger or smaller
scanning angle can of course be used. The divergence angle for the
beam for one application can be one degree or less when it is
desired to illuminate an object at a considerable distance from the
vehicle such as from less than fifty meters to 200 meters or more.
In other cases, where objects are to be illuminated that are closer
to the vehicle, a larger divergence angle can be used. Generally,
it is desirable to have a field of illumination (FOI) approximately
equal to the field of view (FOV) of the camera or other optical
receiver. FIGS. 22A and 22B illustrate the system of FIGS. 21A and
21B for vehicles on a roadway. Note that the divergence angle in
the horizontal plane and vertical plane are not necessarily
equal.
FIGS. 23A and 23B illustrate an alternative mounting location for
laser radar units on or near the roof of a vehicle. They can be
either inside or outside of the vehicle compartment. The particular
design of the laser radar assemblies 262 and 263 are similar to
those used in FIGS. 21A, 21B, 22A and 22B. Although not shown,
other geometries are of course possible such as having the laser
radar assemblies mounted on or near the roof for the rear
assemblies and above the headlights for the frontal assemblies or
vice versa. Also, although assemblies mounted on the corners of the
vehicle are illustrated, in some cases it may be desirable to mount
laser radar assemblies in the center of the front, back and sides
of the vehicle or a combination or center and corner-mounted laser
radar assemblies can be used.
FIG. 24 is a schematic illustration of a typical laser radar
assembly showing the scanning or pointing system with simplified
optics for illustration only. In an actual design, the optics will
typically include multiple lenses. Also, the focal point will
typically not be outside of the laser radar assembly. In this
non-limiting example, a common optical system 267 is used to
control a laser light 265 and an imager or camera 266. The laser
source transmits, usually infrared, light through its optical
sub-system 271 which collimates the radiation. The collimated
radiation is then reflected off mirror 273 to mirror 274 which
reflects the radiation to the desired direction through lens system
267. The direction of the beam is controlled by motor 272 which
rotates both mirror 274 and optical system 267 to achieve the
desired scanning or pointing angle. The radiation leaves the
optical system 267 and illuminates the desired object or target
276. The radiation reflected from object 276 can pass back through
lens 267, reflect off mirror 274, pass through semitransparent
mirror 273 through optic subsystem 268 and onto optical sensitive
surface 266. Many other configurations are possible. The
transmission of the radiation is controlled by optical shutter 270
via controller 275. Similarly, the light that reaches the imager
266 is controlled by controller 275 and optical shutter 269. These
optical shutters 269, 270 can be liquid crystal devices, Kerr or
Pockel cells, garnet films, other spatial light monitors or,
preferably, high speed optical shutters such as described in
patents and patent applications of the 3DV Systems Ltd., of
Yokneam, Israel, as set forth above or equivalent. Since much of
the technology used in this invention related to the camera and
shutter system is disclosed in the 3DV patents and patent
applications, it will not be repeated here, by is incorporated by
reference herein.
In some embodiments, it may be important to assure that the lens
through which the laser radar radiation passes is clean. As a
minimum, a diagnostic system is required to inform the RtZF.TM. or
other system that the lens are soiled and therefore the laser radar
system can not be relied upon. Additionally, in some applications,
means are provided to clean one or more of the lens or to remove
the soiled surface. In the latter case, a roll of thin film can be
provided which, upon the detection of a spoiled lens, rolls up a
portion of the film and thereby provides a new clean surface. When
the roll is used up it can be replaced. Other systems provide one
or more cleaning methods such as a small wiper or the laser radar
unit can move the lens into a cleaning station. Many other methods
are of course possible and the invention here is basically
concerned with ascertaining that the lens is clean and if not
informing the system of this fact and, in some cases, cleaning or
removing the soiled surface.
Note that although laser radar and radar have been discussed
separately, in some implementations, it is desirable to use both a
radar system and a laser radar system. Such a case can be where the
laser radar system is not capable to achieve sufficient range in
adverse weather whereas the radar has the requisite range but
insufficient resolution. The radar unit can provide a warning that
a potentially dangerous situation exists and thus the vehicle speed
should be reduced until the laser radar device and obtain an image
with sufficient resolution to permit an assessment of the extent of
the danger and determine whether appropriate actions should be
undertaken.
5.6 Antennas
When the interrogation system makes use of radar such as systems in
use at 24 GHz and 77 GHz, a key design issue is the antenna. The
inventions herein contemplate the use of various types of antennas
such as dipole and monopole designs, yagi, steerable designs such
as solid state phased array and so called smart antennas. All
combinations of antennas for radar surveillance around a vehicle
are within the scope if the inventions disclosed herein. In
particular, the Rotman lens offers significant advantages as
disclosed in L. Hall, H. Hansen and D. Abbott "Rotman lens for
mm-wavelengths", Smart Structures, Devices, and Systems, SPIE Vol.
4935 (2002). Other antenna designs can be applicable. In some
cases, one radar source can be used with multiple antennas.
6. Smart Highways
The theme of the inventions disclosed herein is that automobile
accidents can be eliminated and congestion substantially mitigated
through the implementation thereof. After sufficient
implementations have occurred, the concept of a smart highway
becomes feasible. When a significant number of vehicles have the
capability of operating in a semi-autonomous manner, then dedicated
highway lanes (like the HOV lanes now in use) can be established
where use of the lanes is restricted to properly equipped vehicles.
Vehicles operating on these lanes can travel in close packed high
speed formations since each of them will know the location of the
road, their location on the road and the location of every other
vehicle in such a lane. Accidents on these lanes will not occur and
the maximum utilization of the roadway infrastructure will have
been obtained. Vehicle owners will be highly motivated to own
equipped vehicles since their travel times will be significantly
reduced and while traveling on such lanes, control of the vehicle
can be accomplished by the system and they are then free to talk on
the telephone, read or whatever.
7. Weather and Road Condition Monitoring
The monitoring of the weather conditions and the control of the
vehicle consistent with those conditions has been discussed herein.
The monitoring of the road conditions and in particular icing has
also been discussed elsewhere herein and in other patents and
patent applications of the current assignee. Briefly, a vehicle
will be controlled so as to eliminate accidents under all weather
and road conditions. This in some cases will mean that the vehicle
velocity will be controlled and, in some cases, travel will be
prohibited until conditions improve.
8. Communication with Other Vehicles--Collision Avoidance
8.1 Requirements
MIR might also be used for vehicle-to-vehicle communication except
that it is line of sight. An advantage is that we can know when a
particular vehicle will respond by range gating. Also, the short
time of transmission permits many vehicles to communicate at the
same time. The preferred system is to use spread spectrum
carrier-less coded channels.
One problem which will require addressing as the system becomes
mature is temporary blockage of a satellite by large trucks or
other movable objects whose location cannot be foreseen by the
system designers. Another concern is to prevent vehicle owners from
placing items on the vehicle exterior that block the GPS and
communication antennas.
The first problem can be resolved if the host vehicle can
communicate with the blocking trucks and can also determine its
relative location, perhaps through using the vehicle exterior
monitoring system. Then the communication link will provide the
location of the adjacent truck and the monitoring system will
provide the relative location and thus the absolute location of the
host vehicle can be determined.
The communication between vehicles for collision avoidance purposes
cannot solely be based on line-of-sight technologies as this is not
sufficient since vehicles which are out of sight can still cause
accidents. On the other hand, vehicles that are a mile away from
one another but still in sight, need not be part of the
communication system for collision avoidance purposes. Messages
sent by each vehicle, in accordance with an embodiment of the
invention, can contain information indicating exactly where it is
located and perhaps information as to what type of vehicle it is.
The type of vehicle can include emergency vehicles, construction
vehicles, trucks classified by size and weight, automobiles, and
oversized vehicles. The subject vehicle can therefore eliminate all
vehicles that are not potential threats, even if such vehicles are
very close, but on the other side of the highway barrier.
The use of a wireless Ethernet protocol can satisfy the needs of
the network, consisting of all threatening vehicles in the vicinity
of the subject vehicle. Alternately, a network where the subject
vehicle transmits a message to a particular vehicle and waits for a
response could be used. From the response time, assuming that the
clocks of both vehicles are or can be synchronized, the relative
position of other vehicles can be ascertained which provides one
more method of position determination. Thus, the more vehicles that
are on the road with the equipped system, the greater accuracy of
the overall system and the safer the system becomes.
To prevent accidents caused by a vehicle leaving the road surface
and impacting a roadside obstacle requires only an accurate
knowledge of the position of the vehicle and the road boundaries.
To prevent collisions with other vehicles requires that the
position of all nearby automobiles must be updated continuously.
Just knowing the position of a threatening vehicle is insufficient.
The velocity, size and/or orientation of the vehicle are also
important in determining what defensive action or reaction may be
required. Once all vehicles are equipped with the system of this
invention, the communication of all relevant information will take
place via a communication link, e.g., a radio link. In addition to
signaling its absolute position, each vehicle will send a message
identifying the approximate mass, velocity, orientation, and/or
other relevant information. This has the added benefit that
emergency vehicles can make themselves known to all vehicles in
their vicinity and all such vehicles can then take appropriate
action to allow passage of the emergency vehicle. The same system
can also be used to relay accident or other hazard information from
vehicle to vehicle through an ad-hoc or mesh network.
8.2 A Preferred System
One preferred method of communication between vehicles uses that
portion of the electromagnetic spectrum that permits only line of
sight communication. In this manner, only those vehicles that are
in view can communicate. In most cases, a collision can only occur
between vehicles that can see each other. This system has the
advantage that the "communications network" only contains nearby
vehicles. This would require that when a truck, for example, blocks
another stalled vehicle that the information from the stalled
vehicle be transmitted via the truck to a following vehicle. An
improvement in this system would use a rotating aperture that would
only allow communication from a limited angle at a time further
reducing the chance for multiple messages to interfere with each
other. Each vehicle transmits at all angles but receives at only
one angle at a time. This has the additional advantage of
confirming at least the direction of the transmitting vehicle. An
infrared rotating receiver can be looked at as similar to the human
eye. That is, it is sensitive to radiation from a range of
directions and then focuses in on the particular direction, one at
a time, from which the radiation is coming. It does not have to
scan continuously. In fact, the same transmitter which transmits
360 degrees could also receive from 360 degrees with the scanning
accomplished using software.
An alternate preferred method is to use short distance radio
communication so that a vehicle can receive position information
from all nearby vehicles such as the DS/SS system. The location
information received from each vehicle can then be used to
eliminate it from further monitoring if it is found to be on a
different roadway or not in a potential path of the subject
vehicle.
Many communications schemes have been proposed for inter-vehicle
and vehicle-to-road communication. At this time, a suggested
approach utilizes DS/SS communications in the 2.4 GHz INS band.
Experiments have shown that communications are 100 percent accurate
at distances up to 200 meters. At a closing velocity of 200 KPH, at
0.5 g deceleration, it requires 30 meters for a vehicle to stop.
Thus, communications accurate to 200 meters is sufficient to cover
all vehicles that are threatening to a particular vehicle.
A related method would be to use a MIR system in a communications
mode. Since the width of the pulses typically used by MIR is less
than a nanosecond, many vehicles can transmit simultaneously
without fear of interference. Other spread spectrum methods based
on ultra wideband or noise radar are also applicable. In
particular, as discussed below, a communication system based on
correlation of pseudorandom or other codes is preferred.
With either system, other than the MIR system, the potential exists
that more than one vehicle will attempt to send a communication at
the same time and there will then be a `data collision`. If all of
the communicating vehicles are considered as being part of a local
area network, the standard Ethernet protocol can be used to solve
this problem. In that protocol, when a data collision occurs, each
of the transmitting vehicles which was transmitting at the time of
the data collision would be notified that a data collision had
occurred and that they should retransmit their message at a random
time later. When several vehicles are in the vicinity and there is
the possibility of collisions of the data, each vehicle can retain
the coordinates last received from the surrounding vehicles as well
as their velocities and predict their new locations even though
some data was lost.
If a line of sight system is used, an infrared, terahertz or MIR
system would be good choices. In the infrared case, and if an
infrared system were also used to interrogate the environment for
non-equipped vehicles, pedestrians, animals etc., as discussed
below, both systems could use some of the same hardware.
If point-to-point communication can be established between
vehicles, such as described in U.S. Pat. No. 5,528,391 to Elrod,
then the need for a collision detection system like Ethernet would
not be required. If the receiver on a vehicle, for example, only
has to listen to one sender from one other vehicle at a time, then
the bandwidth can be considerably higher since there will not be
any interruption.
When two vehicles are communicating their positions to each other,
it is possible through the use of range gating or the sending of a
"clear to send signal" and timing the response to determine the
separation of the vehicles. This assumes that the properties of the
path between the vehicles are known which would be the case if the
vehicles are within view of each other. If, on the other hand,
there is a row of trees, for example, between the two vehicles, a
false distance measurement would be obtained if the radio waves
pass through a tree. If the communication frequency is low enough
that it can pass through a tree in the above example, it will be
delayed. If it is a much higher frequency such that is blocked by
the tree, then it still might reach the second vehicle through a
multi-path. Thus, in both cases, an undetectable range error
results. If a range of frequencies is sent, as in a spread spectrum
pulse, and the first arriving pulse contains all of the sent
frequencies, then it is likely that the two vehicles are in view of
each other and the range calculation is accurate. If any of the
frequencies are delayed, then the range calculation can be
considered inaccurate and should be ignored. Once again, for range
purposes, the results of many transmissions and receptions can be
used to improve the separation distance accuracy calculation.
Alternate methods for determining range can make use of radar
reflections, RFID tags etc.
8.3 Enhancements
In the accident avoidance system of the present invention, the
information indicative of a collision could come from a vehicle
that is quite far away from the closest vehicles to the subject
vehicle. This is a substantial improvement over the prior art
collision avoidance systems, which can only react to a few vehicles
in the immediate vicinity. The system described herein also permits
better simultaneous tracking of several vehicles. For example, if
there is a pileup of vehicles down the highway, then this
information can be transmitted to control other vehicles that are
still a significant distance from the accident. This case cannot be
handled by prior art systems. Thus, the system described here has
the potential to be used with the system of the U.S. Pat. No.
5,572,428 to Ishida, for example.
The network analogy can be extended if each vehicle receives and
retransmits all received data as a single block of data. In this
way, each vehicle is assured in getting all of the relevant
information even if it gets it from many sources. Even with many
vehicles, the amount of data being transmitted is small relative to
the bandwidth of the infrared optical or radio technologies. In
some cases, a receiver and re-transmitter can be part of the
highway infrastructure. Such a case might be on a hairpin curve in
the mountains where the oncoming traffic is not visible.
In some cases, it may be necessary for one vehicle to communicate
with another to determine which evasive action each should take.
This could occur in a multiple vehicle situation when one car has
gone out of control due to a tire failure, for example. In such
cases, one vehicle may have to tell the other vehicle what evasive
actions it is planning. The other vehicle can then calculate
whether it can avoid a collision based on the planned evasive
action of the first vehicle and if not, it can inform the first
vehicle that it must change its evasive plans. The other vehicle
would also inform the first vehicle as to what evasive action it is
planning. Several vehicles communicating in this manner can
determine the best paths for all vehicles to take to minimize the
danger to all vehicles.
If a vehicle is stuck in a corridor and wishes to change lanes in
heavy traffic, the operator's intention can be signaled by the
operator activating the turn signal. This could send a message to
other vehicles to slow down and let the signaling vehicle change
lanes. This would be particularly helpful in an alternate merge
situation and have a significant congestion reduction effect.
8.4 Position-Based Code Communication
In conventional wireless communication such as between cell phones
and a cell phone station or computers in a local area network, a
limited number of clients are provided dedicated channels of
communication with a central server. The number of channels is
generally limited and the data transfer rate is maximized. The
situation of communication between vehicles (cars, trucks, buses,
boats, ships, airplanes) is different in that devices are all peers
and the communication generally depends on their proximity. In
general, there is no central server and each vehicle must be able
to communicate with each other vehicle without going through a
standard server.
Another distinguishing feature is that there may be a large number
of vehicles that can potentially communicate with a particular
vehicle. Thus, there needs to be a large number of potential
channels of communication. One method of accomplishing this is
based on the concept of noise radar as developed by Lukin et al.
and described in the following (all of which are incorporated by
reference herein):
1. K. A. Lukin. Noise Radar Technology for Short Range
Applications, Proc of the. 5th Int. Conference and Exhibition on
Radar Systems, (RADAR'99), May 17 21, Brest, France, 1999, 6
pages;
2. K. A. Lukin. Advanced Noise Radar Technology. Proc. of the PIERS
Workshop on Advances in Radar Methods. Apr. 20 22, 1998, Hotel
Dino, Baveno, Italy, JRC-Ispra 1998, pp. 137 140;
3. W. Keydel and K. Lukin. Summary of Discussion in working Group
V: Unconventional New Techniques and Technologies for Future Radar,
Proc. of the PIERS Workshop in Radar Methods. Apr. 20 22, 1998,
Hotel Dino, Baveno, Italy, 1998, pp. 28 30;
4. Lukin K. A., Hilda A. Cerdeira and Colavita A. A. Chaotic
instability of currents in reverse biased multilayered structure.
Appl. Physics Letter, v. 77(17), 27 Oct. 1997, pp. 2484 2496;
5. K. A. Lukin. Noise Radar Technology for Civil Application. Proc.
of the 1 st EMSL User Workshop. 23 24 Apr. 1996, JRC-Ispra, Italy,
1997, pp. 105 112;
6. A. A. Mogyla. Adaptive signal filtration based on the
two-parametric representation of random processes. Collective
Volume of IRE NASU, Vol. 2, No. 2 pp. 137 141, 1997, (in
Russian);
7. A. A. Mogyla, K. A. Lukin. Two-Parameter Representation of
Non-Stationary Random Signals with a Finite Weighted Average Value
of Energy. The Collective Volume of IRE NASU, No. 1, pp. 118 124,
1996, (in Russian);
8. K. A. Lukin. Noise Radar with Correlation Receiver as the Basis
of Car Collision Avoidance System. 25th European Microwave
Conference, Bologna; Conference Proceedings, UK, Nexus, 1995, pp.
506 507, 1995;
9. K. A. Lukin, V. A. Rakityansky. Dynamic chaos in microwave
oscillators and its applications for Noise Radar development, Proc.
3rd Experimental Chaos Conference, Edinburg, Scotland, UK, 21 23
Aug., 1995;
10. V. A. Rakityansky, K. A. Lukin. Excitation of the chaotic
oscillations in millimeter BWO, International Journal of Infrared
and Millimeter Waves, vol. 16, No. 6, June, pp. 1037 1050,
1995;
11. K. A. Lukin. Ka-band Noise Radar. Proc. of the Millimeter and
Submillimeter Waves, Jun. 7 10 1994, Kharkov, Ukraine; Vol. 2, pp.
322 324, 1994;
12. K. A. Lukin, Y. A. Alexandrov, V. V. Kulik, A. A. Mogila, V. A.
Rakityansky. Broadband millimeter noise radar, Proc. Int. Conf on
Modern Radars, Kiev, Ukraine, pp. 30 31, 1994 (in Russian);
13. K. A. Lukin. High-frequency chaotic oscillations from Chua's
circuit. Journal of Circuits, Systems, and Computers, Vol. 3, No.
2, June 1993, pp. 627 643; In the book: Chua's Circuit Paradigma
for Chaos, World Scientific, Singapore, 1993;
14. K. A. Lukin, V. A. Rakityansky. Application of BWO for
excitation of the intensive chaotic oscillations of millimeter wave
band. 23-rd European Microwave Conference. September 6 9, Madrid,
Spain. Conf. Proceed. pp. 798 799, 1993;
15. K. A. Lukin, V. A. Rakityansky. Excitation of intensive chaotic
oscillations of millimetre wave band. Proc. of ISSSE, Paris,
September 1 4, pp. 454 457, 1992;
16. V. V. Kulik, K. A. Lukin, V. A. Rakityansky. Non-Coherent
Reflectometry Method for Measurement of Plasma Cut-Off Layer
Position, Proc. of the Int. Conference on Millimeter Wave and
Far-Infrared. Technology, Beijing, China, 17 21 Aug., 1992;
17. V. V. Kulik, K. A. Lukin, V. A. Rakityansky. Autodyne effect in
BWO with chaotic dynamic. Collective Volume of IRE NASU, pp. 95
100, 1992, (in Russian);
18. V. V. Kulik, K. A. Lukin, V. A. Rakityansky. Application of
noncoherent reflectometry method for fusion plasma dyagnostic.
Collective Volume of IRE NASU, pp. 13 18, 1992, (in Russian);
19. B. P. Efimov, K. A. Lukin, V. A. Rakityansky. Chaotic
interaction of modes in the electron-wave auto-oscillator with two
feedback channels, Letters in Journal of Technical Physics, v. 15,
No. 18, pp. 9 12, 1989, (in Russian);
20. B. P. Efimov, K. A. Lukin, V. A. Rakityansky. Transformation of
chaotic oscillation power spectrum by reflections. Journal of
Technical Physics, vol. 58, No. 12, pp. 2388 2400, 1988 (in
Russian)).
The concept of noise radar is discussed in detail in the Lukin
references listed above. A description of noise radar is included
elsewhere herein and the discussion here will be limited to the use
of pseudo random noise in a spread spectrum or Ultra-wideband
spectrum environment for communication purposes.
Generally, a particular segment or band of the electromagnetic
spectrum which is compatible with FCC regulations will be selected
for vehicle-to-vehicle communication purposes. Such a band could
include, for example 5.9 GHz to 5.91 GHz. The noise communication
device will therefore transmit information in that band. Each
vehicle will transmit a pseudorandom noise signal in a carrier-less
fashion composed of frequencies within the chosen band. The
particular code transmitted by a particular vehicle should be
unique. Generally, the vehicle will transmit its code repetitively
with a variable or fixed spacing between transmissions. The
information which the vehicle wishes to transmit is encoded using
the vehicle's code by any of a number of different techniques
including phase modulation of distance or time between code
transmissions, phase or amplitude modulation of the code sequences
themselves, changes of the polarity of the entire code sequence or
the individual code segments, or bandwidth modulation of the code
sequence. Other coding technologies would also applicable and this
invention is not limited to any particular coding method.
For example, a vehicle can have a 64 bit code which is a
combination of a vehicle identification number and the GPS
coordinates of the vehicle location. The vehicle would continuously
transmit this 64 bit code using frequencies within the selected
band. The 64 bit code could include both positive and negative bits
in addition to 0 bits. When identifying the vehicle, the receiver
could rectify the bits resulting in a 64 bit code of 0's and 1's.
The information which the transmitting vehicle wishes to send could
be represented by the choice of polarity of each of the code
bits.
Once a particular vehicle begins communicating with another
particular vehicle, the communication channel must remain intact
until the entire message has been transmitted. Since there may be
as many as 100 to 1000 vehicles simultaneously transmitting within
radio range of the receiving vehicle, a transmitting vehicle must
have a code which can be known to the receiving vehicle. One
preferred technique is to make this identification code a function
of the GPS coordinate location of transmitting vehicle. The code
would need to be coarse enough so that information to be
transmitted by the transmitting vehicle is accomplished before the
transmitting vehicle changes its identification. If this
information includes a position and velocity of the transmitting
vehicle, then the receiving vehicle can determine the new
transmitting code of the transmitting vehicle.
For example, the transmitting vehicle determines its location
within one meter. It is unlikely that any other vehicle will be
located within the same meter as the transmitting vehicle. Thus,
the transmitting vehicle will have a unique code which it can send
as a pseudorandom sequence in the noise communication system. A
nearby vehicle can search all information received by its antenna
for a sequence which represents each space within 30 meters of the
receiving vehicle. If it detects such a sequence, it will know that
there are one or more vehicles within 30 meters of the receiving
vehicle. The search can now be refined to locate vehicles based on
their direction since again the receiving vehicle can calculate the
sequences that would be transmitted from a vehicle from any
particular location within the 30 meter range. Once a particular
vehicle has been identified, the receiving vehicle can begin to
receive information from the transmitting vehicle through one or
more of the coding schemes listed above. Since the information will
preferably contain at least the velocity of transmitting vehicle,
the receiving vehicle can predict any code sequence changes that
take place and thus maintain communication with a particular
vehicle even as the vehicle's code changes due to its changing
position. The information being transmitted can also contain
additional information about the vehicle and/or its occupants.
In this manner, a receiving vehicle can selectively receive
information from any vehicle within its listenable range. Such
range may be limited to 100 meters for a highly congested area or
extend to 5000 meters in a rural environment. In this manner, each
vehicle becomes a node on the temporary local area network and is
only identified by its GPS location. Any vehicle can communicate
with any other vehicle and when many vehicles are present, a
priority scheme can be developed based on the urgency of the
message, the proximity of vehicle, the possibility of a collision,
or other desired prioritizing scheme.
The code transmitted by a particular vehicle will begin with a
sequence that indicates, for example, the largest GPS segment that
locates the vehicle which may be a segment 100 km square, for
example. The next bits in the sequence would indicate which of next
lower subsections which, for example, could be 10 km square. The
next set of bits could further refine this to a 1 km square area
and so on down to the particular square meter where the vehicle is
located. Other units, such as angles, degrees, minutes, seconds
etc., could be more appropriate for locating a vehicle on the
surface of spherical earth. By using this scheme, a receiving
vehicle can search for all vehicles located within its 1 km or
square segment and then when a vehicle is found, the search can be
continuously refined until the exact location of the transmitting
vehicle has been determined. This is done through correlation. The
100 or so vehicles transmitting with a range would all transmit low
level signals which would appear as noise to the receiving vehicle.
The receiving vehicle would need to know the code a particular
vehicle was transmitting before it could identify whether that code
was present in the noise. The code derived by the vehicle to be
transmitted must be sufficiently unique that only one vehicle can
have a particular code at a particular time. Since the messages
from different vehicles are separated through correlation
functions, all vehicles must have unique transmission codes which
are not known beforehand by the receiving vehicle yet must be
derivable by the receiving vehicle.
The communication need not be limited to communication between
moving vehicles. This same technology permits communication between
a vehicle and an infrastructure-based station.
There is no limit to the types of information that can be exchanged
between vehicles or between vehicles and infrastructure-based
stations. For example, if an event occurs such as an accident or
avalanche, road erosion, fallen tree, or other event which
temporarily changes the ability to travel safely on a section of a
lane on a highway, an authorized agent can place the transmitting
sign near the affected section of roadway which would transmit
information using the noise communication technique to all oncoming
vehicles within a 1 km range, for example. Prior to the placement
of such a sign, a police vehicle could transmit a similar message
to adjacent vehicles. Even an ordinary driver who first appears on
the scene and identifies a potential hazard can send this message
to vehicles within range of the hazard and can tag this message as
a high priority message. An infrastructure-based receiving station
can receive such a message and notify the emergency crews that
attention is immediately required at a particular location on the
highway. In this manner, all vehicles that could be affected by
such an event as well as urgency response organizations can be
immediately notified as soon as a hazard, such as an accident,
occurs thereby greatly reducing the response time and minimizing
the chance of vehicles engaging the hazardous location.
If a vehicle passes through a precise positioning location as
described elsewhere herein, that vehicle (the vehicle's processor
or computer) momentarily knows or can calculate the errors in the
GPS signals and thus becomes a differential correction station. The
error corrections can then be transmitted to nearby vehicles plus
enhancing their knowledge of their position. If the PPS vehicle
also has an onboard accurate clock, then the carrier phase of the
satellite signals at the PPS location can be predicted and thus, as
the vehicle leaves the PPS station, it can operate on carrier phase
RTK differential GPS and thus know its position within centimeters
or less. Similarly, if the phase of the carrier waves at PPS
station is transmitted to adjacent vehicles, each vehicle also can
operate on RTK carrier phase differential GPS. Thus, as many cars
pass the PPS the accuracy with which each vehicle knows its
position is continuously upgraded and at the time when the
likelihood of collision between vehicles is a maximum, that is when
many vehicles are traveling on a roadway, the accuracy with which
each vehicle knows its location is also maximized. The RtZF.TM.
system automatically improves as the danger of collision
increases.
Other information which a vehicle can transmit relates to the GPS
signals that it is receiving. In this manner, another form of
differential GPS can occur called relative differential GPS.
Without necessarily improving the accuracy with which a given
vehicle precisely knows its position, by comparing GPS signals from
one vehicle to another, the relative location of two vehicles can
again be very accurately determined within centimeters. This of
course is particularly important for collision avoidance.
Other information that can be readily transmitted either from
vehicle to vehicle or from infrastructure-based stations to
vehicles includes any recent map updates. Since a vehicle will
generally always be listening, whenever a map update occurs this
information can be received by a vehicle provided it is within
range of a transmitter. This could occur overnight while the
vehicle is in the garage, for example. Each vehicle would have a
characteristic time indicating the freshness of the information in
its local map database. As the vehicle travels and communicates
with other vehicles, this date can be readily exchanged and if a
particular vehicle has a later map version than the other vehicle,
it would signal the first vehicle requesting that the differences
between the two map databases be transmitted from the first to the
second vehicle. This transmission can also occur between an
infrastructure-based station and a vehicle. Satellites, cell phone
towers, etc. can also be used for map updating purposes.
If the operator of a particular vehicle wishes to send a text or
voice message to another identified vehicle, this information can
also be sent through the vehicle-to-vehicle communication system
described herein. Similarly, interaction with the Internet via an
infrastructure-based station can also be accomplished. In some
cases, it may be desirable to access the Internet using
communication channels with other vehicles. Perhaps, one vehicle
has the satellite, Wi-Fi, Wimax or other link to the Internet while
a second vehicle does not. The second vehicle could still
communicate with the Internet through the first Internet-enabled
vehicle.
Through the communication system based on noise or pseudonoise
communication as described above is ubiquitous, the number of paths
through which information can be transmitted to and from a vehicle
is substantially increased which also greatly increases the
reliability of the system since multiple failures can occur without
affecting the overall system operation. Thus, once again the goal
of zero fatalities.TM. is approached through this use of
vehicle-to-vehicle communication.
By opening this new paradigm for communication between vehicles,
and through the use of message relay from one vehicle to another,
occupants of one vehicle can communicate with any other vehicle on
a road. Similarly, through listening to infrastructure-based
stations, the occupants can communicate with non-vehicle occupants.
In many ways, this system supplements the cell phone system but is
organized under totally different principles. In this case, the
communication takes place without central stations or servers.
Although servers and central stations can be attached to the
system, the fundamental structure is one of independent nodes and
temporary connections based on geographic proximity.
The system is self limiting in that the more vehicles
communicating, the higher the noise level and the more difficult it
will be to separate more distant transmitters. When a vehicle is
traveling in a rural environment, for example, where there are few
sparsely located transmitters, the noise level will be low and
communication with more distant vehicles facilitated. On the other
hand, during rush hour, there will be many vehicles simultaneously
communicating thus raising the noise level and limiting the ability
of a receiver to receive distant transmissions. Thus, the system is
automatically adjusting.
There are several collision avoidance-based radar systems being
implemented on vehicles on the highways today. The prominent
systems include ForeWarn.TM. by Delco division of the Delphi
Corporation and the Eaton Vorad systems. These systems are
acceptable as long as few vehicles on the roads have such system.
As the number of radar-equipped vehicles increases, the reliability
of each system decreases as radar transmissions are received that
originate from other vehicles. This problem can be solved through
the use of noise radar as described in the various technical papers
by Lukin et al. listed above.
Noise radar typically operates in a limited band of frequencies
similarly to spread spectrum technologies. Whereas spread spectrum
utilizes a form of carrier frequency modulation, noise radar does
not. It is carrier-less. Typically, a noise-generating device is
incorporated into the radar transmitter such that the signal
transmitted appears as noise to any receiver. A portion of the
noise signal is captured as it is transmitted and fed to a delay
line for later use in establishing a correlation with a reflected
pulse. In the manner described in the Lukin et al. papers, the
distance and velocity of a reflecting object relative to the
transmitter can be readily determined and yet be detectable by any
other receiver. Thus, a noise radar collision avoidance system such
as discussed in U.S. Pat. No. 6,121,915, U.S. Pat. No. 5,291,202,
U.S. Pat. No. 5,719,579, and U.S. Pat. No. 5,075,863 becomes
feasible. Lukin et al. first disclosed this technology in the
above-referenced papers.
Although noise radar itself is not new, the utilization of noise
radar for the precise positioning system described herein is not
believed to have been previously disclosed by others. Similarly,
the use of noise radar for detecting the presence of an occupant
within a vehicle or of any object within a particular range of a
vehicle is also not believed to have been previously disclosed by
others. By setting the correlation interval, any penetration or
motion of an object within that interval can be positively
detected. Thus, if interval is sent at 2 meters, for example, the
entire interior of a vehicle can be monitored with one simple
device. If any object is moving within the vehicle, then this can
readily detected. Similarly, the space being monitored can be
limited to a portion of the interior of the vehicle such as the
right passenger seat or the entire rear seat. In this manner, the
presence of any moving object within that space can be determined
and thus problems such as a hiding assailant or a child or animal
left in a parked car can be addressed. A device placed in the trunk
can monitor the motion of any object that has been trapped within
the trunk thereby eliminating that well-known problem.
The radar system to be used for the precise positioning system can
also be used for monitoring the space around a vehicle. In this
case, a simple structure involving the placement of four antennas
on the vehicle roof, for example, can be used to locate and
determine the velocity of any object approaching or in the vicinity
of the vehicle. Using neural networks and the reflection received
from the four antennas, the location and velocity of an object can
be determined and by observing the signature using pattern
recognition techniques such as neural networks, the object can be
identified. Each antenna would send and receive noise radar waves
from an angle of, for example, 180 degrees. One forward and one
rear antenna could monitor the left side of the vehicle and one
forward and one rear antenna could monitor the right side.
Similarly, the two rear antennas could monitor the rear of the
vehicle and the two forward antennas could monitor the forward part
of the vehicle. In this manner, one simple system provides rear
impact anticipatory sensing, automatic cruise control, forward
impact anticipatory sensing, blind spot detection, and side impact
anticipatory sensing. Since the duty cycle of the precise
positioning system is small, most of the time would be available
for monitoring the space surrounding the vehicle. Through the
choice of the correlation interval and coding scheme (CDMA, noise,
etc.), the distance monitored can also be controlled.
In addition to the position-based code, an ID related to the type
of vehicle could also be part of the code so that an interested
vehicle may only wish to interrogate vehicles of a certain class
such as emergency vehicles. Also having information about the
vehicle type communicated to the host vehicle can quickly give an
indication of the mass of the oncoming vehicle which, for example,
could aid an anticipatory sensor in projecting the severity of an
impending crash.
Although it has been generally assumed that vehicle-to-vehicle
communication will take place through a direct link or through an
ad-hoc or mesh network, when Internet access becomes ubiquitous for
vehicles, this communication could also take place via the Internet
through a Wi-Fi or Wimax or equivalent link. Additionally, the use
of an ad-hoc or mesh network for vehicle-to-vehicle communication
especially to sending: relative location, velocity and vehicle mass
information for collision avoidance purposes; GPS, DGPS, PPS
related information for location determination and error correction
purposes; traffic congestion or road condition information; weather
or weather related information; and, vehicle type information
particularly for emergency vehicle identification so that the host
vehicle can take appropriate actions to allow freedom of passage
for the emergency vehicle, are considered important parts of the
present inventions. In fact, a mesh or ad-hoc network can greatly
improve the working of an ubiquitous WI-FI, Wimax or equivalent
Internet system thereby extending the range of the wireless
Internet system.
This system also supports emergency vehicles sending warnings to
vehicles that are in its path since it, and only it, will know its
route from its present location to its destination. Such a system
will permit significant advanced warning to vehicles on the route
and also allow for the control of traffic lights based on its
planned route long before it arrives at the lights. In this regard,
see "Private Inventor Files Patent Application For Telematics-Based
Public and Emergency First Responders Safety Advisory System", ITS
America News Release Feb. 13, 2004, for a discussion of a primitive
but similar system.
An alternate approach to using the code based on location system is
to use a vehicle ID system in connection with an easily accessible
central database that relates the vehicle ID to its location. Then
communication can take place via a code based in the vehicle ID, or
some equivalent method.
9. Infrastructure-to-Vehicle Communication
Initial maps showing roadway lane and boundary location for the
CONUS can be installed within the vehicle at the time of
manufacture. The vehicle thereafter would check on a
section-by-section basis whether it had the latest update
information for the particular and surrounding locations where it
is being operated. One method of verifying this information would
be achieved if a satellite or Internet connection periodically
broadcasts the latest date and time or version that each segment
had been most recently updated. This matrix would amount to a small
transmission requiring perhaps a few seconds of airtime. Any
additional emergency information could also be broadcast in between
the periodic transmissions to cover accidents, trees falling onto
roads etc. If the periodic transmission were to occur every five
minutes and if the motion of a vehicle were somewhat restricted
until it had received a periodic transmission, the safety of the
system can be assured. If the vehicle finds that it does not have
the latest map information, vehicle-to-vehicle communication,
vehicle-to-infrastructure communication, Internet communication
(Wi-Fi, Wi-max or equivalent), or the cell phone in the vehicle can
be used to log onto the Internet, for example, and the missing data
downloaded. An alternate is for the GEOs, LEOs, or other
satellites, to broadcast the map corrections directly.
It is also possible that the map data could be off-loaded from a
transmitter on the highway itself or at a gas station, for example,
as discussed above. In that manner, the vehicles would only obtain
that map information which is needed and the map information would
always be up to the minute. As a minimum, temporary data
communication stations can be placed before highway sections that
are undergoing construction or where a recent blockage has
occurred, as discussed above, and where the maps have not yet been
updated. Such an emergency data transfer would be signaled to all
approaching vehicles to reduce speed and travel with care. Such
information could also contain maximum and minimum speed
information which would limit the velocity of vehicles in the
area.
There is other information that would be particularly useful to a
vehicle operator or control system, including in particular, the
weather conditions, especially at the road surface. Such
information could be obtained by road sensors and then transmitted
to all vehicles in the area by a permanently installed system as
disclosed above and in the current assignee's U.S. Pat. No.
6,662,642. Alternately, there have been recent studies that show
that icing conditions on road surfaces, for example, can be
accurately predicted by local meteorological stations and broadcast
to vehicles in the area. If such a system is not present, then the
best place to measure road friction is at the road surface and not
on the vehicle. The vehicle requires advance information of an
icing condition in order to have time to adjust its speed or take
other evasive action. The same road-based or local meteorological
transmitter system could be used to warn the operators of traffic
conditions, construction delays etc. and to set the local speed
limit. Once one vehicle in an area has discovered an icing
condition, for example, this information can be immediately
transmitted to all equipped vehicles through the vehicle-to-vehicle
communication system discussed above.
A number of forms of infrastructure-to-vehicle communication have
been discussed elsewhere herein. These include map and differential
GPS updating methods involving infrastructure stations which may be
located at gas stations, for example. Also communications with
precise positioning stations for GPS independent location
determination have been discussed. Communications via the Internet
using either satellite Internet services with electronic steerable
antennas such as are available from KVH, Wi-Fi or Wimax which will
undoubtedly become available ubiquitously throughout the CONUS, for
example. All of the services that are now available on the Internet
plus may new services will thus be available to vehicle operators
and passengers. The updating of vehicle resident software will also
become automatic via such links. The reporting of actual
(diagnostics) and forecasted (prognostics) vehicle failures will
also able to be communicated via one of these links to the
authorities, the smart highway monitoring system, vehicle dealers
and manufacturers (see U.S. patent application Ser. No. 10/701,361,
now U.S. Pat. No. 6,988,026). This application along with the
inventions herein provide a method of notifying interested parties
of the failure or forecasted failure of a vehicle component using a
vehicle-to-infrastructure communication system. Such interested
parties can include, but are not limited to: a vehicle manufacturer
so that early failures on a new vehicle model can be discovered so
as to permit an early correction of the problem; a dealer so that
it can schedule fixing of the problem so as to provide for the
minimum inconvenience of their customer and even, in some cases,
dispatching a service vehicle to the location of the troubled
vehicle; NHTSA so that they can track problems (such as for
Firestone tire problem) before they become a national issue; the
police, EMS, fire department and other emergency services so that
they can prepare for a potential emergency etc. For example in
"Release of Auto Safety Data Is Disputed", New York Times Dec. 13,
2002 it is written "After Firestone tire failures on Ford Explorers
led to a national outcry over vehicle safety, Congress ordered a
watchdog agency to create an early-warning system for automotive
defects that could kill or injure people." The existence of the
system disclosed herein would provide an automatic method for such
a watchdog group to monitor all equipped vehicles on the nation's
highways. As a preliminary solution, it is certainly within the
state of the art today to require all vehicles to have an emergency
locator beacon or equivalent that is impendent of the vehicle's
electrical system and is activated on a crash, rollover or similar
event.
Although the '361 application primarily discusses diagnostic
information for the purpose of reporting present or forecasted
vehicle failures, there is of course a wealth of additional data
that is available on a vehicle related to the vehicle operation,
its location, its history etc. where an interested party may desire
that such data be transferred to a site remote from the vehicle.
Interested parties could include the authorities, parents,
marketing organizations, the vehicle manufacturer, the vehicle
dealer, stores or companies that may be in the vicinity of the
vehicle, etc. There can be significant privacy concerns here which
have not yet been addressed. Nevertheless, with the proper
safeguards the capability described herein is enabled partially by
the teachings of this invention.
For critical functions where a software-induced system failure
cannot be tolerated, even the processing may occur on the network
achieving what pundits have been forecasting for years that "the
network is the computer". Vehicle operators will also have all of
the functions now provided by specialty products such as PDAs, the
Blackberry, cell phones etc. available as part of the
infrastructure-to-vehicle communication systems disclosed
herein.
There are of course many methods of transferring data wirelessly in
addition to the CDMA system described above. Methods using ultra
wideband signals were first disclosed by the current assignee in
previous patents and are reinforced here. Much depends of the will
of the FCC as to what method will eventually prevail. Ultra
wideband within the frequency limits set by the FCC is certainly a
prime candidate and lends itself to the type of CDMA system where
the code is derivable from the vehicle's location as determined,
for example, by the GPS that this is certainly a preferred method
for practicing the teachings disclosed herein.
Note that different people may operate a particular vehicle and
when a connection to the Internet is achieved, the Internet may not
know the identity of the operator or passenger, for the case where
the passenger wishes to operate the Internet. One solution is for
the operator or passenger to insert a smart card, plug in their PDA
or cell phone or otherwise electronically identify themselves.
Transponders are contemplated by the inventions disclosed herein
including SAW, RFID or other technologies that can be embedded
within the roadway or on objects beside the roadway, in vehicle
license plates, for example. An interrogator within the vehicle
transmits power to the transponder and receives a return signal.
Alternately, as disclosed above, the responding device can have its
own source of power so that the vehicle located interrogator need
only receive a signal in response to an initiated request. The
source of power can be a battery, connection to an electric power
source such as an AC circuit, solar collector, or in some cases the
energy can be harvested from the environment where vibrations, for
example, are present. The range of a license-mounted transponder,
for example, can be greatly increased if such a vibration-based
energy harvesting system is incorporated.
Some of the systems disclosed herein make use of an energy beam
that interrogates a reflector or retransmitting device. Such a
device can be a sign as well as any pole with a mounted reflector,
for example. In some cases, it will be possible for the
infrastructure device to modify its message so that when
interrogated it can provide information in addition to its
location. A speed limit sign, for example, can return a variable
code indicating the latest speed limit that then could have been
set remotely by some responsible authority. Alternately,
construction zones frequently will permit one speed when workers
are absent and another when workers are present. The actual
permitted speed can be transmitted to the vehicle when it is
interrogated or as the vehicle passes. Thus, a sign or reflector
could also be an active sign and this sign could be an active
matrix organic display and solar collector that does not need a
connection to a power line and yet provides both a visual message
and transmits that message to the vehicle for in-vehicle signage.
Each of these systems has the advantage that since minimal power is
required to operate the infrastructure-based sign, it would not
require connection to a power line. It would only transmit when
asked to do so either by a transmission from the vehicle or by
sensing that a vehicle is present.
A key marketing point for OnStar.RTM. is their one button system.
This idea can be generalized in that a vehicle operator can summon
help or otherwise send a desired message to a remoter site by
pushing a single button. The message sent can just be a distress
message or it can perform a particular function selected by the
vehicle depending on the emergency or from a menu selected by the
operator. Thus, the OnStar.RTM. one button concept is retained but
the message can be different for different situations.
10. The RtZF.TM. System
10.1 Technical Issues
From the above discussion, two conclusions should be evident. There
are significant advantages in accurately knowing where the vehicle,
the roadway and other vehicles are and that possession of this
information is the key to reducing fatalities to zero. Second,
there are many technologies that are already in existence that can
provide this information to each vehicle. Once there is a clear,
recognized direction that this is the solution, then many new
technologies will emerge. There is nothing inherently expensive
about these technologies and once the product life cycle is
underway, the added cost to vehicle purchasers will be minimal.
Roadway infrastructure costs will be minimal and system maintenance
costs almost non-existent.
Most importantly, the system has the capability of reducing
fatalities to zero!
The accuracy of DGPS has been demonstrated numerous times in small
controlled experiments, most recently by the University of
Minnesota and SRI.
The second technical problem is the integrity of the signals being
received and the major cause of the lack of integrity is the
multi-path effect. Considerable research has gone into solving the
multi-path effect and Trimble, for example, claims that this
problem is no longer an issue.
The third area is availability of GPS and DGPS signals to the
vehicle as it is driving down the road. The system is designed to
tolerate temporary losses of signal, up to a few minutes. That is
the prime function of the inertial navigation system (INS or IMU).
Prolonged absence of the GPS signal will significantly degrade
system performance. There are two primary causes of lack of
availability, namely, temporary causes and permanent causes.
Temporary causes result from a car driving between two trucks for
an extended period of time, blocking the GPS signals. The eventual
solution to this problem is to change the laws to prevent trucks
from traveling on both sides of an automobile. If this remains a
problem, a warning will be provided to the driver that he/she is
losing system integrity and therefore he/she should speed up or
slow down to regain a satellite view. This could also be done
automatically. Additionally, the vehicle can obtain its location
information through vehicle-to-vehicle communication plus a ranging
system so that if the vehicle learns the exact location of the
adjacent vehicle and its relative location, then it can determine
its absolute location. Of course, if the precise positioning system
is able to interrogate the environment, then the problem is also
solved via the PPS system.
Permanent blockage of the GPS signals, as can come from operating
the vehicle in a tunnel or downtown of a large city, can be
corrected through the use of pseudolites or other guidance systems
such as the SnapTrack system or the PPS described here. This is not
a serious problem since very few cars run off the road in a tunnel
or in downtown areas. Eventually, it is expected that the PPS will
become ubiquitous thereby rendering GPS as the backup system.
Additional methods for location determine to aid in reacquiring the
satellite lock include various methods based on cell phones and
other satellite systems such as the Skybitz system that can locate
a device with minimal information.
The final technical impediment is the operation of the diagnostic
system that verifies that the system is operating properly. This
requires an extensive failure mode and effect analysis and the
design of a diagnostic system that answers all of the concerns of
the FMEA.
10.2 Cost Issues
The primary cost impediment is the cost of the DGPS hardware. A
single base station and roving receiver that will give an accuracy
of 2 centimeters (1.sigma.) currently costs about $25,000. This is
a temporary situation brought about by low sales volume. Since
there is nothing exotic in the receiving unit, the cost can be
expected to follow typical automotive electronic life-cycle costs
and therefore the projected high volume production cost of the
electronics for the DGPS receivers is below $100 per vehicle. In
the initial implementation of the system, an OmniSTAR.TM. DGPS
system will be used providing an accuracy of 6 cm. The U.S.
national DGPS system is now coming on line and thus the cost of the
DGPS corrections will soon approach zero.
A similar argument can be made for the inertial navigation system.
Considerable research and development effort is ongoing to reduce
the size, complexity and cost of these systems. Three technologies
are vying for this rapidly growing market: laser gyroscopes,
fiber-optic lasers, and MEMS systems. The cost of these units today
range from a few hundred to ten thousand dollars each, however,
once again this is due to the very small quantity being sold.
Substantial improvements are being made in the accuracies of the
MEMS systems and it now appears that such a system will be accurate
enough for RtZF.TM. purposes. The cost of these systems in
high-volume production is expected to be on the order of ten
dollars each. This includes at least a yaw rate sensor with three
accelerometers and probably three angular rate sensors. The
accuracy of these units is currently approximately 0.003 degrees
per second. This is a random error which can be corrected somewhat
by the use of multiple vibrating elements. A new laser gyroscope
has recently been announced by Intellisense Corporation which
should provide a dramatic cost reduction and accuracy
improvement.
One of the problems keeping the costs high is the need in the case
of MEMS sensors to go through an extensive calibration process
where the effects of all influences such as temperature, pressure,
vibration, and age is determined and a constitute equation is
derived for each device. A key factor in the system of the
inventions here is that this extensive calibration process is
eliminated and the error corrections for the IMU are determined
after it is mounted on the vehicle through the use of a Kalman
filter, or equivalent, coupled with input from the GPS and DGPS
system and the precise positioning system. Other available sensors
are also used depending on the system. These include a device for
measuring the downward direction of the earth's magnetic field, a
flux gage compass, a magnetic compass, a gravity sensor, the
vehicle speedometer and odometer, the ABS sensors including wheel
speed sensors, and whatever additional appropriate sensors that are
available. Over time, the system can learn of the properties of
each component that makes up the IMU and derive the constituent
equation for that component which, although it will have little
effect on the instantaneous accuracy of the component, it will
affect the long term accuracy and speed up the calculations.
Eventually, when most vehicles on the road have the RtZF.TM.
system, communication between the vehicles can be used to
substantially improve the location accuracy of each vehicle as
described above.
The cost of mapping the CONUS is largely an unknown at this time.
OmniSTAR.RTM. has stated that they will map any area with
sufficient detail at a cost of $300 per mile. They have also
indicated the cost will drop substantially as the number of miles
to be mapped increases. This mapping would be done by helicopter
using cameras and their laser ranging system. Another method is to
outfit a ground vehicle with equipment that will determine the
location of the lane and shoulder boundaries of road and other
information. Such a system has been used for mapping a Swedish
highway. One estimate is that the mapping of a road will be reduced
to approximately $50 per mile for major highways and rural roads
and a somewhat higher number for urban areas. The goal is to map
the country to an accuracy of 2 to 10 centimeters (1.sigma.).
Related to the costs of mapping is the cost of converting the raw
data acquired either by helicopter or by ground vehicle into a
usable map database. The cost for manually performing this
vectorization process has been estimated at $100 per mile by
OmniSTAR.RTM.. This process can be substantially simplified through
the use of raster-to-vector conversion software. Such software is
currently being used for converting hand drawings into CAD systems,
for example. The Intergraph Corp. provides hardware and software
for simplifying this task. It is therefore expected that the cost
for vectorization of the map data will follow proportionately a
similar path to the cost of acquiring the data and may eventually
reach $10 to $20 per mile for the rural mapping and $25 to a $50
per mile for urban areas. Considering that there are approximately
four million miles of roads in the CONUS, and assuming we can
achieve an average of $150 for acquiring the data and converting
the data to a GIS database can be achieved, the total cost for
mapping all of the roads in U.S. will amount to $600 million. This
cost would obviously be spread over a number of years and thus the
cost per year is manageable and small in comparison to the $215
billion lost every year due to death, injury and lost time from
traffic congestion.
Another cost factor is the lack of DGPS base stations. The initial
analysis indicated that this would be a serious problem as using
the latest DGPS technology requires a base station every 30 miles.
Upon further research, however, it has been determined that the
OmniSTAR company has now deployed a nationwide WADGPS system with 6
cm accuracy. The initial goal of the RtZF.TM. system was to achieve
2 cm accuracy for both mapping and vehicle location. The 2 cm
accuracy can be obtained in the map database since temporary
differential base stations will be installed for the mapping
purposes. By relaxing the 2 cm requirement to 6 cm, the need for
base stations every 30 miles disappears and the cost of adding a
substantial number of base stations is no longer a factor.
The next impediment is the lack of a system for determining when
changes are planned for the mapped roads. This will require
communication with all highway and road maintenance organizations
in the mapped area.
A similar impediment to the widespread implementation of this
RtZF.TM. system is the lack of a communication system for supplying
map changes to the equipped vehicles.
10.3 Educational Issues
A serious impediment to the implementation of this system that is
related to the general lack of familiarity with the system, is the
belief that significant fatalities and injuries on U.S. highways
are a fact of life. This argument is presented in many forms such
as "the perfect is the enemy of the good". This leads to the
conclusion that any system that portends to reduce injury should be
implemented rather than taking the viewpoint that driving an
automobile is a process and as such it can be designed to achieve
perfection. As soon as it is admitted that perfection cannot be
achieved, then any fatality gets immediately associated with this
fact. This of course was the prevailing view among all
manufacturing executives until the zero defects paradigm shift took
place. The goal of the "Zero Fatalities".TM. program is not going
to be achieved in a short period of time. Nevertheless, to plan
anything short of zero fatalities is to admit defeat and to thereby
allow technologies to enter the market that are inconsistent with a
zero fatalities goal.
10.4 Potential Benefits when the System is Deployed.
10.4.1 Assumptions for the Application Benefits Analysis
The high volume incremental cost of an automobile will be $200. The
cost of DGPS correction signals will be a onetime charge of $50 per
vehicle. The benefits to the vehicle owner from up-to-date maps and
to the purveyors of services located on these maps. will cover the
cost of updating the maps as the roads change. The cost of mapping
substantially all roads in the CONUS will be $600 million. The
effects of phasing in the system will be ignored. There are 15
million vehicles sold in the U.S. each year. Of the 40,000 plus
people killed on the roadways, at least 10% are due to road
departure, yellow line infraction, stop sign infraction, excessive
speed and other causes which will be eliminated by the Phase Zero
deployment. $165 billion are lost each year due to highway
accidents. The cost savings due to secondary benefits will be
ignored. 10.4.2 Analysis Methods Described.
The analysis method will be quite simple. Assume that 10% of the
vehicles on the road will be equipped with RtZF.TM. systems in the
first year and that this will increase by 10% each year. Ten
percent or 4000 lives will be saved and a comparable percentage of
injuries. Thus, in the first year, one percent of $165 billion
dollars will be saved or $1.65 billion. In the second year, this
saving will be $3.3 billion and the third year $4.95 billion. The
first-year cost of implementation of the system will be $600
million for mapping and $3.75 billion for installation onto
vehicles. The first year cost therefore will be $4.35 billion and
the cost for the second and continuing years will be $3.75 billion.
Thus, by the third year, the benefits exceed the costs and by the
10th year, the benefits will reach $16.5 billion compared with
costs of $3.75 billion yielding a benefits-to-cost ratio of more
than 4.
Before the fifth year of deployment, it is expected that the other
parts of the RtZF.TM. system will begin to be deployed and that the
benefits therefore are substantially understated. It is also
believed that the $250 price for the Phase Zero system on a
long-term basis is high and it is expected that the price to drop
substantially. No attempt has been made to estimate the value of
the time saved in congestion or efficient operation of the highway
system. Estimates that have been presented by others indicate that
as much as a two to three times improvement in traffic through flow
is possible. Thus, a substantial portion of the $50 billion per
year lost in congestion delays will also be saved when the full
RtZF.TM. system is implemented.
It is also believed that the percentage reduction of fatalities and
injuries has been substantially understated. For the first time,
there will be some control over the drunk or otherwise
incapacitated driver. If the excessive speed feature is
implemented, then gradually the cost of enforcing the nation's
speed limits will begin to be substantially reduced. Since it is
expected that large trucks will be among first vehicles to be
totally covered with the system, perhaps on a retrofit basis, it is
expected that the benefits to commercial vehicle owners and
operators will be substantial. The retrofit market may rapidly
develop and the assumptions of vehicles with deployed systems may
be low. None of these effects have been taken into account in the
above analysis.
The automated highway systems resulting from RtZF.TM.
implementation is expected to double or even triple in effective
capacity by increasing speeds and shortening distances between
vehicles. Thus, the effect on highway construction cost could be
significant.
10.5 Initial System Deployment
The initial implementation of the RtZF.TM. system would include the
following services:
1. A warning is issued to the driver when the driver is about to
depart from the road.
2. A warning is issued to the driver when the driver is about to
cross a yellow line or other lane boundary.
3. A warning is provided to the driver when the driver is exceeding
a safe speed limit for the road geometry.
4. A warning is provided to the driver when the driver is about to
go through a stop sign without stopping.
5. A warning is provided to the driver when the driver is about to
run the risk of a rollover.
6. A warning will be issued prior to a rear end impact by the
equipped vehicle.
7. In-vehicle signage will be provided for highway signs (perhaps
with a multiple language option).
8. A recording will be logged whenever a warning is issued.
10.6 Other Uses
The RtZF.TM. system can replace vehicle crash and rollover sensors
for airbag deployment and other sensors now on or being considered
for automobile vehicles including pitch, roll and yaw sensors. This
information is available from the IMU and is far more accurate than
these other sensors. It can also be found by using carrier phase
GPS by adding more antennas to the vehicle. Additionally, once the
system is in place for land vehicles, there will be many other
applications such as surveying, vehicle tracking and aircraft
landing which will benefit from the technology and infrastructure
improvements. The automobile safety issue and ITS will result in
the implementation of a national system which provides any user
with low cost equipment the ability to know precisely where he is
within centimeters on the face of the earth. Many other
applications will undoubtedly follow.
10.7 Road Departure
FIG. 4 is a logic diagram of the system 50 in accordance with the
invention showing the combination 40 of the GPS and DGPS processing
systems 42 and an inertial reference unit (IRU) or inertial
navigation system (INS) or Inertial Measurement Unit (IMU) 44. The
GPS system includes a unit for processing the received information
from the satellites 2 of the GPS satellite system, the information
from the satellites 30 of the DGPS system and data from the
inertial reference unit 44. The inertial reference unit 44 contains
accelerometers and laser or MEMS gyroscopes.
The system shown in FIG. 4 is a minimal RtZF.TM. system that can be
used to prevent road departure, lane crossing and intersection
accidents, which together account for more than about 50% of the
fatal accidents in the U.S.
Map database 48 works in conjunction with a navigation system 46 to
provide a warning to the driver when he or she is about to run off
the road, cross a center (yellow) line, run a stop sign, or run a
red stoplight. The map database 48 contains a map of the roadway to
an accuracy of 2 cm (1.sigma.), i.e., data on the edges of the
lanes of the roadway and the edges of the roadway, and the location
of all stop signs and stoplights and other traffic control devices
such as other types of road signs. Another sensor, not shown,
provides input to the vehicle indicating that an approaching
stoplight is red, yellow or green. Navigation system 46 is coupled
to the GPS and DGPS processing system 42. For this simple system,
the driver is warned if any of the above events is detected by a
driver warning system 45 coupled to the navigation system 46. The
driver warning system 45 can be an alarm, light, buzzer or other
audible noise, or, preferably, a simulated rumble strip for yellow
line and "running off of road" situations and a combined light and
alarm for the stop sign and stoplight infractions.
10.8 Accident Avoidance
FIG. 5 is a block diagram of the more advanced accident avoidance
system of this invention and method of the present invention
illustrating system sensors, transceivers, computers, displays,
input and output devices and other key elements.
As illustrated in FIG. 5, the vehicle accident avoidance system is
implemented using a variety of microprocessors and electronic
circuits 100 to interconnect and route various signals between and
among the illustrated subsystems. GPS receiver 52 is used to
receive GPS radio signals as illustrated in FIG. 1. DGPS receiver
54 receives the differential correction signals from one or more
base stations either directly or via a geocentric stationary or LEO
satellite, an earth-based station or other means. Inter-vehicle
communication subsystem 56 is used to transmit and receive
information between various nearby vehicles. This communication
will in general take place via broad band or ultra-broad band
communication techniques, or on dedicated frequency radio channels,
or in the preferred mode, noise communication system as described
above. This communication may be implemented using multiple access
communication methods including FDMA, TDMA, or CDMA, or noise
communication system, in a manner to permit simultaneous
communication with and between a plurality of vehicles. Other forms
of communication between vehicles are possible such as through the
Internet. This communication may include such information as the
precise location of a vehicle, the latest received signals from the
GPS satellites in view, other road condition information, emergency
signals, hazard warnings, vehicle velocity and intended path, and
any other information which is useful to improve the safety of the
vehicle road system.
Infrastructure communication system 58 permits bidirectional
communication between the host vehicle and the infrastructure and
includes such information transfer as updates to the digital maps,
weather information, road condition information, hazard
information, congestion information, temporary signs and warnings,
and any other information which can improve the safety of the
vehicle highway system.
Cameras 60 are used generally for interrogating environment nearby
the host vehicle for such functions as blind spot monitoring,
backup warnings, anticipatory crash sensing, visibility
determination, lane following, and any other visual information
which is desirable for improving the safety of the vehicle highway
system. Generally, the cameras will be sensitive to infrared and/or
visible light, however, in some cases a passive infrared camera
will the used to detect the presence of animate bodies such as deer
or people on the roadway in front of the vehicle. Frequently,
infrared or visible illumination will be provided by the host
vehicle.
Radar 62 is primarily used to scan an environment close to and
further from the vehicle than the range of the cameras and to
provide an initial warning of potential obstacles in the path of
the vehicle. The radar 62 can also be used when conditions of a
reduced visibility are present to provide advance warning to the
vehicle of obstacles hidden by rain, fog, snow etc. Pulsed,
continuous wave, noise or micropower impulse radar systems can be
used as appropriate. Also, Doppler radar principles can be used to
determine the object to host vehicle relative velocity.
Laser or terahertz radar 64 is primarily used to illuminate
potential hazardous objects in the path of the vehicle. Since the
vehicle will be operating on accurate mapped roads, the precise
location of objects discovered by the radar or camera systems can
be determined using range gating and scanning laser radar as
described above or by phase techniques.
The driver warning system 66 provides visual and/or audible warning
messages to the driver or others that a hazard exists. In addition
to activating a warning system within the vehicle, this system can
activate sound and/or light systems to warn other people, animals,
or vehicles of a pending hazardous condition. In such cases, the
warning system could activate the vehicle headlights, tail lights,
horn and/or the vehicle-to-vehicle, Internet or infrastructure
communication system to inform other vehicles, a traffic control
station or other base station. This system will be important during
the early stages of implementation of RtZF.TM., however as more and
more vehicles are equipped with the system, there will be less need
to warn the driver or others of potential problems.
Map database subsystem 68, which could reside on an external memory
module, will contain all of the map information such as road edges
up to 2 cm accuracy, the locations of stop signs, stoplights, lane
markers etc. as described in detail above. The fundamental map data
can be organized on read-only magnetic or optical memory with a
read/write associated memory for storing map update information.
Alternatively, the map information can be stored on rewritable
media that can be updated with information from the infrastructure
communication subsystem 58. This updating can take place while the
vehicle is being operated or, alternatively, while the vehicle is
parked in a garage or on the street.
Three servos are provided for controlling the vehicle during the
later stages of implementation of the RtZF.TM. product and include
a brake servo 70, a steering servo 72, and a throttle servo 74. The
vehicle can be controlled using deterministic, fuzzy logic, neural
network or, preferably, neural-fuzzy algorithms.
As a check on the inertial system, a velocity sensor 76 based on a
wheel speed sensor, or ground speed monitoring system using lasers,
radar or ultrasonics, for example, can be provided for the system.
A radar velocity meter is a device which transmits a noise
modulated radar pulse toward the ground at an angle to the vertical
and measures the Doppler velocity of the returned signal to provide
an accurate measure of the vehicle velocity relative to the ground.
Another radar device can be designed which measures the
displacement of the vehicle. Other modulation techniques and other
radar systems can be used to achieve similar results. Other systems
are preferably used for this purpose such as the GPS/DGPS or
precise position systems.
The inertial navigation system (INS), sometimes called the inertial
reference unit or IRU, comprises one or more accelerometers 78 and
one or more gyroscopes 80. Usually, three accelerometers would be
required to provide the vehicle acceleration in the latitude,
longitude and vertical directions and three gyroscopes would be
required to provide the angular rate about the pitch, yaw and roll
axes. In general, a gyroscope would measure the angular rate or
angular velocity. Angular acceleration may be obtained by
differentiating the angular rate.
A gyroscope 80, as used herein in the IRU, includes all kinds of
gyroscopes such as MEMS-based gyroscopes, fiber optic gyroscopes
(FOG) and accelerometer based gyroscopes.
Accelerometer-based gyroscopes encompass a situation where two
accelerometers are placed apart and the difference in the
acceleration is used to determine angular acceleration and a
situation where an accelerometer is placed on a vibrating structure
and the Coriolis effect is used to obtain the angular velocity.
The possibility of an accelerometer-based gyroscope 80 in the IRU
is made possible by construction of a suitable gyroscope by
Interstate Electronics Corporation (IEC). IEC manufactures IMUs in
volume based on .mu.SCIRAS (micro-machined Silicon Coriolis
Inertial Rate and Acceleration Sensor) accelerometers. Detailed
information about this device can be found at the IEC website at
iechome.com.
There are two ways to measure angular velocity (acceleration) using
accelerometers. The first way involves installing the
accelerometers at a distance from one another and calculating the
angular velocity by the difference of readings of the
accelerometers using dependencies between the centrifugal and
tangential accelerations and the angular velocity/acceleration.
This way requires significant accuracy of the accelerometers.
The second way is based on the measurement of the Coriolis
acceleration that arises when the mass of the sensing element moves
at a relative linear speed and the whole device performs a
transportation rotation about the perpendicular axis. This
principle is a basis of all mechanical gyroscopes, including
micromachined ones. The difference of this device is that the
micromachined devices aggregate the linear oscillation excitation
system and the Coriolis acceleration measurement system, while two
separate devices are used in the proposed second method. The source
of linear oscillations is the mechanical vibration suspension, and
the Coriolis acceleration sensors are the micromachined
accelerometers. On one hand, the presence of two separate devices
makes the instrument bigger, but on the other hand, it enables the
use of more accurate sensors to measure the Coriolis acceleration.
In particular, compensating accelerometer systems could be used
which are more accurate by an order of magnitude than open
structures commonly used in micromachined gyroscopes.
Significant issues involved in the construction of an
accelerometer-based gyroscope are providing a high sensitivity of
the device, a system for measuring the suspension vibration,
separating the signals of angular speed and linear acceleration;
filtering noise in the output signals of the device at the
suspension frequency, providing a correlation between errors in the
channels of angular speed and linear acceleration, considering the
effect of nonlinearity of the accelerometers and the suspension on
the error of the output signals.
A typical MEMS-based gyroscope uses a quartz tuning fork. The
vibration of the tuning fork, along with applied angular rotation
(yaw rate of the car), creates Coriolis acceleration on the tuning
fork. An accelerometer or strain gage attached to the tuning fork
measures the minute Coriolis force. Signal output is proportional
to the size of the tuning fork. To generate enough output signal,
the tuning fork must vibrate forcefully. Often, this can be
accomplished with a high Q structure. Manufacturers often place the
tuning fork in a vacuum to minimize mechanical damping by air
around the tuning fork. High Q structures can be fairly
fragile.
The gyroscope often experiences shock and vibration because it must
be rigidly connected to the car to accurately measure yaw rate.
This mechanical noise can introduce signals to the Coriolis
pick-off accelerometer that is several orders of magnitude higher
than the tuning-fork-generated Coriolis signal. Separating the
signal from the noise is not easy. Often, the shock or vibration
saturates the circuitry and makes the gyroscope output unreliable
for a short time.
Conventional MEMS-based gyroscopes are usually bulky (100 cm.sup.3
or more is not uncommon). This is partly the result of the addition
of mechanical antivibration mounts, which are incorporated to
minimize sensitivity to external vibration.
New MEMS-based gyroscopes avoid these shortcomings, though. For
example, Analog Devices' iMEMS gyro is expected to be 7 by 7 by 3
mm (0.15 cm.sup.3). Rather than quartz, it uses a resonating
polysilicon beam structure, which creates the velocity element that
produces the Coriolis force when angular rate is presented to it.
At the outer edges of the polysilicon beam, orthogonal to the
resonating motion, a capacitive accelerometer measures the Coriolis
force. The gyroscope has two sets of beams in antiphase that are
placed next to each other, and their outputs are read
differentially, attenuating external vibration sensitivity.
An accelerometer 78, as used herein in the IRU, includes
conventional piezoelectric-based accelerometers, MEMS-based
accelerometers (such as made by Analog Devices) and the type as
described in US06182509 entitled "Accelerometer without proof
mass".
Display subsystem 82 includes an appropriate display driver and
either a heads-up or other display system for providing system
information to the vehicle operator. The information can be in the
form of non-critical information such as the location of the
vehicle on a map, as selected by the vehicle operator and/or it can
include warning or other emergency messages provided by the vehicle
subsystems or from communication with other vehicles or the
infrastructure. An emergency message that the road has been washed
out ahead, for example, would be an example of such a message.
Generally, the display will make use of icons when the position of
the host vehicle relative to obstacles or other vehicles is
displayed. Occasionally, as the image can be displayed especially
when the object cannot be identified.
A general memory unit 84 which can comprise read-only memory or
random access memory or any combination thereof, is shown. This
memory module, which can be either located at one place or
distributed throughout the system, supplies the information storage
capability for the system.
For advanced RtZF.TM. systems containing the precise positioning
capability, subsystem 86 provides the capability of sending and
receiving information to infrastructure-based precise positioning
tags or devices which may be based on noise or micropower impulse
radar technology, radar reflector or RFIR technology or equivalent.
Once again, the PPS system can also be based on a signature
analysis using the adaptive associative memory technology or
equivalent.
In some locations where weather conditions can deteriorate and
degrade road surface conditions, various infrastructure-based
sensors can be placed either in or adjacent to the road surface.
Subsystem 88 is designed to interrogate and obtained information
from such road-based systems. An example of such a system would be
an RFID tag containing a temperature sensor. This device may be
battery-powered or, preferably, would receive its power from the
vehicle-mounted interrogator, or other host vehicle-mounted source,
as the vehicle passes nearby the device. In this manner, the
vehicle can obtain the temperature of the road surface and receive
advanced warning when the temperature is approaching conditions
which could cause icing of the roadway, for example. An RFID based
on a surface acoustic wave (SAW) device is one preferred example of
such a sensor, see U.S. Pat. No. 6,662,642. An infrared sensor on
the vehicle can also be used to determine the road temperature and
the existence of ice or snow.
In order to completely eliminate automobile accidents, a diagnostic
system is required on the vehicle that will provide advanced
warning of any potential vehicle component failures. Such a system
is described in U.S. Pat. No. 5,809,437 (Breed).
For some implementations of the RtZF.TM. system, stoplights will be
fitted with transmitters which will broadcast a signal when the
light is red. Such a system could make use of the vehicle noise
communication system as described above. This signal can be then
received by a vehicle that is approaching the stoplight provided
that vehicle has the proper sensor as shown as 92. Alternatively, a
camera can be aimed in the direction of stoplights and, since the
existence of the stoplight will be known by the system, as it will
have been recorded on the map, the vehicle will know when to look
for a stoplight and determine the color of the light.
An alternative idea is for the vehicle to broadcast a signal to the
stoplight if, via a camera or other means, it determines that the
light is red. If there are no vehicles coming from the other
direction, the light can change permitting the vehicle to proceed
without stopping. Similarly, if the stoplight has a camera, it can
look in all directions and control the light color depending on the
number of vehicles approaching from each direction. A system of
phasing vehicles can also be devised whereby the speed of
approaching vehicles is controlled so that they interleave through
the intersection and the stoplight may not be necessary.
Although atomic clocks are probably too expensive to the deployed
on automobiles, nevertheless there has been significant advances
recently in the accuracy of clocks to the extent that it is now
feasible to place a reasonably accurate clock as a subsystem 94 to
this system. Since the clock can be recalibrated from each DGPS
transmission, the clock drift can be accurately measured and used
to predict the precise time even though the clock by itself may be
incapable of doing so. To the extent that the vehicle contains an
accurate time source, the satellites in view requirement can
temporarily drop from 4 to 3. An accurate clock also facilitates
the carrier phase DGPS implementations of the system as discussed
above. Additionally, as long as a vehicle knows approximately where
it is on the roadway, it will know its altitude from the map and
thus one less satellite is necessary.
Power must be supplied to the system as shown by power subsystem
96. Certain operator controls are also permitted as illustrated in
subsystem 98.
The control processor or central processor and circuit board
subsystem 100 to which all of the above components 52 98 are
coupled, performs such functions as GPS ranging, DGPS corrections,
image analysis, radar analysis, laser radar scanning control and
analysis of received information, warning message generation, map
communication, vehicle control, inertial navigation system
calibrations and control, display control, precise positioning
calculations, road condition predictions, and all other functions
needed for the system to operate according to design.
A display could be provided for generating and displaying warning
messages which is visible to the driver and/or passengers of the
vehicle. The warning could also be in the form of an audible tone,
a simulated rumble strip and light and other similar ways to
attract the attention of the driver and/or passengers.
Vehicle control also encompasses control over the vehicle to
prevent accidents. By considering information from the map database
48 of the navigation system 46, and the position of the vehicle
obtained via GPS systems, a determination can be made whether the
vehicle is about to run off the road, cross a yellow line and run a
stop sign, as well as the existence or foreseen occurrence of other
potential crash situations. The color of an approaching stoplight
can also be factored in the vehicle control.
FIG. 5A shows a selected reduced embodiment of the accident
avoidance system shown in FIG. 5. The system includes an inertial
reference unit including a plurality of accelerometers and
gyroscopes, namely accelerometers 78A, preferably three of any type
disclosed above, and gyroscopes 80A, preferably three of any type
disclosed above. A clock 94A is provided to obtain a time base or
time reference. This system will accurately determine the motion
(displacement, acceleration and/or velocity) of the vehicle in 6
degrees of freedom (3 displacements (longitudinal, lateral and
vertical)) via the accelerometers 78A and three rotations (pitch,
yaw and roll) via the gyroscopes 80A. As such, along with a time
base from clock 94A, the processor 100A can determine that there
was an accident and precisely what type of accident it was in terms
of the motion of the vehicle (frontal, side, rear and rollover).
This system is different from a crash sensor in that this system
can reside in the passenger compartment of the vehicle where it is
protected from actually being in the accident crush and/or crash
zones and thus it does not have to forecast the accident severity.
It knows the resulting vehicle motion and therefore exactly what
the accident was and what the injury potential is. A typical crash
sensor can get destroyed or at least rotated during the crash and
thus will not determine the real severity of the accident.
Processor 100A is coupled to the inertial reference unit and also
is capable of performing the functions of vehicle control, such as
via control of the brake system 70A, steering system 72A and
velocity sensor 74A, crash sensing, rollover sensing, cassis
control sensing, navigation functions and accident prevention as
discussed herein.
Preferably, a Kalman filter is used to optimize the data from the
inertial reference unit as well as other input sources of data,
signals or information. Also, a neural network, fuzzy logic or
neural-fuzzy system could be used to reduce the data obtained from
the various sensors to a manageable and optimal set. The actual
manner in which a Kalman filter can be constructed and used in the
invention would be left to one skilled in the art. Note that in the
system of the inventions disclosed herein, the extensive
calibration process carried on by other suppliers of inertial
sensors is not required since the system periodically corrects the
errors in the sensors and revises the calibration equation. This in
some cases can reduce the manufacturing cost on the IMU by a factor
of ten.
Further, the information from the accelerometers 78A and gyroscopes
80A in conjunction with the time base or reference is transmittable
via the communication system 56A,58A to other vehicles, possibly
for the purpose of enabling other vehicles to avoid accidents with
the host vehicle, and/or to infrastructure.
One particularly useful function would be for the processor to send
data from, or data derived from, the accelerometers and gyroscopes
relating to a crash, i.e., indicative of the severity of the
accident with the potential for injury to occupants, to a
monitoring location for the dispatch of emergency response
personnel, i.e., an EMS facility or fire station. Other telematics
functions could also be provided.
10.9 Exterior Surveillance System
FIG. 6 is a block diagram of the host vehicle exterior surveillance
system. Cameras 60 are primarily intended for observing the
immediate environment of the vehicle. They are used for recognizing
objects that could be most threatening to the vehicle, i.e.,
closest to the vehicle. These objects include vehicles or other
objects that are in the vehicle blind spot, objects or vehicles
that are about to impact the host vehicle from any direction, and
objects either in front of or behind the host vehicle which the
host vehicle is about to impact. These functions are normally
called blind spot monitoring and collision anticipatory
sensors.
As discussed above, the cameras 60 can use naturally occurring
visible or infrared radiation, or other parts of the
electromagnetic spectrum including terahertz and x-rays, or they
may be supplemented with sources of visible or infrared
illumination from the host vehicle. Note that there generally is
little naturally occurring terahertz radiation other than the
amount that occurs in black body radiation from all sources. The
cameras 60 used are preferably high dynamic range cameras that have
a dynamic range exceeding 60 db and preferably exceeding 100 db.
Such commercially available cameras include those manufactured by
the Photobit Corporation in California and the IMS Chips Company in
Stuttgart Germany. Alternately, various other means exist for
increasing the effective dynamic range through shutter control or
illumination control using a Kerr or Pockel cell, modulated
illumination, external pixel integration etc.
These cameras are based on CMOS technology and can have the
important property that pixels are independently addressable. Thus,
the control processor may decide which pixels are to be read at a
particular time. This permits the system to concentrate on certain
objects of interest and thereby make more effective use of the
available bandwidth.
Video processor printed circuit boards or feature extractor 61 can
be located adjacent and coupled to the cameras 60 so as to reduce
the information transferred to the control processor. The video
processor feature extractor 61 can also perform the function of
feature extraction so that all values of all pixels do not need to
be sent to the neural network for identification processing. The
feature extraction includes such tasks as determining the edges of
objects in the scene and, in particular, comparing and subtracting
one scene from another to eliminate unimportant background images
and to concentrate on those objects which had been illuminated with
infrared or terahertz radiation, for example, from the host
vehicle. By these and other techniques, the amount of information
to be transferred to the neural network is substantially
reduced.
The neural network 63 receives the feature data extracted from the
camera images by the video processor feature extractor 61 and uses
this data to determine the identification of the object in the
image. The neural network 63 has been previously trained on a
library of images that can involve as many as one million such
images. Fortunately, the images seen from one vehicle are
substantially the same as those seen from another vehicle and thus
the neural network 63 in general does not need to be trained for
each type of host vehicle.
As the number of image types increases, modular or combination
neural networks can be used to simplify the system.
Although the neural network 63 has in particular been described,
other pattern recognition techniques are also applicable. One such
technique uses the Fourier transform of the image and utilizes
either optical correlation techniques or a neural network trained
on the Fourier transforms of the images rather than on the image
itself. In one case, the optical correlation is accomplished purely
optically wherein the Fourier transform of the image is
accomplished using diffraction techniques and projected onto a
display, such as a garnet crystal display, while a library of the
object Fourier transforms is also displayed on the display. By
comparing the total light passing through the display, an optical
correlation can be obtained very rapidly. Although such a technique
has been applied to scene scanning by military helicopters, it has
previously not been used in automotive applications.
The laser radar system 64 is typically used in conjunction with a
scanner 65. The scanner 65 typically includes two oscillating
mirrors, or a MEMS mirror capable of oscillating in two dimensions,
which cause the laser light to scan the two-dimensional angular
field. Alternately, the scanner can be a solid-state device
utilizing a crystal having a high index of refraction which is
driven by an ultrasonic vibrator as discussed above or rotating
mirrors. The ultrasonic vibrator establishes elastic waves in the
crystal which diffracts and changes the direction of the laser
light.
The laser beam can be frequency, amplitude, time, code or noise
modulated so that the distance to the object reflecting the light
can be determined. The laser light strikes an object and is
reflected back where it is guided onto a pin diode, or other high
speed photo detector. Since the direction of laser light is known,
the angular location of the reflected object is also known and
since the laser light is modulated the distance to the reflected
point can be determined. By varying modulation frequency of the
laser light, or through noise or code modulation, the distance can
be very precisely measured.
Alternatively, the time-of-flight of a short burst of laser light
can be measured providing a direct reading of the distance to the
object that reflected the light. By either technique, a
three-dimensional map can be made of the surface of the reflecting
object. Objects within a certain range of the host vehicle can be
easily separated out using the range information. This can be done
electronically using a technique called range gating, or it can be
accomplished mathematically based on the range data. By this
technique, an image of an object can be easily separated from other
objects based on distance from the host vehicle.
Since the vehicle knows its position accurately and in particular
it knows the lane on which it is driving, a determination can be
made of the location of any reflective object and in particular
whether or not the reflective object is on the same lane as the
host vehicle. This fact can be determined since the host vehicle
has a map and the reflective object can be virtually placed on that
map to determine its location on the roadway, for example.
The laser radar system will generally operate in the near infrared
part of the electromagnetic spectrum. The laser beam will be of
relatively high intensity compared to the surrounding radiation and
thus even in conditions of fog, snow, and heavy rain, the
penetration of the laser beam and its reflection will permit
somewhat greater distance observations than the human driver can
perceive. Under the RtZF.TM. plan, it is recommended that the speed
of the host vehicle be limited such that vehicle can come to a
complete stop in one half or less of the visibility distance. This
will permit the laser radar system to observe and identify
threatening objects that are beyond the visibility distance, apply
the brakes to the vehicle if necessary causing the vehicle to stop
prior to an impact, providing an added degree of safety to the host
vehicle.
Radar system 62 is mainly provided to supplement laser radar
system. It is particularly useful for low visibility situations
where the penetration of the laser radar system is limited. The
radar system, which is most probably a noise or pseudonoise
modulated continuous wave radar, can also be used to provide a
crude map of objects surrounding the vehicle. The most common use
for automotive radar systems is for adaptive cruise control systems
where the radar monitors the distance and, in some cases, the
velocity of the vehicle immediately in front of the host vehicle.
The radar system 62 is controlled by the control processor 100.
The display system 82 was discussed previously and can be either a
heads up or other appropriate display.
The control processor 100 can be attached to a vehicle special or
general purpose bus 110 for transferring other information to and
from the control processor to other vehicle subsystems.
In interrogating other vehicles on the roadway, a positive
identification of the vehicle and thus its expected properties such
as its size and mass can sometimes be accomplished by laser
vibrometry. By this method, a reflected electromagnetic wave can be
modulated based on the vibration that the vehicle is undergoing.
Since this vibration is caused at least partially by the engine,
and each class of engine has a different vibration signature, this
information can be used to identify the engine type and thus the
vehicle. This technique is similar to one used to identify enemy
military vehicles by the U.S. military. It is also used to identify
ships at sea using hydrophones. In the present case, a laser beam
is directed at the vehicle of interest and the returned reflected
beam is analyzed such as with a Fourier transform to determine the
frequency makeup of the beam. This can then be related to a vehicle
to identify its type either through the use of a look-up table or
neural network or other appropriate method. This information can
then be used as information in connection with an anticipatory
sensor as it would permit a more accurate estimation of the mass of
a potentially impacting vehicle.
Once the vehicle knows where it is located, this information can be
displayed on a heads-up display and if an occupant sensor has
determined the location of the eyes of the driver, the road edges,
for example, and other pertinent information from the map database
can be displayed exactly where they would be seen by the driver.
For the case of driving in dense fog or on a snow covered road, the
driver will be able to see the road edges perhaps exactly or even
better than the real view, in some cases. Additionally, other
information gleaned by the exterior monitoring system can show the
operator the presence of other vehicles and whether they represent
a threat to the host vehicle (see for example "Seeing the road
ahead", GPS World Nov. 1, 2003, which describes a system
incorporating many of the current assignee's invention ideas
described herein).
10.10 Corridors
FIG. 7 shows the implementation of the invention in which a vehicle
18 is traveling on a roadway in a defined corridor in the direction
X. Each corridor is defined by lines 14. If the vehicle is
traveling in one corridor and strays in the direction Y so that it
moves along the line 22, e.g., the driver is falling asleep, the
system on board the vehicle in accordance with the invention will
activate a warning. More specifically, the system continually
detects the position of the vehicle, such as by means of the GPS,
DGPS and/or PPS, and has the locations of the lines 14 defining the
corridor recorded in its map database. Upon an intersection of the
position of the vehicle and one of the lines 14 as determined by a
processor, the system may be designed to sound an alarm to alert
the driver to the deviation or possibly even correct the steering
of the vehicle to return the vehicle to within the corridor defined
by lines 14.
FIG. 8 shows the implementation of the invention in which a pair of
vehicles 18, 26 are traveling on a roadway each in a defined
corridor defined by lines 14 and each equipped with a system in
accordance with the invention. The system in each vehicle 18, 26
will receive data informing it of the position of the other vehicle
and prevent accidents from occurring, e.g., if vehicle 18 moves in
the direction of arrow 20. This can be accomplished via direct
wireless broadband communication or any of the other communication
methods described above, or through another path such as via the
Internet or through a base station, wherein each vehicle transmits
its best estimate of its absolute location on the earth along with
an estimate of the accuracy of this location. If one vehicle has
recently passed a precise positioning station, for example, then it
will know its position very accurately to within a few centimeters.
Each vehicle can also send the latest satellite messages that it
received, permitting each vehicle to precisely determine its
relative location to the other since the errors in the signals will
be the same for both vehicles. To the extent that both vehicles are
near each other, even the carrier phase ambiguity can be determined
and each vehicle will know its position relative to the other to
within better than a few centimeters. As more and more vehicles
become part of the community and communicate their information to
each other, each vehicle can even more accurately determine its
absolute position and especially if one vehicle knows its position
very accurately, if it recently passed a PPS for example, then all
vehicles will know their position with approximately the same
accuracy and that accuracy will be able to be maintained for as
long as a vehicle keeps its lock on the satellites in view. If that
lock is lost temporarily, the INS system will fill in the gaps and,
depending on the accuracy of that system, the approximate 2
centimeter accuracy can be maintained even if the satellite lock is
lost for up to approximately five minutes.
A five minute loss of satellite lock is unlikely except in tunnels
or in locations where buildings or geological features interfere
with the signals. In the building case, the problem can be
eliminated through the placement of PPS stations, or through
environmental signature analysis, and the same would be true for
the geological obstruction case except in remote areas where ultra
precise positioning accuracy is probably not required. In the case
of tunnels, for example, the cost of adding PPS stations is
insignificant compared with the cost of building and maintaining
the tunnel.
10.11 Vehicle Control
FIG. 12a is a flow chart of the method in accordance with the
invention. The absolute position of the vehicle is determined at
130, e.g., using a GPS, DGPS PPS system, and compared to the edges
of the roadway at 134, which is obtained from a memory unit 132.
Based on the comparison at 134, it is determined whether the
absolute position of the vehicle is approaching close to or
intersects an edge of the roadway at 136. If not, then the position
of the vehicle is again obtained, e.g., at a set time interval
thereafter, and the process continues. If yes, an alarm and/or
warning system will be activated or the system will take control of
the vehicle (at 140) to guide it to a shoulder of the roadway or
other safe location.
FIG. 12b is another flow chart of the method in accordance with the
invention similar to FIG. 12a. Again the absolute position of the
vehicle is determined at 130, e.g., using a GPS, DGPS PPS system,
and compared to the location of a roadway yellow line at 142 (or
possibly another line which indicates an edge of a lane of a
roadway), which is obtained from a memory unit 132. Based on the
comparison at 144, it is determined whether the absolute position
of the vehicle is approaching close to or intersects the yellow
line 144. If not, then the position of the vehicle is again
obtained, e.g., at a set time interval thereafter, and the process
continues. If yes, an alarm will sound and/or the system will take
control of the vehicle (at 146) to control the steering or guide it
to a shoulder of the roadway or other safe location.
FIG. 12c is another flow chart of the method in accordance with the
invention similar to FIG. 12a. Again the absolute position of the
vehicle is determined at 130, e.g., using a GPS, DGPS PPS system,
and compared to the location of a roadway stoplight at 150, which
is obtained from a memory unit 132. Based on the comparison at 150,
it is determined whether the absolute position of the vehicle is
approaching close to a stoplight. If not, then the position of the
vehicle is again obtained, e.g., at a set interval thereafter, and
the process continues. If yes, a sensor determines whether the
stoplight is red (e.g., a camera) and if so, an alarm will sound
and/or the system will take control of the vehicle (at 154) to
control the brakes or guide it to a shoulder of the roadway or
other safe location. A similar flow chart can be now drawn by those
skilled in the art for other conditions such as stop signs, vehicle
speed control, collision avoidance etc.
10.12 Intersection Collision Avoidance
FIG. 13 illustrates an intersection of a major road 170 with a
lesser road 172. The road 170 has the right of way and stop signs
174 have been placed to control the traffic on the lesser road 172.
Vehicles 18 and 26 are proceeding on road 172 and vehicle 25 is
proceeding on road 170. A very common accident is caused when
vehicle 18 ignores the stop sign 174 and proceeds into the
intersection where it is struck on the side by vehicle 25 or
strikes vehicle 25 on the side.
Using the teachings of this invention, vehicle 18 will know of the
existence of the stop sign and if the operator attempts to proceed
without stopping, the system will sound a warning and if that
warning is not heeded, the system will automatically bring the
vehicle 18 to a stop preventing it from intruding into the
intersection.
Another common accident is where vehicle 18 does in fact stop but
then proceeds forward without noticing vehicle 25 thereby causing
an accident. Since in the fully deployed RtZF.TM. system, vehicle
18 will know through the vehicle-to-vehicle communication the
existence and location of vehicle 25 and can calculate its
velocity, the system can once again take control of vehicle 18 if a
warning is not heeded and prevent vehicle 18 from proceeding into
the intersection and thereby prevent the accident.
In the event that the vehicle 25 is not equipped with the RtZF.TM.
system, vehicle 18 will still sense the presence of vehicle 25
through the laser radar, radar and camera systems. Once again, when
the position and velocity of vehicle 25 is sensed, appropriate
action can be taken by the system in vehicle 18 to eliminate the
accident.
In another scenario where vehicle 18 properly stops at the stop
sign, but vehicle 26 proceeds without observing the presence of the
stopped vehicle 18, the laser radar, radar and camera systems will
all operate to warn the driver of vehicle 26 and if that warning is
not heeded, the system in vehicle 26 will automatically stop the
vehicle 26 prior to its impacting vehicle 18. Thus, in the
scenarios described above the "Road to Zero Fatalities".TM. system
and method of this invention will prevent common intersection
accidents from occurring.
FIG. 14 is a view of an intersection where traffic is controlled by
stoplights 180. If the vehicle 18 does not respond in time to a red
stoplight, the system as described above will issue a warning and
if not heeded, the system will take control of the vehicle 18 to
prevent it from entering the intersection and colliding vehicle 25.
In this case, the stoplight 180 will emit a signal indicating its
color, such as by way of the communication system, and/or vehicle
18 will have a camera mounted such that it can observe the color of
the stoplight. There are of course other information transfer
methods such as through the Internet. In this case, buildings 182
obstruct the view from vehicle 18 to vehicle 25 thus an accident
can still be prevented even when the operators are not able to
visually see the threatening vehicle. If both vehicles have the
RtZF.TM. system, they will be communicating and their presence and
relative positions will be known to both vehicles.
FIG. 15 illustrates the case where vehicle 18 is about to execute a
left-hand turn into the path of vehicle 25. This accident will be
prevented if both cars have the RtZF.TM. system since the locations
and velocities of both vehicles 18, 25 will be known to each other.
If vehicle 25 is not equipped and vehicle 18 is, then the camera,
radar, and laser radar subsystems will operate to prevent vehicle
18 from turning into the path of vehicle 25. Once again, common
intersection accidents are prevented by this invention.
The systems described above can be augmented by
infrastructure-based sensing and warning systems. Camera, laser or
terahertz radar or radar subsystems such as placed on the vehicle
can also be placed at intersections to warn the oncoming traffic if
a collision is likely to occur. Additionally, simple sensors that
sense the signals emitted by oncoming vehicles, including radar,
thermal radiation, etc., can be used to operate warning systems
that notify oncoming traffic of potentially dangerous situations.
Thus, many of the teachings of this invention can be applied to
infrastructure-based installations in addition to the
vehicle-resident systems.
10.13 Privacy
People do not necessarily want the government to know where they
are going and therefore will not want information to be transmitted
that can identify the vehicle. The importance of this issue may be
overestimated. Most people will not object to this minor infraction
if they can get to their destination more efficiently and
safely.
On the other hand, it has been estimated that there are 100,000
vehicles on the road, many of them stolen, where the operators do
not want the vehicle to be identified. If an identification process
that positively identifies the vehicle were made part of this
system, it could thus cut down on vehicle theft. Alternately,
thieves might attempt to disconnect the system thereby defeating
the full implementation of the system and thus increasing the
danger on the roadways and defeating the RtZF.TM. objective. The
state of the system would therefore need to be self-diagnosed and
system readiness must be a condition for entry onto the restricted
lanes.
11. Other Features
11.1 Incapacitated Driver
As discussed herein, the RtZF.TM. system of this invention also
handles the problem of the incapacitated driver thus eliminating
the need for sleep sensors that appear in numerous U.S. patents.
Such systems have not been implemented because of their poor
reliability. The RtZF.TM. system senses the result of the actions
of the operator, which could occur for a variety of reasons
including inattentiveness cause by cell phone use, old age,
drunkenness, heart attacks, drugs as well as falling asleep.
11.2 Emergencies--Car Jacking, Crime
Another enhancement that is also available is to prevent car
jacking in which case the RtZF.TM. system can function like the
Lojack.TM. system. In the case where a car-jacking occurs, the
location of the vehicle can be monitored and if an emergency button
is pushed, the location of the vehicle with the vehicle ID can be
transmitted.
11.3 Headlight Dimmer
The system also solves the automatic headlight dimmer problem.
Since the RtZF.TM. system equipped vehicle knows where all other
RtZF.TM. system equipped vehicles are located in its vicinity, it
knows when to dim the headlights. Since it is also interrogating
the environment in front of the vehicle, it also knows the
existence and approximate location of all non-RtZF.TM. system
equipped vehicles. This is one example of a future improvement to
the system. The RtZF.TM. system is a system which lends itself to
continuous improvement without having to change systems on an
existing vehicle.
11.4 Rollover
It should be obvious from the above discussion that rollover
accidents should be effectively eliminated by the RtZF.TM. system.
In the rare case where one does occur, the RtZF.TM. system has the
capability to sense that event since the location and orientation
of the vehicle is known.
For large trucks that have varying inertial properties depending on
the load that is being hauled, sensors can be placed on the vehicle
that measure angular and linear acceleration of a part of the
vehicle. Since the geometry of the road is known, the inertial
properties of the vehicle with load can be determined and thus the
tendency of the vehicle to roll over can be determined. Since the
road geometry is known the speed of the truck can be limited based
partially on its measured inertial properties to prevent rollovers.
The IMU can play a crucial role here in that the motion of the
vehicle is now accurately known to a degree previously not possible
before the Kalman filter error correction system was employed. This
permits more precise knowledge and thus the ability to predict the
motion of the vehicle. The IMU can be input to the chassis control
system and, through appropriate algorithms, the throttle, steering
and brakes can be appropriately applied to prevent a rollover. When
the system described herein is deployed, rollovers should disappear
as the causes such as road ice, sharp curves and other vehicles are
eliminated.
If a truck or other vehicle is driving on a known roadway where the
vertical geometry (height and angle) has been previously determined
and measured then one or more accelerometers and gyroscopes can be
placed at appropriate points on the truck and used to measure the
response of the vehicle to the disturbance. From the known input
and measured response, the inertial properties of the vehicle can
readily be determined by one skilled in the art. Similarly, if
instead of a knowledge of the road from the map database, the input
to the vehicle from the road can be measured by accelerometers and
gyroscopes placed on the chassis, for example, and then the forcing
function into the truck body is known and by measuring the motion
(accelerations and angular accelerations) the inertial properties
once again can be determined. Finally, the input from the road can
be treated statistically and again the inertial properties of the
truck estimated. If a truck tractor is hauling a trailer then the
measuring devices can be placed at convenient locations of the
trailer such inside the trailer adjacent to the roof at the front
and rear of the trailer.
If the map contains the information, then as the vehicle travels
the road and determines that there has been a change in the road
properties this fact can be communicated via telematics or other
methods to the map maintenance personnel, for example. In this
manner, the maps are kept current and pothole or other evidence of
road deterioration can be sent to appropriate personnel for
attention.
Once the system determines that the vehicle is in danger or a
rollover situation; the operator can be notified with an audible or
visual warning (via a display or light) so that he or she can take
corrective action. Additionally or alternately, the system can take
control of the situation and prevent the rollover through
appropriate application of brakes (either on all wheels or
selectively on particular wheels), throttle or steering.
11.5 Vehicle Enhancements
The RtZF.TM. system can now be used to improve the accuracy of
other vehicle based instruments. The accuracy of the odometer and
yaw rate sensors can be improved over time, for example, by
regression, or through the use of a Kalman filter, against the DGPS
data. The basic RtZF.TM. system contains an IMU which comprises
three accelerometers and three gyroscopes. This system is always
being updated by the DGPS system, odometer, vehicle speed sensor,
magnetic field and field vector sensors, PPS and other available
sensors through a Kalman filter and in some cases a neural
network.
11.6 Highway Enhancements
Enhancements to the roadways that result from the use of the
RtZF.TM. system include traffic control. The timing of the
stoplights can now be automatically adjusted based on the relative
traffic flow. The position of every vehicle within the vicinity of
the light can be known from the communication system discussed
above. When all vehicles have the RtZF.TM. system, many stoplights
will no longer be necessary since the flow of traffic through an
intersection can be accurately controlled to avoid collisions.
Since the road conditions will now be known to the system, an
enhanced RtZF.TM. system will be able to advise an operator not to
travel or, alternately, it can pick an alternate route if certain
roads have accidents or have iced over, for example. Some people
may decide not drive if there is bad weather or congestion. The
important point here is that sensors will be available to sense the
road condition as to both traffic and weather, this information
will be available automatically and not require reporting from
weather stations which usually have only late and inaccurate
information. Additionally, pricing for the use of certain roads can
be based on weather, congestion, time of day, etc. That is, pricing
can by dynamically controlled.
The system lends itself to time and congestion based allocation of
highway facilities. A variable toll can automatically be charged to
vehicles based on such considerations since the vehicle can be
identified. In fact, automatic toll systems now being implemented
will likely become obsolete as will all toll booths.
Finally, it is important to recognize that the RtZF.TM. system is
not a "sensor fusion" system. Sensor fusion is based on the theory
that you can take inputs from different sensors and combine them in
such a way as to achieve more information from the combined sensors
than from treating the sensor outputs independently in a
deterministic manner. The ultimate sensor fusion system is based on
artificial neural networks, sometimes combined with fuzzy logic to
form a neural fuzzy system. Such systems are probabilistic. Thus
there will always be some percentage of cases where the decision
reached by the network will be wrong. The use of such sensor
fusion, therefore, is inappropriate for the "Zero Fatalities" goal
of the invention, although several of the sub-parts of the system
may make use of neural networks.
11.7 Speed Control
Frequently a driver is proceeding down a road without knowing the
allowed speed limit. This can happen if he or she recently entered
a road and a sign has not been observed or perhaps the driver just
was not paying attention or the sign was hidden from view by
another vehicle. If the allowed speed was represented in the map
database then it could be displayed on an in vehicle display since
the vehicle would know its location.
12. Summary
In sum, disclosed above is a computer controlled vehicle and
obstacle location system and method which includes the steps of
receiving continuously from a network of satellites on a first
communication link at one of a plurality of vehicles, GPS ranging
signals for initially accurately determining, in conjunction with
centimeter accurate maps, the host vehicle's position on a roadway
on a surface of the earth; receiving continuously at the host
vehicle on a second communication link from a station, another
vehicle or satellite, DGPS auxiliary range correction signals for
correcting propagation delay errors in the GPS ranging signals;
determining continuously at the host vehicle from the GPS, DGPS,
and accurate map database signals host vehicle's position on the
surface of the earth with centimeter accuracy; communicating the
host vehicle's position to another one of the plurality of
vehicles, and receiving at the host vehicle, location information
from at least one of a plurality of other vehicles; determining
whether the other vehicle represents a collision threat to the host
vehicle based on its position relative to the roadway and the host
vehicle and generating a warning or vehicle control signal response
to control the vehicles motion laterally or longitudinally to
prevent a collision with the other vehicle. In some
implementations, the detecting step includes detecting objects by
scanning with one or more cameras, radars or laser radars located
on the host vehicle. The analyzing step includes processing and
analyzing digital signals indicative of video images detected by
the one or more cameras, radars or laser radars, and processing and
analyzing the digital signals using pattern recognition and range
determination algorithms. The objects detected may include fixed or
moving, or known or unknown obstacles, people, bicycles, animals,
or the like.
An optional feature of this embodiment of the invention is to
operate one or more of the following systems depending on the kind
of response determined by the neural fuzzy logic control system: a
brake pedal, accelerator pedal, steering system (e.g., steering
wheel), horn, light, mirror, defogger and communication
systems.
The first phase of implementation of this invention can be
practiced with only minor retrofit additions to the vehicle. These
include the addition of a differential GPS system, an inertial
measurement unit (IMU) and appropriate circuitry, and an accurate
map database. In this first phase, the driver will only be warned
when he or she is about to depart from the road surface. During the
second phase of practicing this invention, the system will be
augmented with a system that will prevent the operator from leaving
the assigned corridor and in particular leaving the road at high
speed. In further phases of the implementation of this invention,
additional systems will be integrated which will scan the roadway
and act to prevent accidents with vehicles that do not have the
system installed. Also communication systems will be added to
permit the subject vehicle to communicate its position, velocity,
etc., to other nearby vehicles that are also equipped with a
system. This communication system is the main focus herein.
A primary preferred embodiment of the system, therefore, is to
equip a vehicle with a DGPS system, an inertial guidance system (or
IMU), vehicle steering, throttle and brake control apparatus, a
sub-meter accurate digital map system with the relevant maps (or
ability to access the relevant maps), a scanning pulsed infrared
laser radar, a system for sensing or receiving signals from a
highway-based precise position determination system, and
communications systems for (1) sending and receiving data from
similarly equipped vehicles, (2) receiving updated maps and map
status information, and (3) receiving weather and road condition
information. A preferred embodiment for the infrastructure
enhancements includes a DGPS system, a radar reflector based, Radio
Frequency Identification (RFID) based or equivalent precise
position determining system and local weather and road condition
determination and transmission system.
Also disclosed above are methods and apparatus for preventing
vehicle accidents. To this end, a vehicle is equipped with a
differential GPS (DGPS) navigational system as well as an inertial
navigation subsystem. Part of the system can be an array of
infrastructure stations that permit the vehicle to exactly
determine its position at various points along its path. Such
stations would typically be located at intervals such as every 50
miles along the roadway, or more or less frequently depending on
requirements as described below. These stations permit the vehicle
to become its own DGPS station and thus to correct for the GPS
errors and to set the position of the vehicle based initial
guidance system. It also provides sufficient information for the
vehicle to use the carrier frequency to determine its absolute
position to within a few centimeters or better for as long as
satellite locks are maintained. Data is also available to the
vehicle that provides information as to the edges of the roadway,
and edges of the lanes of the roadway, at the location of the
vehicle so that the vehicle control system can continuously
determine its location relative to the roadway edges and/or lane
edges. In the initial implementation, the operator operates his or
her vehicle and is unaware of the presence of the accident
avoidance system. If, however, the operator falls asleep or for
some other reason attempts to drive off the roadway at high speed,
the system will detect that the vehicle is approaching an edge of
the roadway and will either sound an alarm or prevent the vehicle
from leaving the roadway when doing so would lead to an accident.
In some cases, the system will automatically reduce the speed of
the vehicle and stop it on the shoulder of the roadway.
It is important to note that the invention as described in the
above paragraph is in itself a significant improvement to
automotive safety. Approximately half of all fatal accidents
involve only a single vehicle that typically leaves the roadway and
impacts with a roadside obstacle, cross a yellow line or run a red
light or stop sign. This typically happens when the driver in under
the influence of alcohol or drugs, has a medical emergency or
simply falls asleep. If this cause of accidents could be
eliminated, the potential exists for saving many thousands of
deaths per year when all vehicles are equipped with the system of
this invention. This would make this the single greatest advance in
automotive safety surpassing both seatbelts and airbags in
lifesaving potential.
A first improvement to this embodiment of the invention is to
provide the vehicle with a means using radar, laser radar, optical
or infrared imaging, or a similar technology, to determine the
presence, location and velocity of other vehicles on the roadway
that are not equipped with the accident avoidance system. The
accident avoidance system (RtZF.TM.) of this invention will not be
able to avoid all accidents with such vehicles for the reasons
discussed above, but will be able to provide a level of protection
which is believed to surpass all known prior art systems. Some
improvement over prior art systems will result from the fact that
the equipped vehicle knows the location of the roadway edges, as
well as the lane boundaries, not only at the location of the
equipped vehicle but also at the location of the other nearby
vehicles. Thus, the equipped vehicle will be able to determine that
an adjacent vehicle has already left its corridor and warn the
driver or initiate evasive action. In prior art systems, the
location of the roadway is not known leading to significantly less
discrimination ability.
A second improvement is to provide communication ability to other
nearby similarly equipped vehicles permitting the continuous
transmission and reception of the locations of all equipped
vehicles in the vicinity. With each vehicle knowing the location,
and thus the velocity, of all potential impacting vehicles which
are equipped with the RtZF, collisions between vehicles can be
reduced and eventually nearly eliminated when all vehicles are
equipped with the RtZF. One such communication system involves the
use of spread spectrum carrier less communication channels that
make efficient use of the available bandwidth and permit the
simultaneous communication of many vehicles.
A third improvement comprises the addition of software to the
system that permits vehicles on specially designated vehicle
corridors for the operator to relinquish control of the vehicle to
the vehicle-based system, and perhaps to a roadway computer system.
This then permits vehicles to travel at high speeds in a close
packed formation thereby substantially increasing the flow rate of
vehicles on a given roadway. Naturally, in order to enter the
designated corridors, a vehicle would be required to be equipped
with the RtZF. Similarly, this then provides an incentive to
vehicle owners to have their vehicles so equipped so that they can
enter the controlled corridors and thereby shorten their travel
time. Close packed or platooning travel is facilitated in the
invention and thus supportive of the drag reduction advantages of
such travel. But, such travel, although it can be automatically
achieved through implementation of the proper algorithms in a very
simple manner, is not required.
Prior art systems require expensive modifications to highways to
permit such controlled high speed close packed travel. Such
modifications also require a substantial infrastructure to support
the system. The RtZF.TM. of the present invention, in its simplest
form, does not require any modification to the roadway but rather
relies primarily on the GPS or similar satellite system or other
precise locating system. The edge and lane boundary information is
either present within the vehicle RtZF.TM. memory or transmitted to
the vehicle as it travels along the road. The permitted speed of
travel is also communicated to the vehicles on the restricted
corridor and thus each vehicle travels at the appointed speed.
Since each vehicle knows the location of all other vehicles in the
vicinity, should one vehicle slow down, due to an engine
malfunction, for example, appropriate action can be taken to avoid
an accident. Vehicles do not need to travel in groups as suggested
and required by some prior art systems. Rather, each vehicle may
independently enter the corridor and travel at the system defined
speed until it leaves, which may entail notifying the system of a
destination.
Another improvement involves the transmission of additional data
concerning weather conditions, road conditions traffic accidents
etc. to the equipped vehicle so that the speed of that vehicle can
be limited to a safe speed depending on road conditions, for
example. If moisture is present on the roadway and the temperature
is dropping to the point that ice might be building up on the road
surface, the vehicle can be notified by the roadway information
system and prevented from traveling at an unsafe speed.
In contrast to some prior art systems, with the RtZF.TM. system in
accordance with the invention, especially when all vehicles are
appropriately equipped, automatic braking of the vehicle should
rarely be necessary and steering and throttle control should in
most cases be sufficient to prevent accidents. In most cases,
braking means the accident wasn't anticipated.
It is important to understand that this is a process control
problem. The process is designed so that it should not fail and
thus all accidents should be eliminated. Events that are
troublesome to the system include a deer running in front of the
vehicle, a box falling off of a truck, a rock rolling onto the
roadway and a catastrophic failure of a vehicle. Continuous
improvement to the process is thus required before these events are
substantially eliminated. Each vehicle, individual driver and
vehicle control system is part of the system and upon observing
that such an event has occurred, he or she should have the option
of stopping the process to prevent or mitigate an emergency. All
equipped vehicles therefore have the capability of communicating
that the process is stopped and therefore that the vehicle speed,
for example, should be substantially reduced until the vehicle has
passed the troubled spot or until the problem ceases to exist. In
other words, each vehicle and each driver is part of the process.
In one manner, each vehicle is a probe vehicle.
The RtZF.TM. system in accordance with the invention will thus
start simple by reducing single vehicle accidents and evolve. The
system has the capability to solve the entire problem by
eliminating automobile accidents.
Furthermore, disclosed above are methods and apparatus for
eliminating accidents by accurately determining the position of a
vehicle, accurately knowing the position of the road and
communicating between vehicles and between the vehicle and the
infrastructure supporting travel. People get into accidents when
they go too fast for the conditions and when they get out of their
corridor. This embodiment eliminates these and other causes of
accidents. In multilane highways, this system prevents people from
shifting lanes if there are other vehicles in the blind spot, thus,
solving the blind spot problem. The vehicle would always be
traveling down a corridor where the width of the corridor may be a
lane or the entire road width or something in between depending on
road conditions and the presence of other vehicles. This embodiment
is implemented through the use of both an inertial navigation
system (INS) and a DGPS, in some cases with carrier frequency
enhancement. Due to the fact that the signals from at least four
GPS or GLONASS satellites are not always available and to errors
caused by multiple path reception from a given satellite, the DGPS
systems cannot be totally relied upon. Therefore the INS is a
critical part of the system. This will improve as more satellites
are launched and additional ground stations are added. It will also
significantly improve when the WAAS and LAAS systems are
implemented and refined to work with land vehicles as well as
airplanes. It will also be improved with the implementation of
PPS.
Also disclosed above is a method for transferring information
between a vehicle and a transmitter which comprises the steps of
transmitting a unique pseudorandom noise signal by the transmitter
in a carrier-less fashion composed of frequencies within a
pre-selected band, encoding information in the noise or
pseudo-noise signal relating to an identification of the
transmitter and a position of the transmitter and providing the
vehicle with means for extracting the information from the noise or
pseudo-noise signal. The code to use for encoding the noise or
pseudo-noise signal may be selected based on the position of the
transmitter so that analysis of the code, or a portion thereof,
provides an indication of the position of the transmitter.
Information about accidents, weather conditions, road conditions,
map data and traffic control devices and about errors in a GPS
signal can also be encoded in the noise or pseudo-noise signals.
The information may be encoded in the noise or pseudo-noise signal
in various ways, including but not limited to phase modulation of
distance or time between code transmissions, phase or amplitude
modulation of the code sequences, changes of the polarity of the
entire code sequence or the individual code segments, or bandwidth
modulation of the code sequence. The information may be encoded in
the noise or pseudo-noise signal sequentially from general
information to specific information about the position of the
transmitter, e.g., from the country in which the transmitter is
positioned to the actual square meter in which the transmitter is
located. The transmitter may be arranged in a moving object such as
a vehicle to provide vehicle-to-vehicle communications, in which
case, the velocity and optionally direction of travel of the
vehicle is also encoded in the noise or pseudo-noise signal, or at
a fixed location. In the latter case, the location can be used to
correct GPS signals. In this regard, the information encoded in the
noise or pseudo-noise signal may be the GPS coordinate location of
the transmitter.
In a related arrangement, an antenna is arranged on the vehicle to
receive noise or pseudo-noise signals and a processor is coupled to
the antenna. The processor may be constructed or programmed to
analyze the received noise or pseudo-noise signals in order to
determine whether any received noise or pseudo-noise signals
originate from transmitters within a pre-determined distance from
the vehicle. Such analysis can be based on an initial portion of
the noise or pseudo-noise signals, i.e., the processor can scan
through multiple the noise or pseudo-noise signals reading only the
initial part of each to assess which noise or pseudo-noise
signal(s) is/are particularly important and then obtain and process
only those of interest.
While the invention has been illustrated and described in detail in
the drawings and the foregoing description, the same is to be
considered as illustrative and not restrictive in character, it
being understood that only the preferred embodiments have been
shown and described and that all changes and modifications that
come within the spirit of the invention are desired to be
protected.
This application is one in a series of applications covering safety
and other systems for vehicles and other uses. The disclosure
herein goes beyond that needed to support the claims of the
particular invention that is claimed herein. This is not to be
construed that the inventors are thereby releasing the unclaimed
disclosure and subject matter into the public domain. Rather, it is
intended that patent applications have been or will be filed to
cover all of the subject matter disclosed above.
* * * * *