U.S. patent application number 16/734945 was filed with the patent office on 2020-05-07 for systems and methods for controlling autonomous vehicles that provide a vehicle service to users.
The applicant listed for this patent is UATC, LLC. Invention is credited to Michael Aitken, Richard Brian Donnelly.
Application Number | 20200142428 16/734945 |
Document ID | / |
Family ID | 64401112 |
Filed Date | 2020-05-07 |
![](/patent/app/20200142428/US20200142428A1-20200507-D00000.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00001.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00002.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00003.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00004.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00005.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00006.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00007.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00008.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00009.png)
![](/patent/app/20200142428/US20200142428A1-20200507-D00010.png)
View All Diagrams
United States Patent
Application |
20200142428 |
Kind Code |
A1 |
Donnelly; Richard Brian ; et
al. |
May 7, 2020 |
Systems and Methods for Controlling Autonomous Vehicles that
Provide a Vehicle Service to Users
Abstract
Systems and methods for controlling autonomous vehicles are
provided. In one example embodiment, a computer implemented method
includes obtaining data indicative of a location associated with a
user to which an autonomous vehicle is to travel. The autonomous
vehicle is to travel along a first vehicle route that leads to the
location. The method includes obtaining traffic data associated
with a geographic area that includes the location. The method
includes determining an estimated traffic impact of the autonomous
vehicle on the geographic area based at least in part on the
traffic data. The method includes determining vehicle action(s)
based at least in part on the estimated traffic impact and causing
the autonomous vehicle to perform the vehicle action(s) that
include at least one of stopping the autonomous vehicle at least
partially in a travel way within a vicinity of the location or
travelling along a second vehicle route.
Inventors: |
Donnelly; Richard Brian;
(Pittsburgh, PA) ; Aitken; Michael; (Pittsburgh,
PA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
UATC, LLC |
San Francisco |
CA |
US |
|
|
Family ID: |
64401112 |
Appl. No.: |
16/734945 |
Filed: |
January 6, 2020 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15662314 |
Jul 28, 2017 |
10528059 |
|
|
16734945 |
|
|
|
|
62510515 |
May 24, 2017 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G08G 1/0145 20130101;
G08G 1/09623 20130101; G01C 21/3438 20130101; G08G 1/09626
20130101; B60R 25/24 20130101; G08G 1/0137 20130101; G08G 1/096827
20130101; G08G 1/123 20130101; G08G 1/147 20130101; B60R 16/037
20130101; G01C 21/3492 20130101; G08G 1/096791 20130101; G08G 1/09
20130101; G08G 1/096725 20130101; H04W 4/46 20180201; G05D 1/0276
20130101; G08G 1/205 20130101; G08G 1/202 20130101; G01C 21/3676
20130101; G08G 1/0129 20130101; G08G 1/0112 20130101; G08G 1/005
20130101; H04W 4/40 20180201; G08G 1/143 20130101; G05D 1/0088
20130101; G08G 1/096775 20130101 |
International
Class: |
G05D 1/02 20060101
G05D001/02; H04W 4/46 20060101 H04W004/46; G08G 1/123 20060101
G08G001/123; G08G 1/09 20060101 G08G001/09; B60R 25/24 20060101
B60R025/24; B60R 16/037 20060101 B60R016/037; G08G 1/00 20060101
G08G001/00; G08G 1/14 20060101 G08G001/14; G08G 1/0968 20060101
G08G001/0968; G08G 1/0967 20060101 G08G001/0967; G08G 1/0962
20060101 G08G001/0962; G08G 1/01 20060101 G08G001/01; G08G 1/005
20060101 G08G001/005; G01C 21/34 20060101 G01C021/34; G01C 21/36
20060101 G01C021/36; G05D 1/00 20060101 G05D001/00 |
Claims
1.-20. (canceled)
21. A computer-implemented method of controlling autonomous
vehicles, comprising: obtaining, by a vehicle computing system that
comprises one or more computing devices, data indicative of a
location to which an autonomous vehicle is to travel; obtaining, by
the computing system via one or more sensors of the autonomous
vehicle, sensor data associated with a surrounding environment of
the autonomous vehicle; determining, by the computing system, a
level of traffic associated with a geographic area that includes
the location to which the autonomous vehicle is to travel based at
least in part on the sensor data; determining, by the vehicle
computing system, an estimated traffic impact of the autonomous
vehicle based at least in part on the level of traffic that is
based at least in part on the sensor data, wherein the estimated
traffic impact is indicative of an estimated impact of the
autonomous vehicle on one or more objects within the surrounding
environment of the autonomous vehicle in the event that the
autonomous vehicle were to stop; and causing, by the vehicle
computing system, the autonomous vehicle to perform one or more
vehicle actions based at least in part on the estimated traffic
impact.
22. The computer-implemented method of claim 21, wherein the
operations further comprise: obtaining, by the vehicle computing
system, location data associated with a user device; determining,
by the vehicle computing system, an estimated time of user arrival
based at least in part on the location data; and determining, by
the vehicle computing system, the one or more vehicle actions based
at least in part on the estimated time of user arrival.
23. The computer-implemented method of claim 21, further
comprising: determining, by the vehicle computing system, that a
parking location that is out of a travel way is unavailable for the
autonomous vehicle; and determining, by the vehicle computing
system, the one or more vehicle actions based at least in part on
the determination that the parking location is unavailable.
24. The computer-implemented method of claim 21, wherein the one or
more vehicle actions comprise at least one of decelerating the
autonomous vehicle, stopping at least partially in a travel way, or
traveling along a route within the geographic area to return to a
location associated with a user.
25. The computer-implemented method of claim 21, wherein the one or
more vehicle actions comprise the vehicle stopping at least
partially within a travel way for a stopping time duration and
circling a block within the geographic area after expiration of the
stopping time duration.
26. The computer-implemented method of claim 24, wherein stopping
at least partially in the travel way comprises stopping at least
partially within the travel way within a vicinity of a location
associated with a user.
27. The computer-implemented method of claim 25, wherein the
vicinity of the location associated with the user is defined at
least in part by an acceptable walking distance from the location
associated with the user.
28. The computer-implemented method of claim 26, wherein the user
is associated with a request for a vehicle service and wherein the
autonomous vehicle is assigned to perform the vehicle service.
29. The computer-implemented method of claim 28, wherein the
vehicle service comprises at least one of transportation of a
passenger or transportation of an item.
30. The computer-implemented method of claim 21, wherein
determining the estimated traffic impact comprises: comparing, by
the vehicle computing system, the level of traffic to a threshold
level of traffic.
31. The computer-implemented method of claim 30, wherein the
threshold level of traffic is determined at least in part from a
machine-learned model.
32. A computing system comprising: one or more processors; and one
or more memory devices, the one or more memory devices storing
instructions that when executed by the one or more processors cause
the computing system to perform operations, the operations
comprising: obtaining data indicative of a location to which an
autonomous vehicle is to travel; obtaining, via one or more sensors
of the autonomous vehicle, sensor data associated with a
surrounding environment of the autonomous vehicle; determining a
level of traffic associated with a geographic area that includes
the location to which the autonomous vehicle is to travel based at
least in part on the sensor data; determining an estimated traffic
impact of the autonomous vehicle based at least in part on the
level of traffic that is based at least in part on the sensor data,
wherein the estimated traffic impact is indicative of an estimated
impact of the autonomous vehicle on one or more objects within the
surrounding environment of the autonomous vehicle in the event that
the autonomous vehicle were to stop; and causing the autonomous
vehicle to perform one or more vehicle actions based at least in
part on the estimated traffic impact.
33. The computing system of claim 32, wherein the one or more
vehicle actions comprise the autonomous vehicle entering into a
holding pattern.
34. The computing system of claim 33, wherein the holding pattern
comprises the autonomous vehicle re-routing to travel to return to
within a vicinity of a location associated with a user.
35. The computing system of claim 34, wherein re-routing to travel
to return to within the vicinity of the location associated with
the user comprises circling at least one block.
36. The computing system of claim 32, wherein the estimated traffic
impact of the autonomous vehicle is based at least in part on a
time period during which the autonomous vehicle would be stopped at
least partially within a travel way.
37. The computing system of claim 32, further comprising: searching
the surrounding environment of the autonomous vehicle for a parking
location out of a travel way of the geographic area based at least
in part on the sensor data.
38. The computing system of claim 37, wherein the one or more
objects comprise one or more vehicles.
39. An autonomous vehicle comprising: one or more sensors; one or
more processors; and one or more memory devices, the one or more
memory devices storing instructions that when executed by the one
or more processors cause the autonomous vehicle to perform
operations, the operations comprising: obtaining data indicative of
a location to which an autonomous vehicle is to travel; obtaining,
via one or more sensors of the autonomous vehicle, sensor data
associated with the surrounding environment of the autonomous
vehicle; determining a level of traffic associated with a
geographic area that includes the location to which an autonomous
vehicle is to travel based at least in part on the sensor data;
determining an estimated traffic impact of the autonomous vehicle
based at least in part on the level of traffic that is based at
least in part on the sensor data, wherein the estimated traffic
impact is indicative of an estimated impact of the autonomous
vehicle on one or more objects within the surrounding environment
of the autonomous vehicle in the event that the autonomous vehicle
were to stop; and causing the autonomous vehicle to perform one or
more vehicle actions based at least in part on the estimated
traffic impact.
40. The autonomous vehicle of claim 39, wherein the one or more
vehicle actions comprise the autonomous vehicle waiting at a
stopped position for a user associated with a service request for a
vehicle service, and wherein the operations further comprise:
determining a location of the user; determining an updated
estimated traffic impact while the autonomous vehicle is in the
stopped position; and determining whether the autonomous vehicle is
to leave the stopped position based at least in part on the
location of the user and the updated estimated traffic impact.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] The present application is based on and claims priority to
U.S. Provisional Application 62/510,515 having a filing date of May
24, 2017, which is incorporated by reference herein.
FIELD
[0002] The present disclosure relates generally to controlling the
travel holding pattern of an autonomous vehicle that provides a
vehicle service to a user.
BACKGROUND
[0003] An autonomous vehicle is a vehicle that is capable of
sensing its environment and navigating without human input. In
particular, an autonomous vehicle can observe its surrounding
environment using a variety of sensors and can attempt to
comprehend the environment by performing various processing
techniques on data collected by the sensors. Given knowledge of its
surrounding environment, the autonomous vehicle can identify an
appropriate motion path through such surrounding environment.
SUMMARY
[0004] Aspects and advantages of embodiments of the present
disclosure will be set forth in part in the following description,
or may be learned from the description, or may be learned through
practice of the embodiments.
[0005] One example aspect of the present disclosure is directed to
a computer-implemented method of controlling autonomous vehicles.
The method includes obtaining, by a computing system that includes
one or more computing devices, data indicative of a location
associated with a user to which an autonomous vehicle is to travel.
The autonomous vehicle is to travel along a first vehicle route
that leads to the location associated with the user. The method
includes obtaining, by the computing system, traffic data
associated with geographic area that includes the location
associated with the user. The method includes determining, by the
computing system, an estimated traffic impact of the autonomous
vehicle on the geographic area based at least in part on the
traffic data. The method includes determining, by the computing
system, one or more vehicle actions based at least in part on the
estimated traffic impact. The method includes causing, by the
computing system, the autonomous vehicle to perform the one or more
vehicle actions. The one or more vehicle actions include at least
one of stopping the autonomous vehicle at least partially in a
travel way within a vicinity of the location associated with the
user or travelling along a second vehicle route.
[0006] Another example aspect of the present disclosure is directed
to a computing system for controlling autonomous vehicles. The
computing system includes one or more processors and one or more
memory devices. The one or more memory devices store instructions
that when executed by the one or more processors cause the
computing system to perform operations. The operations include
obtaining data indicative of a location associated with a user. The
user is associated with a request for a vehicle service provided by
an autonomous vehicle. The autonomous vehicle is to travel along a
first vehicle route to arrive within a vicinity of the location
associated with the user. The operations include obtaining traffic
data associated with a geographic area that includes the location
associated with the user. The operations include obtaining location
data associated with a user device associated with the user. The
operations include determining at least one of an estimated traffic
impact of the autonomous vehicle on the geographic area based at
least in part on the traffic data or an estimated time of user
arrival based at least in part on the location data associated with
the user device. The operations include determining one or more
vehicle actions based at least in part on at least one of the
estimated traffic impact or the estimated time of user arrival. The
operations include causing the autonomous vehicle to perform the
one or more vehicle actions. The one or more vehicle actions
include at least one of stopping the autonomous vehicle at least
partially in a travel way within a vicinity of the location
associated with the user or travelling along a second vehicle
route.
[0007] Yet another example aspect of the present disclosure is
directed to an autonomous vehicle includes one or more sensors, a
communication system, one or more processors, and one or more
memory devices. The one or more memory devices store instructions
that when executed by the one or more processors cause the
autonomous vehicle to perform operations. The operations include
obtaining data indicative of a location associated with a user. The
user is associated with a request for a vehicle service provided by
the autonomous vehicle. The operations include controlling the
autonomous vehicle to travel along a first vehicle route to arrive
within a vicinity of the location associated with the user. The
operations include obtaining traffic data associated with a
geographic area that includes the location associated with the user
based at least in part on sensor data obtained via the one or more
sensors. The traffic data is indicative of a level of traffic
within a surrounding environment of the autonomous vehicle. The
operations include obtaining, via the communication system,
location data associated with a user device associated with the
user. The location data associated with the user device is
indicative of one or more locations of the user device associated
with the user at one or more times. The operations include
determining an estimated traffic impact of the autonomous vehicle
on the geographic area based at least in part on the traffic data.
The operations include determining an estimated time of user
arrival based at least in part on the location data associated with
the user device. The operations include determining one or more
vehicle actions based at least in part on at least one of the
estimated traffic impact or the estimated time of user arrival. The
operations include causing the autonomous vehicle to perform the
one or more vehicle actions. The one or more vehicle actions
include at least one of stopping the autonomous vehicle within the
vicinity of the location associated with the user or travelling
along a second vehicle route.
[0008] Other example aspects of the present disclosure are directed
to systems, methods, vehicles, apparatuses, tangible,
non-transitory computer-readable media, and memory devices for
controlling autonomous vehicles.
[0009] These and other features, aspects and advantages of various
embodiments will become better understood with reference to the
following description and appended claims. The accompanying
drawings, which are incorporated in and constitute a part of this
specification, illustrate embodiments of the present disclosure
and, together with the description, serve to explain the related
principles.
BRIEF DESCRIPTION OF THE DRAWINGS
[0010] Detailed discussion of embodiments directed to one of
ordinary skill in the art are set forth in the specification, which
makes reference to the appended figures, in which:
[0011] FIG. 1 depicts an example system overview according to
example embodiments of the present disclosure;
[0012] FIG. 2 depicts an example geographic area that includes a
location associated with a user according to example embodiments of
the present disclosure;
[0013] FIG. 3 depicts example information associated with an
acceptable walking distance according to example embodiments of the
present disclosure;
[0014] FIG. 4 depicts an example display device with an example
communication according to example embodiments of the present
disclosure;
[0015] FIG. 5 depicts example information associated with an
estimated traffic impact according to example embodiments of the
present disclosure;
[0016] FIG. 6 depicts an example travel way according to example
embodiments of the present disclosure;
[0017] FIG. 7 depicts a flow diagram of an example method of
determining an estimated time of user arrival according to example
embodiments of the present disclosure;
[0018] FIG. 8A depicts an example portion of a communications
system according to example embodiments of the present
disclosure;
[0019] FIG. 8B depicts an example portion of a communications
system according to example embodiments of the present
disclosure;
[0020] FIG. 8C depicts an example diagram of obtaining location
data according to example embodiments of the present
disclosure;
[0021] FIG. 9 depicts example information associated with a second
vehicle route according to example embodiments of the present
disclosure;
[0022] FIG. 10 depicts an example display device with an example
communication according to example embodiments of the present
disclosure;
[0023] FIG. 11 depicts example information associated with a
vehicle service cancellation threshold according to example
embodiments of the present disclosure;
[0024] FIG. 12 depicts a flow diagram of an example method of
controlling autonomous vehicles according to example embodiments of
the present disclosure;
[0025] FIGS. 13A-B depict a flow diagram of an example method of
controlling autonomous vehicles according to example embodiments of
the present disclosure; and
[0026] FIG. 14 depicts example system components according to
example embodiments of the present disclosure.
DETAILED DESCRIPTION
[0027] Reference now will be made in detail to embodiments, one or
more example(s) of which are illustrated in the drawings. Each
example is provided by way of explanation of the embodiments, not
limitation of the present disclosure. In fact, it will be apparent
to those skilled in the art that various modifications and
variations can be made to the embodiments without departing from
the scope or spirit of the present disclosure. For instance,
features illustrated or described as part of one embodiment can be
used with another embodiment to yield a still further embodiment.
Thus, it is intended that aspects of the present disclosure cover
such modifications and variations.
[0028] Example aspects of the present disclosure are directed to
improving the travel patterns of an autonomous vehicle to account
for potential traffic impacts, while waiting to provide a Vehicle
service to a user. For instance, an entity (e.g., service provider)
can use a fleet of vehicles to provide a vehicle service (e.g.,
transportation service, delivery service, courier service, etc.) to
a plurality of users. The fleet can include, for example,
autonomous vehicles that can drive, navigate, operate, etc. with
minimal and/or no interaction from a human driver. For example, an
autonomous vehicle can receive data indicative of a location
associated with a user that has requested a vehicle service, such
as a transportation service. The autonomous vehicle can
autonomously navigate along a route towards the location associated
with the user. When within a vicinity of the location, the
autonomous vehicle can attempt to identify a parking spot that is
out of a travel way (e.g., out of a traffic lane). Such parking
spots may not, however, be available to the autonomous vehicle. In
such a case, the autonomous vehicle can determine whether the
vehicle should stop in a travel lane to pick up the user and/or
whether the vehicle should enter a holding pattern whereby the
vehicle continues to travel (e.g., around the block) to return to
the user's location. For example, as further described herein, the
autonomous vehicle can estimate the impact the autonomous vehicle
may have on traffic if the autonomous vehicle were to stop in a
travel way to wait for the user, given, for example, the user's
proximity to the vehicle. In the event that the impact would be
low, the autonomous vehicle can stop in the travel way and alert
the user of the vehicle's location. In the event that the impact
would be high, or becomes high after the vehicle has stopped, the
autonomous vehicle can be re-routed (e.g., around the block) so
that the vehicle can again attempt to pick up the user. In this
way, the system and methods of the present disclosure can improve
the situational awareness of an autonomous vehicle that is waiting
for a user of a vehicle service (e.g., while attempting to pick up
a user for transport).
[0029] More particularly, an entity (e.g., service provider, owner,
manager) can use one or more vehicles (e.g., ground-based vehicles)
to provide a vehicle service such as a transportation service
(e.g., rideshare service), a courier service, a delivery service,
etc. The vehicle(s) can be autonomous vehicles that include various
systems and devices configured to control the operation of the
vehicle. For example, an autonomous vehicle can include an onboard
vehicle computing system for operating the vehicle (e.g., located
on or within the autonomous vehicle). The vehicle computing system
can receive sensor data from sensor(s) onboard the vehicle (e.g.,
cameras, LIDAR, RADAR), attempt to comprehend the vehicle's
surrounding environment by performing various processing techniques
on the sensor data, and generate an appropriate motion plan through
the vehicle's surrounding environment. Moreover, the autonomous
vehicle can be configured to communicate with one or more computing
devices that are remote from the vehicle. For example, the
autonomous vehicle can communicate with an operations computing
system that can be associated with the entity. The operations
computing system can help the entity monitor, communicate with,
manage, etc. the fleet of vehicles.
[0030] An autonomous vehicle can be configured to operate in a
plurality of operating modes. For example, an autonomous vehicle
can be configured to operate in a fully autonomous (e.g.,
self-driving) operating mode in which the autonomous vehicle can
drive and navigate with no interaction from a human driver present
in the vehicle. In some implementations, a human driver may not be
present in the autonomous vehicle. The autonomous vehicle can also
be configured to operate in an approach mode in which autonomous
vehicle performs various functions as it approaches a location
associated with a user, such as searching its surrounding
environment for a parking location. The approach operating mode can
be utilized, for example, when the autonomous vehicle is
approaching a user that has requested a vehicle service, as further
described herein.
[0031] A user can make a request for a vehicle service provided by
the autonomous vehicle. For instance, a user can provide (e.g., via
a user device) a request to the operations computing system of an
entity (e.g., service provider, manager, owner) that is associated
with the autonomous vehicle. The request can indicate the type of
vehicle service that the user desires (e.g., a transportation
service, a delivery service, a courier service, etc.), a location
associated with the user (e.g., a current location of the user, a
different location, etc.), an identifier (e.g., phone number,
Bluetooth, WiFi, Cellular, other data that can be used to contact
the user, etc.) associated with the user device that provided the
request, and/or other information.
[0032] The operations computing system can process the request and
select an autonomous vehicle to provide the requested vehicle
service to the user. The operations computing system can provide,
to the autonomous vehicle, data indicative of a location to which
the autonomous vehicle is to travel. The location can be associated
with the user that requested the vehicle service. For example, the
location can be the current location of the user and/or a different
location, such as for example a location at which the user would
like to be picked up by the autonomous vehicle, provide an item to
the autonomous vehicle, retrieve an item from the autonomous
vehicle, etc. The location can be expressed as a coordinate (e.g.,
GPS coordinate, latitude-longitude coordinate pair), an address, a
place name, and/or other geographic reference that can be used to
identify the location. The location associated with the user can be
represented, for example, as a pin on a map user interface.
[0033] The autonomous vehicle can obtain, from the operations
computing system, the data indicative of the location associated
with the user. The autonomous vehicle can also obtain a first
vehicle route that leads to the location associated with the user.
The first vehicle route can be, for example, a route from the
current location of the vehicle to the location associated with the
user. In some implementations, the operations computing system can
provide the first vehicle route to the autonomous vehicle.
Additionally, or alternatively, the onboard vehicle computing
system of the autonomous vehicle can determine the first Vehicle
route.
[0034] The autonomous vehicle can travel in accordance with the
first vehicle route to arrive within a vicinity of the location
associated with the user. The vicinity can be defined by a distance
(e.g., a radial distance) from the location associated with the
user. The distance can be indicative of an acceptable walking
distance from the location associated with the user. For example,
an acceptable walking distance can be distance that a user would be
willing walk to arrive at a vehicle. The acceptable walking
distance can be determined based on a variety of information. Such
information can include, for example, specific user preferences
stored in a profile associated with the user, weather information
(e.g., gathered via sensors onboard the vehicle, provided by a
third party source, etc.), traffic conditions (e.g., current,
historic, future predicted traffic conditions), historic vehicle
services data (e.g., historic pickup data for previous
transportation services), and/or other types of information. The
autonomous vehicle can provide a communication to the user
indicating that the autonomous vehicle is within the vicinity of
the location associated with the user. For instance, such a
communication can be in the form of a textual message stating "your
ride is arriving please prepare to board".
[0035] Once the autonomous vehicle is within the vicinity of the
location associated with the user, the autonomous vehicle can begin
to search for a parking location. For example, the autonomous
vehicle can enter into the approach operating mode when the vehicle
is within the vicinity of the location associated with the user.
While it is within the vicinity, the autonomous vehicle can search
for a parking location before it reaches the location associated
with the user (e.g., before the GPS pin coordinate on a map) and/or
after it passes the location associated with the user, but is still
within the vicinity of the user (e.g., within acceptable walking
distance for the user).
[0036] To identify a parking spot, the autonomous vehicle can
obtain sensor data associated with one or more objects that are
proximate to the vehicle (e.g., within a field of view of one or
more of the vehicle's onboard sensor(s)). The sensor data can
include image data, radar data, LIDAR data, and/or other data
acquired by the vehicle's sensor(s). The object(s) can include, for
example, pedestrians, vehicles, bicycles, and/or other objects. The
sensor data can be indicative of locations associated with the
object(s) within the surrounding environment of the vehicle at one
or more times. The autonomous vehicle can process the sensor data
to determine if there are any available parking locations that are
not currently occupied by the objects (e.g., other vehicles) within
the vehicle's surrounding environment. In some implementations, the
autonomous vehicle can utilize map data to determine if there are
any designated parking locations (e.g., parking lots, pullover
lanes, etc.) within the vicinity of the location associated with
the user.
[0037] In the event the autonomous vehicle is able to identify a
parking location that is out of a travel way and within the
vicinity of the location associated with the user (e.g., out of a
traffic lane), the autonomous vehicle can position itself into that
parking location accordingly. The autonomous vehicle can send a
communication to a user device associated with the user. The
communication can indicate that the vehicle has arrived as well as
the location of the autonomous vehicle. In some implementations,
the user device can display a map user interface that includes a
user route. The user route can be a route along which a user can
travel to arrive at the autonomous vehicle. In the event that the
autonomous vehicle is unable to identify a parking location that is
out of the travel way (and within the vicinity of the user), the
autonomous vehicle can decide whether or not to stop at least
partially in a travel way to wait for the user.
[0038] To help decide whether to stop within a travel way, the
autonomous vehicle can obtain traffic data associated with a
geographic area that includes the location associated with the
user. The traffic data can include various types of data such as
historic traffic data, predicted traffic data, and/or current
traffic data associated with the geographic area (e.g., within the
vicinity of the location of the user, a wider area, etc.). The
traffic data can be obtained from a variety of sources such as
other autonomous vehicles (e.g., within the vehicle fleet), the
operations computing system, third party sources (e.g., regional
traffic management entities, etc.), as well as the autonomous
vehicle itself.
[0039] By way of example, the autonomous vehicle can obtain the
current traffic data associated with the geographic area that
includes the location associated with the user. To do so, the
autonomous vehicle can obtain sensor data (e.g., via its onboard
sensors) associated with the surrounding environment of the
autonomous vehicle that is within the vicinity of the location
associated with the user, as described herein. The sensor data can
be indicative of one or more objects within the surrounding
environment of the autonomous vehicle. The autonomous vehicle can
process the sensor data to classify which of the object(s) would be
impacted (e.g., caused to stop) by the autonomous vehicle stopping
in a travel way. For example, the autonomous vehicle can classify
the vehicles that are behind the autonomous vehicle (in the same
travel lane) as objects that would be impacted in the event the
autonomous vehicle were to stop at least partially in the travel
way. The autonomous vehicle can also identify object(s) that would
not be affected by the vehicle stopping in the travel way. For
example, the autonomous vehicle can determine that object(s) that
have a path to travel around the autonomous vehicle (e.g.,
bicycles, vehicles in adjacent lanes, vehicles with clear paths to
change lanes, etc.), may not be impacted and/or may be impacted to
an insignificant degree. After such classification, the autonomous
vehicle can determine a level of traffic associated with the
geographic area (e.g., within the vicinity of the user's location)
based at least in part on the sensor data. For example, the level
of traffic can be based at least in part on the number of object(s)
within the surrounding environment of the autonomous vehicle that
would be impacted by the autonomous vehicle stopping at least
partially in the travel way, while filtering out those object(s)
that would not be impacted. Similar such information could be
acquired via one or more other vehicles in the associated vehicle
fleet.
[0040] The autonomous vehicle can determine an estimated traffic
impact of the autonomous vehicle on the geographic area based at
least in part on the traffic data. The estimated traffic impact can
be indicative of an estimated impact of the autonomous vehicle on
one or more objects within a surrounding environment of the
autonomous vehicle in the event that the autonomous vehicle were to
stop at least partially in the travel way (e.g., in the vicinity of
the location associated with the user). For example, in some
implementations, the autonomous vehicle can compare the level of
traffic (e.g., based on the sensor data) to a traffic constraint to
determine whether the estimated traffic impact would be high or
low. The traffic constraint can be implemented in a variety of
forms. For example, the traffic constraint can include a traffic
threshold that is indicative of an acceptable level of traffic
(e.g., an acceptable number of objects) that would be impacted by
the autonomous vehicle stopping at least partially in the travel
way. A traffic level that exceeds the traffic threshold would be
considered a high impact on traffic. In some implementations, the
traffic constraint can be implemented as cost data (e.g., one or
more cost function(s)). For example, the autonomous vehicle's
onboard vehicle computing system can include cost data that
reflects the cost(s) of stopping vehicle motion, the cost(s) of
causing traffic build-up, the cost(s) of illegally stopping in a
travel way, etc.
[0041] The traffic constraint can be based on a variety of
information. In some implementations, the traffic constraint can be
based at least in part on historic traffic data that indicates the
level of traffic previously occurring in that geographic area. For
example, if the geographic area normally experiences a high level
of traffic build-up, a corresponding traffic threshold can be
higher (and/or the cost of stopping can be lower). Additionally, or
alternatively, the traffic constraint can be based at least in part
on real-time traffic data (e.g., from other vehicles in the fleet,
from the autonomous vehicle, other sources). For example, in the
event that there is already a traffic jam in the vicinity of the
location of the user, the traffic threshold could be higher (and/or
the cost of stopping could be lower). In some implementations, the
traffic constraint can be based at least in part on the typical
travel expectations of individuals in the geographic area. For
example, individuals that are located in City A may be more patient
when waiting in traffic than those in City B. Thus, a traffic
threshold may be higher in City A than in City B (and/or the cost
of stopping may be lower in City A than in City B). In some
implementations, the traffic constraint can be based at least in
part on map data. For example, in the event that the autonomous
vehicle is traveling on a wide travel way in which impacted
vehicles could eventually travel around the autonomous vehicle, the
traffic threshold could be higher and/or the cost of stopping could
be lower).
[0042] In some implementations, the traffic constraint can be
determined at least in part from a model, such as a machine-learned
model. For example, the machine-learned model can be or can
otherwise include one or more various model(s) such as, for
example, models using boosted random forest techniques, neural
networks (e.g., deep neural networks), or other multi-layer
non-linear models. Neural networks can include recurrent neural
networks (e.g., long short-term memory recurrent neural networks),
feed-forward neural networks, and/or other forms of neural
networks. For instance, supervised training techniques can be
performed to train the model (e.g., using previous driving logs) to
determine a traffic constraint based at least in part on input
data. The input data can include, for example, traffic data as
described herein, map data, data from a traffic management entity,
driving characteristics of individuals in an associated geographic
area, complaints received from operators of vehicles that were
caused to stop by autonomous vehicles, etc. The machine-learned
model can provide, as an output, data indicative of a recommended
traffic constraint. In some implementations, the recommended
traffic constraint can be specific to the geographic area.
[0043] The autonomous vehicle can also, or alternatively, determine
an estimated time of user arrival in order to help determine
whether or not to stop at least partially in the travel way. For
instance, the autonomous vehicle can obtain location data
associated with a user device associated with a user. The location
data can be indicative of the position of the user device
associated with the user. The autonomous vehicle can use one or
more identifier(s) of the user device (e.g, provided by the
operations computing system) to scan for and/or communicate with
the user device when the vehicle is within the vicinity of the
user. For example, the autonomous vehicle can utilize multiple
input/multiple output communication, Bluetooth low energy protocol,
RF signaling, and/or other communication technologies to obtain the
location data. The autonomous vehicle can determine the estimated
time of user arrival based at least in part on the location data
associated with the user device. The estimated time of user arrival
can be indicative of, for example, a time at which the user is
estimated to complete boarding of the autonomous vehicle (e.g., for
a transportation service). The estimated time of user arrival can
be expressed as a time duration (e.g., user estimated to arrive in
1 minute) and/or a point in time (e.g., user estimated to arrive at
10:31 am (PT)). In this way, the autonomous vehicle can determine
an amount of stopping time that the object(s) within its
surroundings would be impacted as the autonomous vehicle waits for
the user's arrival (e,g., how long other vehicles would be caused
to stop while waiting for the user).
[0044] By way of example, the operations computing system can
provide the autonomous vehicle with one or more identifier(s) of a
user device associated with the user. The autonomous vehicle can
use the identifier(s) (e.g., Bluetooth, WiFi, Cellular, other
identifier) to determine the location of the user. For example,
when the autonomous vehicle is within a vicinity of the location
associated with the user, the autonomous vehicle can scan for the
user device based at least in part on the identifier(s). Once the
user device is found, the autonomous vehicle can track changes in
the signal strength (e.g., radio signal strength identifier) to
determine the approximate heading of the user as well as the
approximate distance between the user and autonomous vehicle (e.g.,
without authenticated connection). In some implementations, the
autonomous vehicle can determine differences in Bluetooth Low
Energy beacon radio signal strength identifiers over time and/or
inertial measurement unit changes, which can indicate distance and
direction of the autonomous vehicle from the user device (e.g.,
mobile phone associated with the user). The autonomous vehicle can
calculate the estimated time of user arrival based at least on the
heading of the user and the approximate distance between the user
and the vehicle (and/or the estimated velocity of the user), as
further described herein. In some implementations, the autonomous
vehicle can obtain the location data using multiple input, multiple
output communication between the autonomous vehicle and a user
device associated with the user. This can allow the autonomous
vehicle to take advantage of the multiple antennas included in the
vehicle's communication system as well as those of the user device
to increase accuracy of the location data associated with the user.
In some implementations, the estimated time to user arrival can be
based at least in part on historic data. Such historic data can
include, for example, previous correlations between changes in the
signal strength of an identifier and the user's time to arriving at
a vehicle.
[0045] Additionally, or alternatively, the autonomous vehicle,
systems, and methods described herein can utilize other
communication methods. Such communication methods can include, for
example, autonomous vehicle based triangulation (e.g., on vehicle
triangulated RF), on-vehicle multi-range beacon (e.g., Bluetooth
Low Energy) hardware (e.g., paired with a software application on a
user device), application triangulation, user device to vehicle
handshake (e.g., light signal handshake), autonomous vehicle
perception of the user (e.g., via processing of sensor data to
perceive the user and the user's location, distance, heading,
velocity, and/or other state data associated therewith), GPS
location of the user device, device specific techniques (e.g.,
associated with a specific device/model type), the autonomous
vehicle serving as a localized base station (e.g., GPS, WiFi, etc.)
for the user device, autonomous vehicle localization and user
device image based localization via one or more network(s), and/or
other techniques.
[0046] In some implementations, in order to determine whether the
amount of stopping time is acceptable, the autonomous vehicle can
compare the estimated time of user arrival to a time constraint.
The time constraint can be expressed as a time threshold (e.g.,
indicating an acceptable amount of stopping time) and/or cost data
(e.g., cost functions expressing a cost in relation to stopping
time). Similar to the traffic constraint, the time constraint can
be based on historic data (e.g., indicating historic wait times),
real-time data (e.g., indicating that the vehicles are already
waiting due to another traffic build-up in front of the autonomous
vehicle), expectations of individuals in the geographic area,
machine-learned model(s), and/or other information.
[0047] In some implementations, the estimated time of user arrival
can also factor in additional amounts of time that can impact the
objects) within the surrounding environment of the autonomous
vehicle while the autonomous vehicle is stopped, awaiting the user.
Example instances requiring such additional amounts of time can be
associated with a user getting into the vehicle and securely
fastening his/her seatbelt, a user helping other passengers enter
and become securely positioned within the vehicle (e.g., children
or others requiring assistance), a user loading luggage or other
items for transportation within the vehicle, a user receiving
delivered item(s) from/placing item(s) within the vehicle, etc. In
some implementations, the additional amounts of time can be
determined based at least in part on information provided with a
service request (e.g., type of service, destination, number of
passengers, child's car seat requested, etc.
[0048] The autonomous vehicle can determine one or more vehicle
actions based at least in part on the estimated traffic impact
and/or the estimated time of user arrival. In some implementations,
vehicle action(s) can include stopping within the vicinity of the
location associated with the user (e.g., at least partially in the
travel way). For example, in the event that the level of traffic
(e.g., the number of other vehicles that would be impacted by an
in-lane stop) is below a traffic threshold, the autonomous vehicle
can determine that the vehicle can stop within the travel way to
wait for the user to arrive at the vehicle. In another example, in
the event that the estimated time of user arrival is below the time
threshold, the autonomous vehicle can determine that the vehicle
can stop art least partially within the travel way.
[0049] In some implementations, the autonomous vehicle can base its
determination to stop at least partially within the travel way on
both the estimated traffic impact and the estimated time of user
arrival. For instance, the autonomous vehicle can weigh each of
these estimates to determine whether it would be appropriate for
the vehicle to stop at least partially in the travel way to wait
for the user. The autonomous vehicle can apply a first weighting
factor to the estimated traffic impact and a second weighting
factor to the estimated time of user arrival. The first weighting
factor can be different than the second weighting factor. By way of
example, the estimated impact on traffic may be high while the
estimated time of user arrival may be short. Accordingly, the
autonomous vehicle may determine that it can stop within the travel
way because although a higher number of objects (e.g., other
vehicles) may be caused to stop, it would only be for a short
period of time because the user is close in distance to the
autonomous vehicle. In such a case, the estimated time of user
arrival can he given a greater weight than the estimated traffic
impact. In another example, the estimated impact on traffic may be
low while the estimated time of user arrival may be long.
Accordingly, the autonomous vehicle may determine that it should
not stop within the travel way because although only a few objects
(e.g., other vehicles) may be caused to stop, it would be for a
greater period of time because the user is farther from the
autonomous vehicle. In such a case, the estimated traffic impact
can be given a greater weight than the estimated time of user
arrival.
[0050] The vehicle action(s) can also, or alternatively, include
the autonomous vehicle entering into a holding pattern. In such a
case, the vehicle action(s) can include traveling along a second
vehicle route (e.g., an optimal holding pattern route). For
example, the autonomous vehicle may be unable to find a parking
location before and/or after the location associated with the user
(e.g., before and/or after a pin location of the user).
Additionally, the autonomous vehicle may determine that it should
not stop within a travel way to wait for the user's arrival, as
described herein. Thus, the autonomous vehicle can be re-routed
along a second vehicle route that is at least partially different
than the first vehicle route. The second vehicle route can be a
path along which the autonomous vehicle can travel to re-arrive
within the vicinity of the location of the user. For example, the
second vehicle route can be a path along which the autonomous
vehicle can travel around a block, back to the location associated
with the user. This can afford the user additional time to arrive
at the vehicle, without the autonomous vehicle impacting traffic
(e.g., due to a stop).
[0051] In some implementations, the autonomous vehicle can
determine that it can stop within the travel way, but later
determine that it must begin to travel again (e.g., according to a
holding pattern route). For example, the autonomous vehicle can
determine that it would be appropriate to stop at least partially
within the travel way to wait for the user based at least in part
on the estimated traffic impact and/or the estimated time of user
arrival. While the autonomous vehicle is stopped, the traffic
impact may increase (e.g., due to an increase in the number of
other vehicle(s) stopped behind the autonomous vehicle) and/or the
user may take longer than estimated to arrive at the autonomous
vehicle. As such, the autonomous vehicle can determine an updated
estimated traffic impact (e.g., based on the number of vehicles
that have already stopped and/or additional vehicles that may be
caused to stop) and/or an updated estimated time of user arrival
(e.g., based on a change in the user device location data, if any).
The autonomous vehicle can then determine that it can no longer
remain stopped to wait for the user based at least in part on the
updated estimates. As such, the autonomous vehicle can begin to
travel again, for example, along the second vehicle route (e.g.,
around the block).
[0052] The vehicle computing system of the autonomous vehicle can
implement the determined vehicle action(s). For example, in the
event that the autonomous vehicle has determined to stop in the
travel way, the vehicle computing system can cause the autonomous
vehicle to stop by sending one or more control signals to the
braking control system(s) of the autonomous vehicle. In the event
that the autonomous vehicle has determined to travel along a second
vehicle route (e.g., in accordance with the holding pattern), the
vehicle computing system can obtain data associated with the second
vehicle route and implement the route accordingly. For example, the
vehicle computing system can request and obtain data indicative of
the second vehicle route from the operations computing system.
Additionally, or alternatively, the vehicle computing system can
determine the second vehicle route onboard the vehicle. In either
case, the vehicle computing system can send one or more control
signals to cause a motion planning system of the autonomous vehicle
to plan the motion of the vehicle in accordance with the second
vehicle route (e.g., to implement a vehicle trajectory in
accordance with the second vehicle action).
[0053] The autonomous vehicle can provide the user with one or more
communications indicating the actions taken by the autonomous
vehicle. For example, in the event that the autonomous vehicle
stops within the travel way, the autonomous vehicle can provide a
communication to the user indicating that the vehicle is waiting
for the user (e.g., "your vehicle has arrived, please proceed
quickly to the vehicle"). In response to receiving such a
communication, a user device associated with the user can display a
map user interface indicating the vehicle location of the
autonomous vehicle and a user route to the vehicle's location. In
the event that the autonomous vehicle does not find a parking spot
and does not stop in the travel way, the autonomous vehicle can
provide a communication to the user indicating as such (e.g., "I
could not locate you at the pin drop, traffic forced me to go
around the block. Please proceed to the pin drop").
[0054] In some implementations, the autonomous vehicle may be
relieved of its responsibility to provide a vehicle service to the
user. For instance, in the event that a vehicle computing system
and/or operations computing system determines that an autonomous
vehicle should travel along the second vehicle route, such
computing system(s) can determine whether it would be advantageous
(e.g., more time efficient, more fuel efficient, etc.) for another
autonomous vehicle in the area to be routed to the user. In the
event that it would be advantageous, the computing system(s) can
provide data to the autonomous vehicle indicating that the
autonomous vehicle is no longer responsible for the service
request. Additionally, or alternatively, the computing system(s)
can re-route the autonomous vehicle to provide a vehicle service to
another user.
[0055] In some implementations, the autonomous vehicle can cancel
the service request of the user. By way of example, the autonomous
vehicle may be caused to re-route (e.g., circle the block) a
certain number of times and/or the user may not arrive at the
autonomous vehicle within a certain timeframe (e.g., above a trip
cancellation threshold). In response, the autonomous vehicle can
send data to the operations computing system requesting the
cancellation of the user's service request. The operations
computing system can cancel the service request (and inform the
user accordingly) and/or re-route the autonomous vehicle to provide
a vehicle service to another user. In some implementations, the
autonomous vehicle can communicate directly with a user device
associated with the user to cancel the service request and inform
the user accordingly. The autonomous vehicle can report such a
cancellation to the operations computing system.
[0056] The systems and methods described herein may provide a
number of technical effects and benefits. For instance, the systems
and methods enable an autonomous vehicle to determine a holding
pattern, for waiting for a user, onboard the autonomous vehicle. As
such, the autonomous vehicle need not communicate with a remote
computing system (e.g., operations computing system) each time the
vehicle must decide whether to stop in a travel way or re-route the
vehicle (e.g., around the block). This can help improve the
response time of the vehicle computing system when deciding and/or
implementing such a holding pattern. Moreover, by enabling the
vehicle computing system to determine whether to stop and/or to
re-route onboard the Vehicle, the systems and methods described
herein can save computational resources of the operations computing
system (that would otherwise be required for such determination).
Accordingly, the computational resources of the operations
computing system can be allocated to other core functions such as
the management of service requests, routing of autonomous vehicles,
etc.
[0057] The systems and methods of the present disclosure also
provide an improvement to vehicle computing technology, such as
autonomous vehicle computing technology. For instance, the
computer-implemented methods and systems improve the situational
awareness of the autonomous vehicle to provide a vehicle service to
a user. The systems and methods can enable a computing system to
obtain data indicative of a location associated with a user to
which an autonomous vehicle is to travel. The autonomous vehicle
can travel along a first vehicle route that leads to the location
associated with the user. The computing system can obtain traffic
data associated with a geographic area that includes the location
associated with the user and location data associated with a user
device associated with the user. The computing system can determine
an estimated traffic impact of the autonomous vehicle on the
geographic area based at least in part on the traffic data and/or
an estimated time of user arrival based at least in part on the
location data associated with the user device. The computing system
can determine one or more vehicle actions based at least in part on
the estimated traffic impact and/or the estimated time of user
arrival. The computing system can cause the autonomous vehicle to
perform the one or more vehicle actions. As described herein, the
vehicle action(s) can include at least one of stopping the
autonomous vehicle at least partially in a travel way within a
vicinity of the location associated with the user or travelling
along a second vehicle route. In this way, the vehicle computing
system can improve the holding pattern of the autonomous vehicle
that is waiting for a user. For example, the holding pattern can be
customized based on the estimated impact on the traffic surrounding
the autonomous vehicle and/or the estimated time it will take for
the specific user to arrive at the vehicle. As such, the systems
and methods can improve the vehicle computing system's situational
awareness by allowing it to take into account (e.g., in real-time)
such circumstances when making a determination as to how to best
provide vehicle services. Such approach can also increase the
efficiency of implementing a holding pattern e.g., by avoiding the
aforementioned latency issues) while providing an additional
benefit of minimizing the autonomous vehicle's impact on
traffic.
[0058] Additionally, the systems and methods of the present
disclosure can enhance the user experience associated with the
autonomous vehicle. For instance, the systems and methods described
herein provide a systematic approach that enables users to engage
autonomous vehicles and receive effectively communicated
information regarding expected autonomous locations including, for
example, arrival times, use of holding patterns when needed, etc.
By way of example, the communications and user interfaces described
herein provide the user with updated information regarding the
autonomous vehicle's actions and locations, thereby increasing the
user's knowledge and understanding of the autonomous vehicle's
intentions. In addition to these advantages, the user experience is
further enhanced in that the described systems and methods can
decrease the likelihood that a user will be subjected to potential
frustration from other drivers or pedestrians in the surrounding
environment that could be impacted while the user is arriving to
and/or boarding the autonomous vehicle.
[0059] With reference now to the FIGS., example embodiments of the
present disclosure will be discussed in further detail. FIG. 1
depicts an example system 100 according to example embodiments of
the present disclosure. The system 100 can include a vehicle
computing system 102 associated with a vehicle 104 and an
operations computing system 106 that is remote from the vehicle
104.
[0060] The vehicle 104 incorporating the vehicle computing system
102 can be a ground-based autonomous vehicle (e.g., car, truck,
bus. etc.), an air-based autonomous vehicle (e.g., airplane, drone,
helicopter, or other aircraft), or other types of vehicles (e.g.,
watercraft, etc.). The vehicle 104 can be an autonomous vehicle
that can drive, navigate, operate, etc. with minimal and/or no
interaction from a human driver. A human operator can be omitted
from the vehicle 104 (and/or also omitted from remote control of
the vehicle 104).
[0061] The vehicle 104 can be configured to operate in a plurality
of operating modes 108A-D. The vehicle 104 can be configured to
operate in a fully autonomous (e.g., self-driving) operating mode
108A in which the vehicle 104 can drive and navigate with no input
from a user present in the vehicle 104. The vehicle 104 can be
configured to operate in a semi-autonomous operating mode 108B in
which the vehicle 104 can operate with some input from a user
present in the vehicle 104. The vehicle 104 can enter into a manual
operating mode 108C in which the vehicle 104 is fully controllable
by a user (e.g., human driver) and can be prohibited from
performing autonomous navigation (e.g., autonomous driving). In
some implementations, the vehicle 104 can implement vehicle
operating assistance technology (e.g., collision mitigation system,
power assist steering, etc.) while in the manual operating mode
108C to help assist the operator of the vehicle 104.
[0062] The vehicle 104 can also be configured to operate in an
approach mode 108D in which vehicle 104 performs various functions
as it approaches a location associated with a user. For example,
the vehicle 104 can enter into the approach mode 108D when the
vehicle is within a vicinity of a user to which the vehicle 104 is
to provide a vehicle service, as further described herein. While in
the approach mode 108D, the vehicle 104 can search its surrounding
environment for a parking location as well as communicate with a
user (e.g., via a user device associated with the user).
Additionally, or alternatively, the vehicle 104 can be configured
to evaluate its estimated impact on traffic and/or an estimated
time of user arrival, as further described herein, when the vehicle
104 is in the approach mode 108D.
[0063] The operating mode of the vehicle 104 can be adjusted in a
variety of manners. In some implementations, the operating mode of
the vehicle 104 can be selected remotely, off-board the vehicle
104. For example, an entity associated with the vehicle 104 (e.g.,
a service provider) can utilize the operations computing system 106
to manage the vehicle 104 (and/or an associated fleet). The
operations computing system 106 can send one or more control
signals to the vehicle 104 instructing the vehicle 104 to enter
into, exit from, maintain, etc. an operating mode. By way of
example, the operations computing system 106 can send one or more
control signals to the vehicle 104 instructing the vehicle 104 to
enter into the fully autonomous operating mode 108A. In some
implementations, the operating mode of the vehicle 104 can be set
onboard and/or near the vehicle 104. For example, the vehicle
computing system 102 can automatically determine when and where the
vehicle 104 is to enter, change, maintain, etc. a particular
operating mode (e.g., without user input). Additionally, or
alternatively, the operating mode of the vehicle 104 can be
manually selected via one or more interfaces located onboard the
vehicle 104 (e.g., key switch, button, etc.) and/or associated with
a computing device proximate to the vehicle 104 (e.g., a tablet
operated by authorized personnel located near the vehicle 104). In
some implementations, the operating mode of the vehicle 104 can be
adjusted based at least in part on a sequence of interfaces located
on the vehicle 104. For example, the operating mode may be adjusted
by manipulating a series of interfaces in a particular order to
cause the vehicle 104 to enter into a particular operating
mode.
[0064] The vehicle computing system 102 can include one or more
computing devices located onboard the vehicle 104. For example, the
computing device(s) can be located on and/or within the vehicle
104. The computing device(s) can include various components for
performing various operations and functions. For instance, the
computing device(s) can include one or more processor(s) and one or
more tangible, non-transitory, computer readable media (e.g.,
memory devices). The one or more tangible, non-transitory, computer
readable media can store instructions that when executed by the one
or more processor(s) cause the vehicle 104 (e.g., its computing
system, one or more processors, etc.) to perform operations and
functions, such as those described herein for controlling an
autonomous vehicle.
[0065] As shown in FIG. 1, the vehicle 104 can include one or more
sensors 112, an autonomy computing system 114, and one or more
vehicle control systems 116. One or more of these systems can be
configured to communicate with one another via a communication
channel. The communication channel can include one or more data
buses (e.g., controller area network (CAN)), on-board diagnostics
connector (e.g., OBD-II), and/or a combination of wired and/or
wireless communication links. The onboard systems can send and/or
receive data, messages, signals, etc. amongst one another via the
communication channel.
[0066] The sensor(s) 112 can be configured to acquire sensor data
118 associated with one or more objects that are proximate to the
vehicle 104 (e.g., within a field of view of one or more of the
sensor(s) 112). The sensor(s) 112 can include a Light Detection and
Ranging (LIDAR) system, a Radio Detection and Ranging (RADAR)
system, one or more cameras (e.g., visible spectrum cameras,
infrared cameras, etc.), motion sensors, and/or other types of
imaging capture devices and/or sensors. The sensor data 118 can
include image data, radar data, LIDAR data, and/or other data
acquired by the sensor(s) 112. The object(s) can include, for
example, pedestrians, vehicles, bicycles, and/or other objects. The
object(s) can be located in front of, to the rear of, and/or to the
side of the vehicle 104. The sensor data 118 can be indicative of
locations associated with the object(s) within the surrounding
environment of the vehicle 104 at one or more times. The sensor(s)
112 can provide the sensor data 118 to the autonomy computing
system 114.
[0067] In addition to the sensor data 118, the autonomy computing
system 114 can retrieve or otherwise obtain map data 120. The map
data 120 can provide detailed information about the surrounding
environment of the vehicle 104. For example, the map data 120 can
provide information regarding: the identity and location of
different roadways, road segments, buildings, or other items or
objects (e.g., lampposts, crosswalks, curbing, etc.) the location
and directions of traffic lanes (e.g., the location and direction
of a parking lane, a turning lane, a bicycle lane, or other lanes
within a particular roadway or other travel way and/or one or more
boundary markings associated therewith); traffic control data
(e.g., the location and instructions of signage, traffic lights, or
other traffic control devices); and/or any other map data that
provides information that assists the vehicle 104 in comprehending
and perceiving its surrounding environment and its relationship
thereto.
[0068] The vehicle 104 can include a positioning system 122. The
positioning system 122 can determine a current position of the
vehicle 104. The positioning system 122 can be any device or
circuitry for analyzing the position of the vehicle 104. For
example, the positioning system 122 can determine position by using
one or more of inertial sensors, a satellite positioning system,
based on IP/MAC address, by using triangulation and/or proximity to
network access points or other network components (e.g., cellular
towers, WiFi access points, etc.) and/or other suitable techniques.
The position of the vehicle 104 can be used by various systems of
the vehicle computing system 102 and/or provided to one or more
remote computing device(s) (e.g., of the operations computing
system 106). For example, the map data 120 can provide the vehicle
104 relative positions of the surrounding environment of the
vehicle 104. The vehicle 104 can identify its position within the
surrounding environment (e.g., across six axes) based at least in
part on the data described herein. For example, the vehicle 104 can
process the sensor data 118 (e.g., LIDAR data, camera data) to
match it to a map of the surrounding environment to get an
understanding of the vehicle's position within that
environment.
[0069] The autonomy computing system 114 can include a perception
system 124, a prediction system 126, a motion planning system 128,
and/or other systems that cooperate to perceive the surrounding
environment of the vehicle 104 and determine a motion plan for
controlling the motion of the vehicle 104 accordingly. For example,
the autonomy computing system 114 can receive the sensor data 118
from the sensor(s) 112, attempt to comprehend the surrounding
environment by performing various processing techniques on the
sensor data 118 (and/or other data), and generate an appropriate
motion plan through such surrounding environment. The autonomy
computing system 114 can control the one or more vehicle control
systems 116 to operate the vehicle 104 according to the motion
plan.
[0070] The autonomy computing system 114 can identify one or more
objects that are proximate to the vehicle 104 based at least in
pail on the sensor data 118 and/or the map data 120. For example,
the perception system 124 can obtain state data 130 descriptive of
a current state of an object that is proximate to the vehicle 104.
The state data 130 for each object can describe, for example, an
estimate of the object's: current location (also referred to as
position); current speed (also referred to as velocity); current
acceleration; current heading; current orientation; size/footprint
(e.g., as represented by a bounding shape); class (e.g., pedestrian
class vs. vehicle class vs. bicycle class), and/or other state
information. The perception system 124 can provide the state data
130 to the prediction system 126 (e.g., for predicting the movement
of an object).
[0071] The prediction system 126 can create predicted data 132
associated with each of the respective one or more objects
proximate to the vehicle 104. The predicted data 132 can be
indicative of one or more predicted future locations of each
respective object. The predicted data 132 can be indicative of a
predicted path (e.g., predicted trajectory) of at least one object
within the surrounding environment of the vehicle 104. For example,
the predicted path (e.g., trajectory) can indicate a path along
which the respective object is predicted to travel over time
(and/or the speed at which the object is predicted to travel along
the predicted path). The prediction system 126 can provide the
predicted data 132 associated with the object(s) to the motion
planning system 128.
[0072] The motion planning system 128 can determine a motion plan
134 for the vehicle 104 based at least in part on the predicted
data 132 (and/or other data). The motion plan 134 can include
vehicle actions with respect to the objects proximate to the
vehicle 104 as well as the predicted movements. For instance, the
motion planning system 128 can implement an optimization algorithm
that considers cost data associated with a vehicle action as well
as other objective functions (e.g., cost functions based on speed
limits, traffic lights, etc.), if any, to determine optimized
variables that make up the motion plan 134. By way of example, the
motion planning system 128 can determine that the vehicle 104 can
perform a certain action (e.g., pass an object) without increasing
the potential risk to the vehicle 104 and/or violating any traffic
laws (e.g., speed limits, lane boundaries, signage). The motion
plan 134 can include a planned trajectory, speed, acceleration,
other actions, etc. of the vehicle 104.
[0073] The motion planning system 128 can provide the motion plan
134 with data indicative of the vehicle actions, a planned
trajectory, and/or other operating parameters to the vehicle
control system(s) 116 to implement the motion plan 134 for the
vehicle 104. For instance, the vehicle 104 can include a mobility
controller configured to translate the motion plan 134 into
instructions. By way of example, the mobility controller can
translate a determined motion plan 134 into instructions to adjust
the steering of the vehicle 104 "X" degrees, apply a certain
magnitude of braking force, etc. The mobility controller can send
one or more control signals to the responsible vehicle control
component (e.g., braking control system, steering control system,
acceleration control system) to execute the instructions and
implement the motion plan 134.
[0074] The vehicle 104 can include a communications system 136
configured to allow the vehicle computing system 102 (and its
computing device(s)) to communicate with other computing devices.
The vehicle computing system 102 can use the communications system
136 to communicate with the operations computing system 106 and/or
one or more other remote computing device(s) over one or more
networks (e.g., via one or more wireless signal connections). In
some implementations, the communications system 136 can allow
communication among one or more of the system(s) on-board the
vehicle 104. The communications system 136 can also be configured
to enable the autonomous vehicle to communication and/or otherwise
receive data from a user device 138 associated with a user 110. The
communications system 136 can utilize various communication
technologies such as, for example, Bluetooth low energy protocol,
radio frequency signaling, etc. In some implementations, the
communications systems 136 can enable the vehicle 104 to function
as a WiFi base station for a user device 138 and/or implement
localization techniques. The communications system 136 can include
any suitable components for interfacing with one or more
network(s), including, for example, transmitters, receivers, ports,
controllers, antennas, and/or other suitable components that can
help facilitate communication.
[0075] The vehicle 104 can include one or more human-machine
interfaces 139. For example, the vehicle 104 can include one or
more display devices located onboard the vehicle 104. A display
device (e.g., screen of a tablet, laptop, etc.) can be viewable by
a user of the vehicle 104 that is located in the front of the
vehicle 104 (e.g., driver's seat, front passenger seat).
Additionally, or alternatively, a display device can be viewable by
a user of the vehicle 104 that is located in the rear of the
vehicle 104 (e.g., back passenger seat(s)).
[0076] In some implementations, the vehicle 104 can be associated
with an entity (e.g., a service provider, owner, manager). The
entity can be one that provides one or more vehicle service(s) to a
plurality of users via a fleet of vehicles that includes, for
example, the vehicle 104. In some implementations, the entity can
be associated with only vehicle 104 (e.g., a sole owner, manager).
In some implementations, the operations computing system 106 can be
associated with the entity. The vehicle 104 can be configured to
provide one or more vehicle services to one or more users. The
vehicle service(s) can include transportation services (e.g.,
rideshare services in which the user rides in the vehicle 104 to be
transported), courier services, delivery services, and/or other
types of services. The vehicle service(s) can be offered to users
by the entity, for example, via a software application (e.g., a
mobile phone software application). The entity can utilize the
operations computing system 106 to coordinate and/or manage the
vehicle 104 (and its associated fleet, if any) to provide the
vehicle services to a user 110.
[0077] The operations computing system 106 can include one or more
computing devices that are remote from the vehicle 104 (e.g.,
located off-board the vehicle 104). For example, such computing
device(s) can be components of a cloud-based server system and/or
other type of computing system that can communicate with the
vehicle computing system 102 of the vehicle 104. The computing
device(s) of the operations computing system 106 can include
various components for performing various operations and functions.
For instance, the computing device(s) can include one or more
processor(s) and one or more tangible, non-transitory, computer
readable media (e.g., memory devices). The one or more tangible,
non-transitory, computer readable media can store instructions that
when executed by the one or more processor(s) cause the operations
computing system 106 (e.g., the one or more processors, etc.) to
perform operations and functions, such as coordinating vehicles to
provide vehicle services.
[0078] For example, a user 110 can request a vehicle service
provided by the vehicle 104. For instance, a user can provide
(e.g., via a user device 138) data indicative of a request 140 to
the operations computing system 106 (e.g., of the entity that is
associated with the vehicle 104). In some implementations, the
request 140 can be generated based at least in part on user input
to a user interface displayed on the user device 138 (e.g., a user
interface associated with a software application of the entity).
The request 140 can indicate the type of vehicle service that the
user 110 desires (e.g., a transportation service, a delivery
service, a courier service, etc.) and a location associated with
the user 110 (e.g., a current location of the user, a different
location, etc.). The request 140 can also include an identifier
(e.g., phone number, Bluetooth, WiFi, Cellular, IP address, other
information, etc.) associated with the user device 138 that
provided the request 140 (and/or other user device). The identifier
can be used by the vehicle computing system 102 to communicate with
the user device 138 and/or otherwise provide/obtain data associated
therewith, as further described herein. In some implementations,
such an identifier can be retrieved from a memory that securely
stores such information in a profile/account associated with the
user 110 (e.g., such that the request 140 need not provide the
identifier).
[0079] The operations computing system 106 can process the request
140 and select the vehicle 104 to provide the requested vehicle
service to the user 110. The operations computing system can
provide, to the vehicle 104, data 142 indicative of a location to
which the vehicle 104 is to travel. For example, FIG. 2 depicts an
example geographic area 200 that includes a location 202 associated
with the user 110 according to example embodiments of the present
disclosure. The location 202 can be associated with the user 110
that requested the vehicle service. For example, the location 202
can be the current location of the user 110, as specified by the
user 110 and/or determined based on user device location data
(e.g., provided with the request 140 and/or otherwise obtained).
The location 202 can also be a location that is different than a
current location of a user 110, such as for example a location at
which the user 110 would like to be picked-up by the vehicle 104,
provide an item to the vehicle 104, retrieve an item from the
vehicle 104, and/or otherwise interact with the vehicle 104. The
location 202 can be expressed as a coordinate (e.g., GPS
coordinate, latitude-longitude coordinate pair), an address, a
place name, and/or another geographic reference that can be used to
identify the location 202. The location 202 associated with the
user 110 can be represented, for example, as a pin on a map user
interface.
[0080] The vehicle computing system 102 can obtain the data 142
indicative of the location 202 associated with the user 110 to
which the vehicle 104 is to travel. As described herein, the user
110 can be associated with a request 140 for a vehicle service
provided by the vehicle 104. The vehicle 104 can obtain the data
142 indicative of the location 202 associated with the user 110
from the operations computing system 106. In some implementations,
the user 110 may communicate directly with the vehicle 104 to
request the vehicle service. For example, the user 110 may use the
user device 138 to send the request 140 to the vehicle computing
system 102. In such a case, the vehicle computing system 102 can
process the request 140 and determine the location 202 of the user
110.
[0081] The vehicle 104 can also obtain a first vehicle route 204
that leads to the location 202 associated with the user 110. The
first vehicle route 204 can be, for example, a route from the
current location of the vehicle 104 to the location 202 associated
with the user 110. In some implementations, the operations
computing system 106 can provide the first vehicle route 204 to the
vehicle 104. Additionally, or alternatively, the vehicle computing
system 102 of the vehicle 104 can determine the first vehicle route
204. For example, the vehicle computing system 102 can determine
the first vehicle route 204 based at least in part on the map data
120.
[0082] The vehicle 104 is to travel in accordance with the first
vehicle route 204 to arrive within a vicinity 206 of the location
202 associated with the user 110. For example, the vehicle
computing system 102 can control the vehicle 104 (e.g., via a
motion plan 134 implemented by the control system(s) 116) to travel
along a first vehicle route 204 to arrive within a vicinity 206 of
the location 202 associated with the user 110. The vicinity 206 of
the location 202 associated with the user 110 can be defined at
least in part by a distance (e.g., a radial distance) from the
location 202 associated with the user 110. The distance can be
indicative of an acceptable walking distance from the location 202
associated with the user 110. For example, an acceptable walking
distance can be distance that a user would be willing walk (or
otherwise travel) to arrive at a vehicle. In some implementations,
the operations computing system 106 can determine the acceptable
walking distance, as described herein, and provide such information
to the vehicle 104. In some implementations, the vehicle 104 can
determine the acceptable walking distance. In some implementations,
the acceptable walking distance can be determined by another
computing system and provided to the operations computing system
106 and/or the vehicle 104.
[0083] The acceptable walking distance can be determined based on a
variety of information. FIG. 3 depicts example information 300
associated with an acceptable walking distance according to example
embodiments of the present disclosure. The acceptable walking
distance 302 can be determined based at least in part on at least
one of current traffic data 304A, historic data 304B, user
preference data 304C, local vehicle weather data 304D, regional
weather data 304E, and/or other data. The current traffic data 304A
can be indicative of a current level of traffic within the
geographic area 200 and/or an area that would affect the geographic
area 200. The current traffic data 304A can be obtained from
another computing system (e.g., city management database, the
operations computing system 106, another vehicle, etc.) and/or
determined by a vehicle 104, as further described herein. For
example, in the event that the current traffic level is high such
that it would take a longer time for the vehicle 104 to reach the
location 202, the acceptable walking distance 302 may be higher so
that the user 110 isn't waiting a greater time to board the vehicle
104, place an item in the vehicle 104, retrieve an item from the
vehicle, etc.
[0084] The historic data 304B can include historic data associated
with providing vehicle services to a user. For instance, the
historic data 304B can be indicative of previously calculated
acceptable walking distances for the specific user 110 and/or for
other user(s) of the vehicle services. By way of example, in the
event that the user (or similarly situated user(s)) has shown a
willingness to walk a certain distance to a vehicle, the acceptable
walking distance 302 can be determined to reflect the historically
acceptable walking distance. In some implementations, the historic
data 3049 can include historic traffic data.
[0085] Additionally, or alternatively, the acceptable walking
distance 302 can be based at least in part on preferences of the
user 110. For instance, the entity associated with the vehicle 104
can maintain an account/profile associated with the user 110. In
some implementations, the user 110 can specify an acceptable
walking distance (e.g., via user input to a user interface). The
user-specified acceptable walking distance can be securely stored
and used to determine the acceptable walking distance 302 when the
user 110 requests a vehicle service. In some implementations, the
user 110 may provide feedback regarding the distance the user 110
walked to arrive at a vehicle that is providing the user 110
vehicle services. The user 110 can provide feedback data (e.g., via
user input to a user interface) indicating whether the distance was
acceptable or unacceptable (e.g., as prompted by a software
application). Accordingly, the acceptable walking distance 302 can
be based at least in part on such feedback data. In some
implementations, acceptable walking distance 302 can be based at
least in part on preferences of other users such as those similar
situated to the user 110, within the geographic area, within a
similar geographic area, etc.
[0086] The acceptable walking distance 302 can also, or
alternatively, be based at least in part on weather data. In some
implementations, the acceptable walking distance 302 can be based
at least in part on local weather data 304D obtained via a vehicle.
For example, the vehicle 104 can include a rain sensor,
thermometer, humidity sensor, and/or other types of sensor(s) that
can be used to determine weather conditions within the surrounding
environment of the vehicle 104. The vehicle 104 can also be
configured to determine the presence of one or more weather
conditions (e.g., rain, sleet, snow, etc.) based at least in part
on the sensor data 118. Additionally, or alternatively, the
acceptable walking distance 302 can be based at least in part on
regional weather data 304E (e.g., from a third party weather
source). The acceptable walking distance 302 can be adjusted
depending on the weather conditions indicated by the weather data.
By way of example, in the event that the local weather data 304D
and/or the regional weather data 304E indicate that the geographic
area 200 is experiencing (and/or will experience) rain, the
acceptable walking distance 302 may be decreased to a short
distance from the location 202. In the event that the local weather
data 304D and/or the regional weather data 304E indicate that the
geographic area 200 is experiencing (and/or will experience) clear
skies, the acceptable walking distance 302 can be greater distance
from the location 202.
[0087] In some implementations, the acceptable walking distance 302
can be determined at least in part from a model, such as a
machine-learned model. For example, the machine-learned model can
be or can otherwise include one or more various model(s) such as,
for example, models using boosted random forest techniques, neural
networks (e.g., deep neural networks), or other multi-layer
non-linear models. Neural networks can include recurrent neural
networks (e.g., long short-term memory recurrent neural networks),
feed-forward neural networks, and/or other forms of neural
networks. For instance, supervised training techniques can be
performed to train the model using historical acceptable walking
distances, weather data, user feedback data, etc. to determine an
acceptable walking distance 302 based at least in part on input
data. The input data can include, for example, the various types of
information 300, as described herein. The machine-learned model can
provide, as an output, data indicative of an acceptable walking
distance 302. In some implementations, the acceptable walking
distance 302 can be specific to the geographic area. For example,
the model can be trained based on information associated with the
geographic area 200 such that the outputted acceptable walking
distance 302 is specific to the geographic area 200.
[0088] In some implementations, the vehicle 104 can provide a
communication to the user 110 indicating that the vehicle 104 is
within the vicinity 206 of the location 202 associated with the
user 110. For example, the vehicle computing system 102 can
determine that the vehicle 104 is within the acceptable walking
distance 302 from the location 202 of the user 110 (e.g., based on
the positioning system 144). In response, the vehicle computing
system 102 can send a communication to the user device 138
associated with the user 110. The communication can indicate that
the vehicle 104 is within the vicinity 206 of the user 110 (e.g.,
within the acceptable walking distance 302). For instance, for a
transportation service, such a communication can be in the form of
a textual message, auditory message, etc. stating "your vehicle is
arriving please prepare to board".
[0089] The vehicle 104 can begin to search for a parking location
(e.g., out of the vehicle's travel way) when the vehicle 104 is
within the vicinity of the location 202 associated with the user
110. For example, the vehicle 104 can enter into the approach
operating mode 108D when the vehicle 104 is within the vicinity 206
of the location 202 associated with the user 110. While within the
vicinity 206, the vehicle 104 can search for a parking location
before it reaches the location 202 associated with the user 110
(e.g., before the GPS pin coordinate on a map). In some
implementations, the vehicle 104 can search for a parking location
after it passes the location 202 associated with the user 110 and
is still within the vicinity 206 of the user 110 (e.g., within the
acceptable walking distance 302).
[0090] To identify a parking location, the vehicle computing system
102 can obtain sensor data 118 associated with one or more objects
that are proximate to the vehicle 104. As described herein, the
sensor data 118 can be indicative of locations associated with the
object(s) within the surrounding environment of the vehicle 104 at
one or more times. The vehicle 104 can process the sensor data 118
to determine if there are any available parking locations out of
the vehicle's travel way (e.g., out of a traffic lane) that are not
currently occupied by the objects (e.g., other vehicles) within the
vehicle's surrounding environment. In some implementations, the
vehicle 104 can utilize the map data 120 to determine whether any
designated parking locations (e.g., parking lots, pullover lanes,
etc. are located and/or are available within the vicinity 206 of
the location 202 associated with the user 110. In the event the
vehicle 104 can identify a parking location that is out of a travel
way and within the vicinity 206 of the location 202 associated with
the user 110 (e.g., out of a traffic lane), the vehicle 104 can
position itself into that parking location accordingly.
[0091] The vehicle computing system 102 can send a communication to
the user device 138 associated with the user 110 indicating that
the vehicle 104 has parked. For example, FIG. 4 depicts an example
display device 400 with an example communication 402 according to
example embodiments of the present disclosure. The display device
400 (e.g., display screen) can be associated with the user device
138 associated with the user 110. The communication 402 can be
presented via a user interface 404 on the display device 400. The
communication 402 can indicate that the vehicle 104 has arrived. In
some implementations, the communication 402 and/or another portion
of the user interface 404 can be indicative of a location of the
vehicle 104. For example, the display device 400 can display a map
user interface 406 that includes a user route 408. The user route
408 can be a route along which a user 110 can travel to arrive at
the vehicle 104.
[0092] In the event that the vehicle 104 is unable to identify a
parking location that is out of the travel way (and within the
vicinity 206 of the user 110), the vehicle 104 can decide whether
or not to stop at least partially in a travel way to wait for the
user 110. To help do so, the vehicle computing system 102 of the
vehicle 104 can obtain traffic data associated with the geographic
area 200 that includes the location 202 associated with the user
110.
[0093] FIG. 5 depicts example traffic data 500 that can be obtained
by the vehicle computing system 102 according to example
embodiments of the present disclosure. The traffic data 500 can be
obtained from a variety of sources such as other vehicles (e.g.,
other autonomous vehicles within the vehicle fleet), the operations
computing system 106, third party sources (e.g., traffic management
entities, etc.), as well as the vehicle 104 itself. The traffic
data 500 can include, for example, in-lane traffic data 502A,
out-of-lane traffic data 502B, other vehicle traffic data 502C,
current wider traffic data 502D, historic traffic data 502E, and/or
other types of traffic data. The traffic data 500 can be updated
periodically, as scheduled, upon request, in real-time, and/or in
near real-time.
[0094] The in-lane traffic data 502A and/or out-of-lane traffic
data 502B can be indicative of a level of traffic within the
surrounding environment of the vehicle 104. For instance, the
in-lane traffic data 502A can be indicative of the number of
objects, object locations, and the speed of the respective objects
within the current travel lane (and/or other designated travel
boundaries) of the vehicle 104 (e.g., the other vehicles to the
rear and front of the vehicle 104). The out-of-lane traffic data
502B can be indicative of the number of objects, object locations,
and the speed of the respective objects within the surrounding
environment, other than in the current travel lane (or other
boundaries) of the vehicle 104 (e.g., other vehicles in the other
lanes, all other classified objects around the vehicle 104, etc.).
The in-lane traffic data 502A and/or out-of-lane traffic data 502B
can be based at least in part on the sensor data 118 associated
with the surrounding environment of the vehicle 104, as further
described herein.
[0095] The vehicle computing system 102 can obtain the other
vehicle traffic data 502C from one or more other vehicles, such as
one or more other autonomous vehicles in a fleet that includes the
vehicle 104. The other vehicle traffic data 502C can be indicative
of the number of objects, object locations, and the speed of the
respective objects within the surrounding environment of the other
vehicle(s), while the other vehicle(s) are within the geographic
area 200, the vicinity 206 of the location 202 associated with the
user 110, and/or another location which may have a traffic effect
on the geographic area 200 and/or the vicinity 206 of the location
202 associated with the user 110. The vehicle computing system 102
can obtain the other vehicle traffic data 502C via vehicle to
vehicle communication and/or via other computing device(s) (e.g.,
the operations computing system 106).
[0096] The current wider traffic data 502D can be indicative of the
traffic within a larger regional area that includes the geographic
area 200. For instance, the vehicle computing system 102 can obtain
the current wider traffic data 502D that indicates the current
traffic patterns, build-ups, etc. within a region that includes the
geographic area 200. Such data can be obtained via a third patty
source such as, for example, a traffic management entity associated
with the region.
[0097] The historic traffic data 502E can include previously
collected traffic data of the types of traffic data described
herein and/or other types of traffic data. For example, the
historic traffic data 502E can include traffic data previously
obtained by the vehicle computing system 102 (e.g., in-lane,
out-of-lane traffic data), traffic data previously obtained by
other vehicle(s) (e.g., associated with the geographic area 200,
the vicinity 206 of the location 202, etc.), historic traffic data
(e.g., traffic patterns) associated with a region that includes the
geographic area 200, and/or other types of historic traffic
data.
[0098] The vehicle computing system 102 can determine an estimated
traffic impact 504 of the vehicle 104 on the geographic area 200
based at least in part on the traffic data 500. The estimated
traffic impact 504 can be indicative of an estimated impact of the
vehicle 104 on one or more objects within a surrounding environment
of the vehicle 104 in the event that the vehicle 104 were to stop
at least partially in the travel way within the vicinity 206 of the
location 202 associated with the user 110. In some implementations,
the estimated traffic impact 504 can estimate the likelihood that
an approach object (e.g., another vehicle) can pass safely without
endangering the user 110.
[0099] In some implementations, the estimated traffic impact 504
can be based at least in part on a comparison of a level of traffic
to a traffic constraint 506. The vehicle computing system 102 can
determine a level of traffic (e.g., a number of objects that could
be impacted by the vehicle 104 stopping at least partially in a
travel way) within the surrounding environment of the vehicle 104
based at least in part on one or more of the types of traffic data
500. The vehicle computing system 102 can compare the level of
traffic to a traffic constraint 506 to determine whether the
estimated traffic impact 504 would be high or low (e.g.,
significant or insignificant, unacceptable or acceptable,
etc.).
[0100] The traffic constraint 506 can be implemented in a variety
of forms. For example, the traffic constraint 506 can include a
traffic threshold that is indicative of an acceptable level of
traffic (e.g., an acceptable number of objects) that would be
impacted by the vehicle 104 stopping at least partially in the
travel way. A level of traffic that exceeds the traffic threshold
would be considered a high impact on traffic. In some
implementations, the traffic constraint 506 can be implemented as
cost data (e.g., one or more cost function(s)). For example, the
vehicle computing system 102 (e.g., the motion planning system 128)
can include cost data that reflects the cost(s) of stopping vehicle
motion, the cost(s) of causing traffic build-up, the cost(s) of
illegally stopping in a travel way, etc.
[0101] The traffic constraint 506 can be based on a variety of
information. In some implementations, the traffic constraint 506
can be based at least in part on historic traffic data that
indicates the level of traffic previously occurring in that
geographic area. For example, if the geographic area 200 normally
experiences a high level of traffic build-up, a corresponding
traffic threshold can be higher (and/or the cost of stopping can be
lower). Additionally, or alternatively, the traffic constraint 506
can be based at least in part on real-time traffic data (e.g., from
other vehicles in the fleet, from the vehicle 104, other sources).
For example, in the event that there is already a traffic jam in
the vicinity 206 of the location 202 of the user 110, the traffic
threshold could be higher (and/or the cost of stopping could be
lower). In some implementations, the traffic constraint 506 can be
based at least in part on the typical travel expectations of
individuals in the geographic area 200. For example, individuals
that are located in City A may he more patient when waiting in
traffic than those in City B. Thus, a traffic threshold may be
higher in City A than in City B (and/or the cost of stopping may be
lower in City A than in City B). In some implementations, the
traffic constraint 506 can be based at least in part on map data
120 (and/or other map data). For example, in the event that the
vehicle 104 is traveling on a wide travel way in which impacted
vehicles could eventually travel around the vehicle 104, the
traffic threshold could be higher (and/or the cost of stopping
could be lower). The traffic constraint 506 can be determined
dynamically, in real-time (and/or near real-time) to reflect the
conditions currently faced by the vehicle 104.
[0102] In some implementations, the traffic constraint 506 can be
determined at least in part from a model, such as a machine-learned
model. For example, the machine-learned model can he or can
otherwise include one or more various model(s) such as, for
example, models using boosted random forest techniques, neural
networks (e.g., deep neural networks), or other multi-layer
non-linear models. Neural networks can include recurrent neural
networks (e.g., long short-term memory recurrent neural networks),
feed-forward neural networks, and/or other forms of neural
networks. For instance, supervised training techniques can he
performed to train the model (e.g., using previous driving logs) to
determine a traffic constraint 506 based at least in part on input
data. The input data can include, for example, the traffic data 500
as described herein, map data, data from a traffic management
entity, driving characteristics of individuals in an associated
geographic area, complaints received from operators of vehicles
that were caused to stop by autonomous vehicles, etc. The
machine-learned model can provide, as an output, data indicative of
a recommended traffic constraint. In some implementations, the
recommended traffic constraint can be specific to the geographic
area 200. For example, the model can be trained based at least in
part on data associated with geographic area 200 such that the
recommended traffic constraint is specific for that particular
region.
[0103] Additionally, or alternatively, the estimated traffic impact
504 can be based at least in part on a model. The vehicle computing
system 102 and/or the operations computing system 106 can obtain
data descriptive of the model (e.g., machine learned model). The
vehicle computing system 102 and/or the operations computing system
106 can provide input data to the model. The input data can include
the one or more of the types of traffic data 500 (e.g., associated
with the geographic area 200). In some implementations, the input
data can include the map data 120. The model can determine the
estimated traffic impact 504 that the vehicle 104 would have on the
geographic area 200 if the vehicle 104 were to stop at least
partially within a travel way (e.g., within a current lane of
travel). By way of example, the model can evaluate the traffic data
500 to determine the level of traffic in the vehicle surrounding
environment. The model can analyze the traffic data 500 with
respect to the traffic constraint 506 to determine whether the
estimated traffic impact 504 is high or low, significant or
insignificant, unacceptable or acceptable, etc. The output of such
a model can be the estimated traffic impact 504, which can be
indicative of, for example, a number of objects that would be
impacted by the vehicle 104 stopping at least partially in a travel
way and/or whether the estimated traffic impact is high or low,
significant or insignificant, acceptable or unacceptable, etc.
[0104] In some implementations, the output of the model can be
provided as an input to the model for another set of traffic data
(e.g., at a subsequent time step). In such fashion, confidence can
be built that a determined estimated traffic impact is the
accurate. Stated differently, in some implementations, the process
can be iterative such that the estimated traffic impact can be
recalculated over time as it becomes clearer what the estimated
traffic impact is on the respective geographic area. For example,
the model can include one or more autoregressive models. In some
implementations, the model can include one or more machine-learned
recurrent neural networks. For example, recurrent neural networks
can include long or short-term memory recurrent neural networks,
gated recurrent unit networks, or other forms of recurrent neural
networks.
[0105] As described herein, the estimated traffic impact 504 can be
based at least in part on the sensor data 118 acquired onboard the
vehicle 104. For instance, the vehicle computing system 102 can
obtain traffic data associated with the geographic area 200 that
includes the location 202 associated with the user 110 based at
least in part on the sensor data 118 obtained via the one or more
sensors 112. By way of example, the vehicle computing system 102
can obtain the in-lane traffic data 502A and/or out-of-lane traffic
data 502B associated with the geographic area 200 that includes the
location 202 associated with the user 110. In some implementations,
the vehicle computing system 102 can obtain sensor data 118 (e.g.,
via the one or more sensors 112) associated with the surrounding
environment of the vehicle 104 that is within the vicinity 206 of
the location 202 associated with the user 110, as described herein.
The sensor data 118 can be indicative of one or more objects within
the surrounding environment of the vehicle 104 that is within the
vicinity 206 of the location 202. The vehicle computing system 102
can process the sensor data 118 to classify which of the object(s)
would be impacted (e.g., caused to stop) by the vehicle 104
stopping in a travel way.
[0106] By way of example, FIG. 6 depicts an example travel way 600
according to example embodiments of the present disclosure. The
travel way 600 can be associated with the geographic area 200. For
example, the travel way 600 can be located within the vicinity 206
of the location 202 associated with the user 110 (e.g., the street
on which the user 110 is located). The travel way 600 can include
the current travel lane 602 (and/or other designated travel
boundaries) of the vehicle 104. The travel way 600 can include
other travel lanes, such as other lane 604 (e.g., a lane adjacent
to the current travel lane 602). The vehicle computing system 102
can classify the objects within the surrounding environment that
would be impacted in the event the vehicle 104 were to stop at
least partially in the travel way 600. For example, the object 606
(e.g., another vehicle behind the vehicle 104 in the same travel
lane 602) can be identified as an object that would be impacted in
the event that the vehicle 104 stops at least partially in the
current lane 602.
[0107] The vehicle computing system 102 can also identify object(s)
that would not be affected by the vehicle 104 stopping at least
partially in the travel way 600. Such object(s) can include one or
more objects in another travel lane as well as objects with the
same travel lane as the vehicle 104. For example, the vehicle
computing system 102 can determine that the object 608 located in
the other travel lane 604 (e.g., another vehicle in an adjacent
travel lane) will not be impacted by the vehicle 104 stopping at
least partially in the travel way 600 because the object 608 can
continue passed the vehicle 104 via the other travel lane 604. In
another example, the vehicle computing system 102 can determine
that an object 610 (e.g., a bicycle) may not be impacted by the
vehicle 104 stopping (and/or may be impacted to an insignificant
degree) because the object 608 may be parked and/or have the
opportunity to travel around the vehicle 104 (e.g., via a lane
change into the adjacent lane, maneuver around the vehicle 104
within the same lane, etc.).
[0108] After such classification, the vehicle computing system 102
can determine a level of traffic associated with the geographic
area 200 (e.g., within the vicinity of the user's location) based
at least in part on the sensor data 118. For example, the level of
traffic can be based at least in part on the number of object(s)
within the surrounding environment of the vehicle 104 that would be
impacted by the vehicle 104 stopping at least partially in the
travel way 600 (e.g., the current travel lane 602), while filtering
out those object(s) that would not be impacted. Similar such
information could be acquired via one or more other vehicles in the
associated vehicle fleet. The vehicle computing system 102 can
determine the estimated traffic impact by comparing the level of
traffic to the traffic constraint 506 (e.g., a traffic threshold
indicative of a threshold level of traffic). In the event that the
level of traffic exceeds the traffic constraint 506, the vehicle
computing system 102 can determine that the estimated traffic
impact 504 on the geographic area 200 would be high.
[0109] The vehicle computing system 102 can also, or alternatively,
determine an estimated time of user arrival. The estimated time of
user arrival can be indicative of an amount of time needed for the
user 110 to interact with the vehicle 104 (e.g., before the vehicle
104 can begin moving again). For example, the vehicle computing
system 102 can determined an estimated time until the user 110
completes boarding of the vehicle 104 (e.g., enters the vehicle 104
and fasten a seatbelt for a transportation service). This can be
made up of an estimated time until the user 110 starts boarding the
vehicle 104 (e.g., the time needed to walk to the vehicle 104, the
time until the vehicle doors unlock, etc.) and an estimated time of
boarding duration for a user 110 (e.g., to load luggage, help other
passengers board, etc.). In some implementations, the vehicle
computing system 102 can determine an estimated time until the user
110 completes the retrieval of an item from the vehicle 104 (e.g.,
completely removes an item from the vehicle 104 for a delivery
service). In some implementations, the vehicle computing system 102
can determine an estimated time until the user 110 places an item
in the vehicle 104 (e.g., for a courier service). The estimated
time estimated time of user arrival (e.g., estimated time until the
user 110 completes boarding) can help the vehicle computing system
102 determine whether or not to stop at least partially in the
travel way 600. Such time estimate(s) can be expressed as a time
duration (e.g., user estimated to arrive in 1 minute) and/or a
point in time (e.g., user estimated to arrive at 10:31 am
(PT)).
[0110] The estimated time of user arrival (e.g., the estimated time
until the user 110 completes boarding, item retrieval, item
placement, etc.) can be based on a variety of information. FIG. 7
depicts example depicts a flow diagram of an example method 700 of
determining an estimated time of user arrival (e.g., autonomous
vehicle user boarding times) according to example embodiments of
the present disclosure. While the following provides examples of
the method 700 with respect to determining autonomous vehicle user
boarding times, a similar approach can be taken for determining an
estimated time until the user 110 completes the retrieval of an
item from the vehicle 104 and/or an estimated time until the user
110 places an item in the vehicle 104. One or more portion(s) of
the method 700 can be implemented by one or more computing devices
such as, for example, the one or more computing device(s) of the
vehicle computing system 102 and/or other systems (e.g., as
computing operations). Each respective portion of the method 700
can be performed by any (or any combination) of the one or more
computing devices. Moreover, one or more portion(s) of the method
700 can be implemented as an algorithm on the hardware components
of the device(s) described herein (e.g., as in FIGS. 1 and 14), for
example, to control an autonomous vehicle. FIG. 7 depicts elements
performed in a particular order for purposes of illustration and
discussion. Those of ordinary skill in the art, using the
disclosures provided herein, will understand that the elements of
any of the methods discussed herein can be adapted, rearranged,
expanded, omitted, combined, and/or modified in various ways
without deviating from the scope of the present disclosure.
[0111] At (702), the method 700 can include obtaining one or more
identifiers of a user device. The vehicle computing system 102 can
use one or more identifier(s) of the user device 138 (e.g.,
obtained with the request 140, obtained by the vehicle 104,
provided by the operations computing system 106, etc.) to scan for
and/or communicate with the user device 138 when the vehicle 104 is
within the vicinity 206 of the user 110. By way of example, the
operations computing system 106 can provide the vehicle computing
system 102 with one or more identifier(s) of a user device 138
associated with the user 110. The identifiers can be, for example,
unique radio identifiers (e.g., Bluetooth, WiFi, Cellular, friendly
names, contact number, other identifier) collected via a software
application and provided to the operations computing system 106.
The vehicle computing system 102 can utilize the identifiers to
communicate and/or locate a user device 138 associated with the
user 110. For example, when the vehicle 104 is within a vicinity
206 of the location 202 associated with the user 110, the vehicle
computing system 102 can scan for the user device 138 (e.g., opt-in
radios) based at least in part on the identifier(s) (e.g., when the
vehicle 104 is in the approach mode 108D).
[0112] At (704), the method 700 can include obtaining location data
associated with the user device. For instance, the vehicle
computing system 102 can obtain (e.g., via the communication system
136) location data associated with a user device 138 associated
with a user 110. The location data associated with the user device
138 can be indicative of one or more locations of the user device
138 associated with the user 110, at one or more times. In some
implementations, the vehicle computing system 102 can use the
identifier(s) to determine the location of the user 110. The
vehicle computing system 102 can be configured to utilize a variety
of communication technologies to obtain the location data
associated with the user device 138. For example, the vehicle
computing system 102 can obtain the location data based at least in
part on a triangulation of signals via at least one of multiple
input, multiple output communication between the vehicle 104 and
the user device 138, one or more Bluetooth low energy beacons
located onboard the vehicle 104, or a light signal handshake
between the user device 138 and the vehicle 104, as further
described herein.
[0113] In some implementations, the vehicle computing system 102
can use radio frequency (RF) signaling to obtain location data
associated with the user 138. For example, FIG. 8A depicts an
example portion 800 of a communications system 136 according to
example embodiments of the present disclosure. The communications
system 136 can include one or more electronic devices 802 (e.g., RF
module) configured to transmit and/or obtain one or more RF signals
(e.g., from the user device 138) via one or more communication
device(s) 804 (e.g., transmitters, receivers, transmitters, RF
sensors, etc.). For example, once the user device 138 is found
(e.g., via the identifiers), the vehicle computing system 102 can
track changes in the signal strength (e.g., radio signal strength
identifier (RSSI)) to determine various information about the user
device 138. For example, the vehicle computing system 102 can
determine the approximate distance of the user device 138 (and the
user 110) to the vehicle 104 (e.g., the difference between RSSI
triangulation and improvements over time can be used to determine
the approximate distance of the user 110 to the vehicle 104).
Additionally, or alternatively, the vehicle computing system 102
can determine a heading of the user 110 based at least in part on
RF signal(s) associated with the user device 138 (e.g., the
communication device(s) 804 with the best/strongest RSSI reading
can indicate direction/heading of the user 110). The vehicle
computing system 102 can also determine a speed/velocity of the
user device 138 using such a technique. The approximate distance,
heading, speed, etc. measurements can be determined with or without
an authenticated connection.
[0114] Additionally, or alternatively, the vehicle computing system
102 can utilize Bluetooth low energy protocol to obtain location
data associated with the user device 138. For example. FIG. 8B
depicts an example portion 810 of a communications system 136
according to example embodiments of the present disclosure. The
communications system 136 can include one or more Bluetooth low
energy beacons 812 that are located onboard the vehicle 104. The
vehicle computing system 102 can obtain the location data
associated with the user device 138 based at least in part on the
one or more Bluetooth low energy beacons 812 located onboard the
vehicle 104, in some implementations, the vehicle computing system
102 can determine differences in Bluetooth Low Energy (BLE) beacon
radio signal strength identifiers over time and/or inertial
measurement unit changes, which can indicate a distance between the
user device 138 of the user 110 (e.g., a mobile phone associated
with the user 110) and the vehicle 104. Additionally, or
alternatively, the vehicle computing system 102 can determine a
heading of the user device 138 (and the user 110) based at least in
part on the Bluetooth low energy beacon(s) 812 (e.g., the beacon(s)
with the best/strongest RSSI reading can indicate direction/heading
of the user 110). The vehicle computing system 102 can also
determine a speed/velocity of the user device 138 using such a
technique. In some implementations, the user device 138 can
determine its location (e.g., relative to the vehicle 104) based at
least in part on signals transmitted from one or more beacons 812
located onboard the vehicle 104. The user device 138 can provide
data indicative of its location to one or more remote computing
device(s) (e.g., the operations computing system 106, cloud-based
system). The remote computing device(s) can provide data indicative
of the location of the user device 138 (determined based on the
beacons 812) to the vehicle computing system 102. The vehicle
computing system 102 can determine the location of the user device
138 (and the user 110) based on such data. In some implementations,
the remote computing device(s) can process data from the user
device 138 (e.g., associated with the beacons 812) to determine a
location of the user device 138 and provide data associated
therewith to the vehicle computing system 102.
[0115] In some implementations, the vehicle computing system 102
can utilize one or more altimeters (and/or other measuring device)
to obtain location data associated with the user device 138. For
example, FIG. 8C depicts an example diagram 820 of obtaining
location data according to example embodiments of the present
disclosure. The vehicle 104 can include one or more altimeters 822
located onboard the vehicle 104. The user device 138 can include
one or more altimeters 824. As shown, the user device 138 (and the
user 110) can be located on the second story of a building 826. The
vehicle computing system 102 can obtain location data associated
with the user device 138 via at least one altimeter 822 located
onboard the vehicle 104. By way of example, the vehicle computing
system 102 can obtain location data associated with altimeter(s)
824 of the user device 138 via one or more network(s) 828. The
vehicle computing system 102 can compare the location data
associated with altimeter(s) 824 of the user device 138 to location
data associated with altimeter(s) 822 of the vehicle 104. The
vehicle computing system 102 can determine an elevation/altitude of
the user device 138 (and the user 110) based at least in part on
this comparison (e.g., a difference between the altimeter(s) 824 of
the user device 138 and the altimeter(s) 822 of the vehicle 104 can
indicate elevation/altitude difference). The elevation/altitude can
be relative to the position of the vehicle 104 and/or another
reference (e.g., ground level, sea level. etc.).
[0116] In some implementations, the vehicle computing system 102
can utilize other communication technologies to obtain location
data associated with the user device 138. For example, the vehicle
computing system 102 can obtain the location data associated with
the user device 138 based at least in part on multiple input,
multiple output communication between the vehicle 104 and the user
device 138 associated with the user 110. This can allow the vehicle
104 to take advantage of the multiple antennas included in the
vehicle's communication system 136 as well as those of the user
device 138 to increase accuracy of the location data 704A
associated with the user 110. For example, the user device 138 can
obtain an identifier (e.g., Radio Network Temporary Identifier)
that can be associated with the user device 138. The user device
138 can provide data indicative of the identifier to one or more
remote computing device(s) (e.g., the operations computing system
106, cloud-based system). The remote computing device(s) can
provide data indicative of the identifier to the vehicle computing
system 102. The vehicle computing system 102 can locate the user
device 138 (and the user 110) based at least in part on the
identifier via the multiple antennas onboard the vehicle 104
(and/or the antenna(s) of the user device 138). In some
implementations, the vehicle computing system 102 can utilize a
handshake (e.g., light signal handshake) between the user device
138 and the vehicle 104.
[0117] In some implementations, the vehicle computing system 102
can obtain location data associated with the user device 138 based
at least in part on image data. For example, the user 110 can
obtain image data associated with the user 110 (e.g., via the user
device 138). The image data can be indicative of one or more
characteristics (e.g., buildings, street signs, etc) of the
geographic area and/or surrounding environment of the user 110. In
some implementations, the user 110 may be included in the image
data. The user device 138 can process the image data to determine
the location of the user 110 (e.g., via a comparison of image data
features to known geographic features). The user device 138 can
provide the determined location to one or more remote computing
device(s) (e.g., the operations computing system 106 and/or other
cloud-based system). The remote computing device(s) can provide
data indicative of the location of the user 110 to the vehicle
computing system 102. The vehicle computing system 102 can obtain
the data indicative of the location of the user 110 determined
based at least in part on image data associated with the user 110.
The estimated time until the user 110 starts boarding the vehicle
104 can be based at least in part on the location of the user 110.
In some implementations, the remote computing device (e.g., the
operations computing system 106 and/or other cloud-based system)
can obtain the image data associated with the user 110, process the
image data (e.g., as described herein), and provide the data
indicative of the location of the user 110 to the vehicle computing
system 102. This can be helpful to save the processing resources of
a computationally limited device (e.g., mobile device).
[0118] In some implementations, the vehicle computing system 102
can obtain the image data associated with the user 110 (e.g., via
the user device 138 and/or the remote computing device). The
vehicle computing system 102 can determine a location of the user
110 based at least in part on the image data. For example, the
vehicle computing system 102 can analyze the features of the image
data (e.g., the background, street signs, buildings, etc.) and
compare the features to other data (e.g., sensor data 118, map data
120, other data, etc.). The vehicle computing system 102 can
determine a location of the user based at least in part on this
comparison.
[0119] In some implementations, the vehicle computing system 102
can utilize other communication techniques. These techniques can
include, for example, vehicle perception of the user 110 (e.g., via
processing of sensor data 118 to perceive the user 110 and the
user's location, distance, heading, velocity, and/or other state
data 130 associated therewith), GPS location of the user device
138, device specific techniques (e.g., specific device/model type),
the vehicle 104 serving as a localized base station (e.g., GPS,
WiFi, etc.), and/or other techniques.
[0120] Returning to FIG. 7, the method 700 can include determining
an estimated time until the user starts interaction with the
vehicle, at (706). For instance, the vehicle computing system 102
can determine an estimated time until the user starts boarding the
vehicle 104 (ETSB) based at least in part on the location data
associated with the user device 138. The estimated time of until
the user 110 starts interaction with the vehicle 104 (e.g., the
estimated time until the user starts boarding the vehicle 104
(ETSB)) can be indicative of, for example, a countdown in time
until the user's location is near/at the vehicle 104. until the
user 110 begins to board, until the doors of the vehicle 104 are
unlocked, until the doors are opened, etc. The user 110 can, but
need not, physically interact with the vehicle 104 for this time
estimate.
[0121] The estimated time until the user 110 starts interaction
with the vehicle 104 (e.g., the estimated time until the user
starts boarding the vehicle 104 (ETSB)) can be based on a variety
of data. For example, the vehicle computing system 102 can
determine a distance 708 between the user 110 and the vehicle 104
based at least in part on the location data associated with the
user device 138, as described herein. The estimated time until the
user 110 starts interaction with the vehicle 104 (e.g., the
estimated time until the user starts boarding the vehicle 104
(ETSB)) can be based at least in part on the distance 708 between
the user 110 and the vehicle 104. Additionally, or alternatively,
the vehicle computing system 102 can determine an
elevation/altitude 710 of the user 110 based at least in part on
the location data associated with the user device 138 (e.g.,
obtained via the one or more altimeters onboard the vehicle 104),
as described herein. The estimated time until the user 110 starts
interaction with the vehicle 104 (e.g., the estimated time until
the user starts boarding the vehicle 104 (ETSB)) can be based at
least in part on the elevation/altitude 710 of the user 110 (e.g.,
if the user 110 is in a tall building then additional time should
be added). Additionally, or alternatively, the vehicle computing
system 102 can determine a heading 712 of the user 110 based at
least in part on the location data associated with the user device
138, as described herein. The estimated time until the user 110
starts interaction with the vehicle 104 (e.g., the estimated time
until the user starts boarding the vehicle 104 (ETSB)) can be based
at least in part on the heading 712 of the user 110 (e.g., if the
user 110 is across the street then additional time can be added).
Additionally, or alternatively, the vehicle computing system 102
can determine a location of the user 110 based at least in part on
image data, as described herein. The estimated until the user 110
starts interaction with the vehicle 104 (e.g., the estimated time
until the user starts boarding the vehicle 104 (ETSB)) can be based
at least in part on the location of the user 100, as determined
from the image data.
[0122] In some implementations, the vehicle computing system 102
can obtain historic data 714 to help determine the estimated time
until the user 110 starts interaction with the vehicle 104 (e.g.,
the estimated time until the user starts boarding the vehicle 104
(ETSB)). The historic data 714 can be indicative of, for example,
historic start boarding times of one or more other users (and or
the user 110). In some implementations, a machine learned model can
be trained based on such historic data to determine the estimated
time until the user 110 starts interaction with the vehicle 104
(e.g., the estimated time until the user starts boarding the
vehicle 104 (ETSB)). For example, the model can be trained based on
training data indicative of previous location data, user distances,
altitudes, headings, etc. labeled with the times of user arrival
(e.g., the times when the users started boarding the vehicle). The
model can be trained to receive input data location data, user
distances, altitudes, headings, etc.) and provide, as an output, an
until the user 110 starts interaction with the vehicle 104 (e.g.,
an estimated time until the user starts boarding a vehicle (ETSB)).
In some implementations, the estimated time until the user 110
starts interaction with the vehicle 104 (e.g., the estimated time
until the user starts boarding the vehicle 104 (ETSB)) can be based
at least in part on the historic data 714. For example, vehicle
computing system 102 can determine the estimated time until the
user 110 starts interaction with the vehicle 104 (e.g., the
estimated time until the user starts boarding the vehicle 104
(ETSB)) based at least in part on such a machine-learned model
(e.g., as an output thereto).
[0123] In some implementations, the estimated time until the user
110 starts interaction with the vehicle 104 (e.g., the estimated
time until the user starts boarding the vehicle 104 (ETSB)) can he
used to determine one or more vehicle actions, at (716). For
example, the estimated time until the user 110 starts interaction
with the vehicle 104 (e.g., the estimated time until the user
starts boarding the vehicle 104 (ETSB)) can indicate that the user
110 is close to the vehicle 104 and/or heading toward the vehicle
104. The vehicle computing system 102 can cause one or more doors
of the vehicle 102 to unlock based at least in part on the
estimated time until the user 110 starts interaction with the
vehicle 104 (e.g., the estimated time until the user starts
boarding the vehicle 104 (ETSB)). To do so, the vehicle computing
system 102 can provide control signal(s) to an associated door
controller. Additionally, or alternatively, the vehicle computing
system 102 can cause the vehicle 104 to implement one or more
vehicle settings (e.g., temperature, music, etc.) associated with
the user 110 based at least in part on the estimated time until the
user 110 starts interaction with the vehicle 104 (e.g., the
estimated time until the user starts boarding the vehicle 104
(ETSB)). For example, the vehicle computing system 102 can access a
profile associated with the user 110 to identity the user's
preferred vehicle settings and can provide one or more control
signals to the appropriate systems onboard the vehicle 104 (e.g.,
temperature control system, sound system etc.) to implement the
vehicle settings.
[0124] At (718), the method 700 can include determining an
estimated time of interaction duration between the user and the
vehicle. This estimated time can be indicative of the time it will
take for the user 110 to interact with the vehicle 102. By way of
example, the estimated time of interaction duration can include an
estimated board duration (EBD) that is indicative of how long the
user 110 may take to load him/herself, children, luggage, etc. into
the vehicle 104, to securely fastening seatbelts, and/or undertake
other tasks (e.g., for a transportation service). In some
implementations, the estimated time of interaction duration can be
indicative of how long the user 110 may take to unsecure and remove
an item from the vehicle 104 (e.g., for a delivery service). In
some implementations, the estimated time of interaction duration
can be indicative of how long the user 110 may take to securely
place an item into the vehicle 104 (e.g., for a courier
service).
[0125] To help determine the estimated time of interaction duration
between the user 110 and the vehicle 104, the vehicle computing
system 102 can obtain data associated with the user 110. The
vehicle computing system 102 can determine the estimated time of
interaction duration (e.g., an estimated time of boarding duration
for the user 110) based at least in part on the data associated
with the user 110. The data associated with the user 110 can
include, for example, data indicative of one or more preferences
720 of the user 110 and/or one or more vehicle service parameters
722 associated with the user 110. The preferences 720 can be
indicative of the user's destination (e.g., airport, train station,
etc.), service type, and/or other information specified by the user
110 (e.g., when requesting the vehicle service). The one or more
vehicle service parameters 722 can be indicative of number of
passengers, child's car seat request, presence/amount of luggage,
etc. The vehicle service parameters 722 can also be specified by
the user 110 (e.g., when requesting the vehicle service).
[0126] In some implementations, the data associated with the user
110 can include historic data 724 (e.g., indicative of a boarding
behavior associated with one or more other users). For example, the
historic data 724 can be indicative of historic wait time(s)
associated with other users in the geographic area 200 and/or a
greater region, similarly situated users, etc. In some
implementations, the historic data 724 can be associated with the
specific user 110. The historic data 724 can include, for example,
previous correlations between changes in the signal strength of an
identifier and a user's time to arriving at a vehicle. For example,
the historic data 724 can indicate historic RSSI changes as a
countdown to user arrival.
[0127] At (726), the method 700 can include determining an
estimated time of user arrival with the vehicle 104. The estimated
time of user arrival can include, for example, an estimated time
until the user completes hoarding of the vehicle 104 (ETCB) (e.g.,
for a transportation service), an estimated time until the user 110
finishes retrieving an item from the vehicle 104 (e.g., for a
delivery service), and estimated time until the user 110 finishes
placing an item in the vehicle 104 (e.g., for a courier service),
etc. The vehicle computing system 102 can determine an estimated
time of user arrival based at least in part on the location data
associated with the user device 138. The vehicle computing system
102 can determine the estimated time of user arrival based at least
in at part on the estimated time until the user 110 starts
interaction with the vehicle 104 and the estimated time of
interaction duration (e.g., a sum of these estimated times) way of
example, the vehicle computing system 102 can determine an
estimated time until the user 110 completes boarding of the vehicle
102 based at least in part on the location data associated with the
user device 110 and the data associated with the user 110. More
particularly, the vehicle computing system 102 can determine an
estimated time until the user 110 completes boarding of the vehicle
104 based at least in part on the estimated time until the user 110
starts boarding the vehicle 104 and the estimated time of boarding
duration for the user 110. The vehicle computing system 102 can
determine the estimated time until the user 110 completes hoarding
of the vehicle 104 via the addition of these two time estimates
(e.g., ETCB=ETSB+EBD). In this way, the vehicle computing system
102 can determine the amount of time that the object(s) within its
surroundings would be impacted as the vehicle 104 waits for the
user 110 (e.g., how long other vehicles would be caused to stop
while waiting for the user 110 to board the vehicle 104).
[0128] In some implementations, in order to determine whether the
estimated time of user arrival (e.g., the estimated time until the
user 110 completes boarding of the vehicle 104) is acceptable, the
vehicle computing system 102 can compare the estimated time for
interaction between the user 110 and the vehicle 104 (e.g., the
estimated time until the user 110 completes boarding of the vehicle
104) to a time constraint 750. The time constraint 750 can be
expressed as a time threshold (e.g., indicating an acceptable
amount of stopping time) and/or cost data (e.g., cost functions
expressing a cost in relation to stopping time). This can allow the
vehicle computing system 102 to determine whether the amount of
stopping time is acceptable.
[0129] Similar to the traffic constraint 506, the time constraint
750 can be based on historic data (e.g., indicating historic wait
times), real-time data (e.g., indicating that the vehicles are
already waiting due to another traffic build-up in front of the
autonomous vehicle), expectations of individuals in the geographic
area, machine-learned model(s), and/or other information. For
example, in the event that there is already a traffic jam in front
of the vehicle 104, the time constraint 750 (e.g., indicative an
acceptable wait time) can higher.
[0130] In some implementations, the time constraint 750 can be
determined at least in part from a model, such as a machine-learned
model. For example, the machine-learned model can be or can
otherwise include one or more various model(s) such as, for
example, models using boosted random forest techniques, neural
networks (e.g., deep neural networks), or other multi-layer
non-linear models. Neural networks can include recurrent neural
networks (e.g., long short-term memory recurrent neural networks),
feed-forward neural networks, and/or other forms of neural
networks. For instance, supervised training techniques can be
performed to train the model (e.g., using historical wait time
data) to determine a time constraint 750 based at least in part on
input data. The input data can include, for example, the data
described herein with reference to FIGS. 7 and 8A-C, driving
characteristics of individuals in an associated geographic area,
complaints received from operators of vehicles that were caused to
stop by autonomous vehicles, etc. The machine-learned model can
provide, as an output, data indicative of a recommended time
constraint. In some implementations, the recommended time
constraint can be specific to the geographic area 200. For example,
the model can be trained based at least in part on data associated
with geographic area 200 such that the recommended time constraint
is specific for that particular region.
[0131] Additionally, or alternatively, one or more time estimates
of FIG. 7 can be based at least in part on a model. The vehicle
computing system 102 can obtain data descriptive of the model
(e.g., machine learned model). The vehicle computing system 102 can
provide input data to the model. The input data can include the one
or more of the types of data described herein with reference to
FIGS. 7 and 8A-C. For example, the model can determine an estimated
time of user arrival (e.g., an estimated time until the user 110
completes boarding of the vehicle 104 (ETCB)), an estimated time
until the user 110 starts interaction with the vehicle 104 (e.g.,
an estimate time until the user 110 starts boarding the vehicle 104
(ETSB)), and/or an estimated time of interaction duration between
the user 110 and the vehicle 104 (e.g., an estimated time of
hoarding duration for a user 110 (EBD)). For example, the estimated
time of user arrival to the vehicle 104 (e.g., ETSB) can be based
at least in part on the input data (e.g., the user's location,
heading, distance, altitude, etc.). In some implementations, the
model can analyze the input data with respect to the time
constraint 750 to determine whether the estimated time of user
arrival (e.g., the estimated time until the user 110 completes
boarding of the vehicle 104 (ETCB)) is high or low, significant or
insignificant, acceptable or unacceptable, etc. The output of such
a model can be the estimated time for interaction between the user
110 and the vehicle 104 (e.g., the estimated time until the user
110 completes boarding of the vehicle 104 (ETCB)) and/or whether it
is high or low, significant or insignificant, acceptable or
unacceptable, etc.
[0132] In some implementations, the output of the model can be
provided as an input to the model for another set of data (e.g., at
a subsequent time step). In such fashion, confidence can be built
that a determined time estimate is the accurate. Stated
differently, in some implementations, the process can be iterative
such that the time estimate can be recalculated over time as it
becomes clearer what the time estimate is with respect to the user
110. For example, the model can include one or more autoregressive
models. In some implementations, the model can include one or more
machine-learned recurrent neural networks. For example, recurrent
neural networks can include long or short-term memory recurrent
neural networks, gated recurrent unit networks, or other forms of
recurrent neural networks.
[0133] Returning to FIG. 6, the vehicle computing system 102 can
determine one or more vehicle actions based at least in part on at
least one of the estimated traffic impact 504 or the estimated time
of user arrival (e.g., one and/or both of the estimates). In some
implementations, the vehicle computing system 102 can determine the
vehicle actions) based at least in part on the estimated traffic
impact 504. By way of example, the vehicle action(s) can include
stopping within the vicinity 206 of the location 202 associated
with the user 110 (e.g., at least partially in the travel way 600).
In the event that the level of traffic (e.g., the number of other
vehicles that would be impacted by an in-lane stop) is below a
traffic threshold (e.g., the estimated traffic impact 504 is low),
the vehicle computing system 102 can determine that the vehicle 104
can stop within the travel way 600 to wait for the user 110 to
arrive at the vehicle 104. In some implementations, the vehicle
computing system 102 can determine the vehicle action(s) based at
least in part on the estimated time of user arrival. For instance,
in the event that the estimated time of user 702 arrival is below
the time constraint 706 (e.g., the estimated time of user arrival
is low), the vehicle computing system 102 can determine that the
vehicle 104 can stop at least partially within the travel way 600.
Such a stop can occur, for example, as close as possible (e.g., for
the vehicle 104) to the location 202 associated with the user
110.
[0134] In some implementations, the vehicle computing system 102
can base its determination to stop at least partially within the
travel way 600 on both the estimated traffic impact 504 and the
estimated time of user arrival. For instance, vehicle computing
system 102 can weigh each of these estimates to determine whether
it would be appropriate for the vehicle 104 to stop at least
partially in the travel way 600 to wait for the user 110. The
vehicle computing system 102 can apply a first weighting factor to
the estimated traffic impact 504 and a second weighting factor to
the estimated time of user arrival. The first weighting factor can
be different than the second weighting factor. In some
implementations, the first weighting factor can be inversely
related to the second weighting factor. An example equation can
include: ETI*w1+ETUR*w2, where "ETT" is the estimated traffic
impact 504, "w1" is the first weighing factor (e.g., 0 to 1),
"ETUR" is the estimated time to user arrival 702, and "w2" is the
second weighting factor (e.g., 0 to 1). By way of example, the
estimated traffic impact 504 may be high while the estimated time
of user arrival may be short. Accordingly, the vehicle computing
system 102 may determine that it can stop within the travel way 600
because although a higher number of objects (e.g., other vehicles)
may be caused to stop, it would only be for a short period of time
because the user 110 is close in distance to and/or quickly heading
toward the vehicle 104. In such a case, the estimated time of user
arrival can be given a greater weight than the estimated traffic
impact 504. In another example, the estimated impact on traffic 504
may be low while the estimated time of user arrival may be long.
Accordingly, the vehicle computing system 102 can determine that it
should not stop within the travel way 600 because although only a
few objects (e.g., other vehicles) may be caused to stop, it would
be for a greater period of time because the user 110 is farther
from (and/or moving slowly, moving away from, etc.) the vehicle
104. In such a case, the estimated time of user arrival can be
given a greater weight than the estimated traffic impact 504 In
some implementations, the first and second weighting factors can be
manually and/or automatically adjusted depending on the
circumstances (e.g., a VIP user is being provided the vehicle
services) and/or the geographic area 200.
[0135] In some implementations, the estimated traffic impact 504
can be adjusted based at least in part on the estimated time to
user arrival 702. For instance, in the event that the estimated
time to user arrival 702 is long, the estimated traffic impact 504
can be higher.
[0136] With reference again to FIG. 2, the vehicle computing system
102 can also, or alternatively determine that the vehicle 104 is to
enter into a holding pattern. For instance, the vehicle action(s)
can include traveling along a second vehicle route 208 (e.g., an
optimal holding pattern route). In some implementations, the
vehicle computing system 102 can cause the vehicle 104 to enter
into a particular operating mode in which the vehicle 104
implements a holding pattern (e.g., a holding pattern operating
mode). The vehicle 104 may be unable to find a parking location
before and/or after travelling passed the location 202 associated
with the user 110. Additionally, the vehicle 104 may determine that
it should not stop within a travel way 600 to wait for the user's
arrival, as described herein. Thus, the vehicle 104 can be
re-routed along a second vehicle route 208 that is at least
partially different than the first vehicle route. The second
vehicle route 208 can be a path along which the vehicle 104 can
travel to re-arrive within the vicinity 206 of the location 202 of
the user 110. For example, the second vehicle route 208 can be a
path along which the vehicle 104 can travel around a block, back to
the location associated with the user. In some implementations,
such a path may be similar to and/or the same as a portion of the
first vehicle route 204 (e.g., along the street associated with the
user 110). In some implementations, the second vehicle route 208
can be completely different from the first vehicle route 204 such
that no portion of the second vehicle route 208 overlaps with the
first vehicle route 204.
[0137] The determination of the second vehicle route 208 can be
based on a variety of information. For example, FIG. 9 depicts
example information 900 associated with a second vehicle route 208
according to example embodiments of the present disclosure. In some
implementations, the vehicle computing system 102 can determine the
second vehicle route 208 based at least in part on such
information. In some implementations, the second vehicle route 208
can be determined off-board the vehicle 104 by another computing
system (e.g., the operations computing system 106) and data
indicative of the second vehicle route 208 can be provided to the
vehicle computing system 102.
[0138] The second vehicle route 208 can be determined based at
least in part on current and/or historic traffic data 902A. For
example, the second vehicle route 208 can be determined to
implement the route that will allow the vehicle 104 to arrive back
within the vicinity 206 of the location 202 of the user 110 within
the shortest amount of time and/or distance. The second vehicle
route 208 can take into account the current traffic (and/or
historic traffic patterns) within the geographic area 200 such that
the vehicle 104 is minimally impeded by such traffic (e.g., such
that the second vehicle route 208 is the fastest and/or shortest
vehicle route to the location 202 of the user 110).
[0139] Additionally, or alternatively, the second vehicle route 208
can be determined based at least in part on map data 902B. For
example, the map data 902B can be used to determine the path (e.g.,
roads, other terrain, etc.) the vehicle 104 is to travel along to
arrive back within the vicinity 206 of the user 110.
[0140] In some implementations, the second vehicle route 208 can be
based on data 902C associated with other vehicle(s) in the
geographic area 200. For instance, the data 902C associated with
other vehicle(s) can include additional traffic data associated
with the geographic area 200 (e.g., indicating a certain road is
impeded by heavy traffic). The data 902C associated with the other
vehicle(s) can also include the location of such vehicles. For
example, the vehicle computing system 102 and/or the operations
computing system 106 can determine the second vehicle route 208
(e.g., optimal vehicle holding pattern) by processing map data and
traffic data to establish an estimated time back to the location
202 associated with the user 110. If another vehicle (e.g., another
autonomous vehicle in an associated fleet) can arrive at the
location 202 associated with the user 110 in a shorter time period
than the vehicle 104, the other vehicle (rather than the vehicle
104) can be routed to the location 202. If the vehicle 104 would
arrive to the location 202 the fastest, then the vehicle 104 can be
routed in accordance with the second vehicle route 208. In some
implementations, the second vehicle route 208 can be based on a
model, such as a machine-learned model, in a manner similar to that
described herein with respect to the estimated traffic impact 504,
the estimated time to user arrival 702, etc.
[0141] In some implementations, the vehicle computing system 102
can determine that the vehicle 104 can stop within the travel way
600, but later determine that the vehicle 104 should begin to
travel again (e.g., according to a holding pattern route). For
example, the vehicle computing system 102 can determine that it
would be appropriate for the vehicle 104 to stop at least partially
within the travel way 600 to wait for the user 110 based at least
in part on the estimated traffic impact 504 and/or the estimated
time of user arrival, as described herein. The vehicle computing
system 102 can be configured to update (e.g., continuously,
periodically, as scheduled, in real-time, in near real-time, etc.)
the estimated traffic impact 504 and/or the estimated time of user
arrival. For example, while the vehicle 104 is stopped, the traffic
impact may increase (e.g., due to an increase in the number of
other vehicle(s) stopped behind the vehicle 104) and/or the user
110 may take longer than estimated to arrive at the vehicle 104.
The vehicle computing system 102 can determine at least one of an
updated estimated traffic impact (e.g., based on the number of
vehicles that have already stopped and/or additional vehicles that
may be caused to stop) or an updated estimated time of user arrival
(e.g., based on a change in the user device location data, if any).
The vehicle computing system 102 can determine that the vehicle 104
can no longer remain stopped to wait for the user 110 based at
least in part on at least one of the updated estimated traffic
impact or the updated estimated time of user arrival. Accordingly,
the vehicle computing system 102 can cause the vehicle 104 to
travel along the second vehicle route 208 based at least in part on
at least one of the updated estimated traffic impact or the updated
estimated time of user arrival.
[0142] The vehicle computing system 102 can cause the vehicle 104
to perform one or more vehicle action(s). As described herein, the
vehicle action(s) can include at least one of stopping the vehicle
104 (e.g., at least partially within the travel way 600) within the
vicinity 206 of the location 202 associated with the user 110 or
travelling along a second vehicle route 208. In the event that the
one or more vehicle actions include stopping the vehicle 104 within
the vicinity 206 of the location 202 associated with the user 110,
the vehicle computing system 102 can cause the vehicle 104 to stop.
For example, the vehicle computing system 102 can provide one or
more control signals to a control system 116 of the vehicle 104
(e.g., braking control system) to cause the vehicle 104 to
decelerate to a stopped position that is at least partially in a
travel way 600 within the vicinity 206 of the location 202
associated with the user 110. In the event that the vehicle
computing system 102 has determined that the vehicle 104 is to
travel along a second vehicle route 208 (e.g., in accordance with
the holding pattern), the vehicle computing system 102 can obtain
data associated with the second vehicle route 208 and implement the
second vehicle route 208 accordingly. For example, the vehicle
computing system 102 can request and obtain data indicative of the
second vehicle route 208 from the operations computing system 106.
Additionally, or alternatively, the vehicle computing system 102
can determine the second vehicle route 208 onboard the vehicle 104.
The vehicle computing system 102 can provide one or more control
signals to cause the vehicle 104 to implement a motion plan that
causes the vehicle 104 to travel in accordance with the second
vehicle route 208 (e.g., to implement one or more vehicle
trajectories in accordance with the second vehicle route 208).
[0143] The vehicle computing system 102 can provide the user 110
with one or more communications indicating the actions performed by
(or to be performed by) the vehicle 104. For example, in the event
that the vehicle 104 does not find a parking location and does not
stop in the travel way, the vehicle computing system 102 can
provide, via the communication system 136, a communication to the
user device 138 associated with the user 110. The communication can
include, for example, a textual message, auditory message, etc.
that indicates the vehicle actions (e.g., "I could not locate you
at the pin drop, traffic forced me to go around the block. Please
proceed to the pin drop").
[0144] In the event that the vehicle 104 stops at least partially
within the travel way 600, the vehicle computing system 102 can
provide a communication (e.g., data) to a user device 138
associated with the user 110. The communication can indicate that
the vehicle 104 is travelling to return to the location 202
associated with the user 110. In response to receiving such a
communication, the user device 138 associated with the user 110 can
display a user interface indicative of the communication. For
example, FIG. 10 depicts an example display device 1000 with an
example communication 1002 according to example embodiments of the
present disclosure. The display device 1000 (e.g., display screen)
can be associated with the user device 138 associated with the user
110. The communication 1002 can be presented via a user interface
1004 on the display device 1000. The communication 1002 can
indicate that the vehicle 104 has arrived and is waiting in the
travel way (e.g., in a current traffic lane). In sonic
implementations, the communication 1002 and/or another portion of
the user interface 1004 can be indicative of a location of the
vehicle 104. For example, the display device 1000 can display a map
user interface 1006 that includes a user route 1008. The user route
1008 can be a route along which a user 110 can travel to arrive at
the vehicle 104.
[0145] Returning to FIG. 2, in some implementations, the vehicle
104 may be relieved of its responsibility to provide a vehicle
service to the user 110. For instance, in the event that a vehicle
computing system 102 (and/or operations computing system 106)
determines that a vehicle 104 is to travel along the second vehicle
route 208, such computing system(s) can determine whether it would
be advantageous (e.g., more time efficient, more fuel efficient,
etc.) for another vehicle 210 (e.g., another autonomous vehicle)
within the geographic area 200 to be routed to the user 110. In the
event that it would be advantageous (e.g., because the other
vehicle 210 can arrive sooner), the operations computing system 106
can provide data to the vehicle 104 indicating that the vehicle 104
is no longer responsible for the request 140. The other vehicle 210
can be routed to the user 110 in the manner described herein.
Additionally, or alternatively, the operations computing system 106
(and/or the vehicle computing system 102 of the vehicle 104) can
re-route the vehicle 104 to provide a vehicle service to another
user.
[0146] In some implementations, the vehicle computing system 102
can cancel the request 140 associated with the user 110. By way of
example, the vehicle 104 may be caused to re-route e.g., circle the
block) a certain number of times and/or the user 110 may not arrive
at the vehicle 104 within a certain timeframe. The vehicle
computing system 102 can determine whether to cancel the request
140 based at least in part on a vehicle service cancellation
threshold.
[0147] FIG. 11 depicts example information 1100 associated with a
vehicle service cancellation threshold 1102 according to example
embodiments of the present disclosure. The vehicle service
cancellation threshold 1102 can he indicative of a threshold time
and/or distance that the vehicle 104 is in the holding pattern. For
instance, the vehicle service cancellation threshold 1102 can be
indicative of a time between when the vehicle 104 initially arrived
within a vicinity 206 of the user 110 (and/or passed the location
202) to the current time. Additionally, or alternatively, the
vehicle service cancellation threshold 1102 can be indicative of a
distance travelled by the vehicle 104 while in a holding pattern
(e.g., number of times the vehicle is re-routed to arrive at the
user 110). The vehicle service cancellation threshold 1102 can be
determined and updated continuously, periodically, as scheduled, on
request, in real-time, in near real-time, etc. (e.g., per trip,
while on a trip, etc.). The vehicle service cancellation threshold
1102 can be determined by the vehicle computing system 102 and/or
off-board the vehicle 104 and provided to the vehicle computing
system 102.
[0148] The information 1100 can include vehicle service demand data
1104A, historic vehicle service data 1104B, geographic area
preferences 1104C, data 1104D associated with other vehicle(s)
within the geographic area, user selected holding patterns 1104E,
and/or other types of information. The vehicle service demand data
1104A can include a current level of demand (e.g., number of
current service requests) for vehicle services (e.g., within the
geographic area 200). Additionally, or alternatively, the vehicle
service demand data 1104A can include a historic level of demand
for vehicle services at a certain time, day, etc. (e.g., within the
geographic area, similarly situated area, etc.). In the event that
the demand is lower, the vehicle service cancellation threshold
1102 can be higher (e.g., because the vehicle 104 may not be needed
for other vehicle service requests). In the event that the demand
is higher, the vehicle service cancellation threshold 1102 can be
lower (e.g., because the vehicle 104 may be needed for other
vehicle service requests).
[0149] The historic vehicle service data 1104B can be associated
with the specific user 110 and/or one or more other user(s). For
example, the historic vehicle service data 1104B can indicate that
the user 110 typically takes a longer amount of time to arrive at
the vehicle 104 (e.g., due to a disability). As such, the vehicle
service cancellation threshold 1102 can be higher in order to cause
the vehicle 104 to remain in the holding pattern longer, thereby
giving the user 110 a greater opportunity to arrive at the vehicle
104. Additionally, or alternatively, the historic vehicle service
data 1104B can indicate that it generally takes user(s) of within
the geographic area 200 longer to arrive at the vehicle 104. As
such, the vehicle service cancellation threshold 1102 may be
higher. In the event that the historic vehicle service data 1104B
indicates that the user 110 (and/or other user(s)) typically
arrives at the vehicle 104 in a relatively short timeframe, the
vehicle service cancellation threshold 1102 may be lower.
[0150] The geographic area preferences 1104C can be descriptive of
the preferences associated with a geographic area 200 (e.g., as
indicated by the managers of the geographic area 200). For example,
the geographic area 200 may prefer that a vehicle 104 only remain
in a holding pattern (e.g., circle the block) for a certain time
period and/or distance so as not to affect local traffic.
[0151] In some implementations, the vehicle service cancellation
threshold 1102 can be based at least in part on data 1104D
associated with one or more other vehicles (e.g., other autonomous
vehicles in an associated fleet). For example, the data 1104D can
be indicative of the location(s) of other vehicle(s) (e.g., within
the geographic area 200) and/or whether the other vehicle(s) are
available to provide a vehicle service (e.g., whether or not the
other vehicle is assigned to a service request, currently providing
a vehicle service, etc.). In the event that another vehicle is
located close to the location 202 associated with the user 110, the
vehicle service cancellation threshold 1102 may be lower (e.g.,
because the user 110 can be serviced by the other vehicle in the
event a new request is made after cancellation). In the event that
another vehicle is not located close to the location 202 associated
with the user 110, the vehicle service cancellation threshold 1102
may be higher (e.g., because another vehicle is not readily
available in the event the user 110 makes a new request after
cancellation).
[0152] In some implementations, the vehicle service cancellation
threshold 1102 can be based at least in part on user selected
holding patterns 1104E. The user selected holding patterns 1104E
can include data indicative of a vehicle service cancellation
threshold 1102 selected by a user 110. For example, the user 110
can purchase (e.g., via a user interface associated with a software
application) a higher vehicle service cancellation threshold 1102,
such that the vehicle 104 will remain in the holding pattern (e.g.,
circle the block) for a longer time/distance before the vehicle
service request is cancelled. In some implementations, a user 110
can have a higher vehicle service cancellation threshold 1102 due
to a higher user rating, specialized treatment (e.g., frequent
user), and/or based on other conditions.
[0153] In some implementations, the vehicle service cancellation
threshold 1102 can be based at least in part on a model, such as a
machine learned model. For example, the model can be trained based
on previously obtained information 1100 and labeled data indicative
of the vehicle service cancellation thresholds associated
therewith. In a manner similar to that described herein for the
estimated traffic impact 504 and the estimated time of user
arrival, the vehicle computing system 102 (or other computing
system) can provide input data(e.g., the information 1100) into
such a model and receive, as an output, a recommended vehicle
service cancellation threshold.
[0154] The vehicle computing system 102 can cancel the request 140
associated with the user 110 in the event that the user 110 has not
arrived at the vehicle 104 and the vehicle 104 has exceeded the
vehicle service cancellation threshold 1102. In response, the
vehicle computing system 102 can provide data indicating that the
request 140 for the vehicle service provided by the vehicle 104 is
cancelled to the operations computing system 106 (and/or one or
more other computing devices that are remote from the vehicle
computing system 102). In some implementations, such data can
request the cancellation of the user's service request 140. The
operations computing system 106 can cancel the service request 140
(and inform the user 110 accordingly) and/or re-route the vehicle
104 to provide a vehicle service to another user. In some
implementations, the vehicle computing system 102 can communicate
directly with a user device 138 associated with the user 110 to
cancel the service request 140 and/or inform the user 110 of the
vehicle service cancellation. The vehicle computing system 102 can
report such a cancellation to the operations computing system
106.
[0155] FIG. 12 depicts a flow diagram of an example method 1200 of
controlling autonomous vehicles according to example embodiments of
the present disclosure. One or more portion(s) of the method 1200
can be implemented by one or more computing devices such as, for
example, the one or more computing device(s) of the vehicle
computing system 102 and/or other systems. Each respective portion
of the method 1200 (e.g., 1202-1222) can be performed by any (or
any combination) of the one or more computing devices. Moreover,
one or more portion(s) of the method 1200 can be implemented as an
algorithm on the hardware components of the device(s) described
herein (e.g., as in FIGS. 1 and 14), for example, to control an
autonomous vehicle. FIG. 12 depicts elements performed in a
particular order for purposes of illustration and discussion. Those
of ordinary skill in the art, using the disclosures provided
herein, will understand that the elements of any of the methods
discussed herein can be adapted, rearranged, expanded, omitted,
combined, and/or modified in various ways without deviating from
the scope of the present disclosure.
[0156] At (1202), the method 1200 can include obtaining data
indicative of a location associated with a user. For instance, the
vehicle computing system 102 can obtain data 142 indicative of a
location 202 associated with a user 110 to which the vehicle 104 is
to travel. The vehicle 104 is to travel along a first vehicle route
204 that leads to the location 202 associated with the user 110. As
described herein, the user 110 can be associated with a request 140
for a vehicle service. The vehicle computing system 104 and/or the
operations computing system 106 can determine the first vehicle
route 204 so that the vehicle 104 can travel to the user 110 to
provide the user 110 with the requested vehicle service (e.g., pick
up the user 110 for a transportation service, deliver an item for a
delivery service, receive an item for courier service, and/or
provide another service).
[0157] At (1204), the method 1200 can include travelling along a
first vehicle route. A human operator may not be located within the
vehicle 104. The vehicle computing system 102 can provide one or
more control signals to the motion planning system 128 and/or the
vehicle's control systems 116 to cause the vehicle 104 to plan its
motion and/or implement a motion plan to travel in accordance with
first vehicle route 204 (e.g., autonomously, without input from a
human operator to the vehicle 104).
[0158] The first vehicle route 204 can bring the vehicle 104 within
the vicinity 206 of the location 202 associated with the user 110.
The vicinity of the location 202 associated with the user 110 can
be defined at least in part by a distance from the location 202
associated with the user 110. In some implementations, the distance
from the location 202 associated with the user 110 can be based at
least in part on an acceptable walking distance 302 from the
location 202 associated with the user 110, as described herein.
[0159] At (1206), the method 1200 can include determining whether a
parking location out of a travel way is available for the vehicle.
For instance, the vehicle computing system 102 can determine
whether a parking location within the vicinity 206 of the location
202 associated with the user 110 is available for the vehicle 104
(e.g., based on sensor data 118, map data 120, etc.). The vehicle
computing system 102 may search for a parking location while the
vehicle 104 is in an approach operating mode 108D, as described
herein. The vehicle computing system 102 may search for a parking
location before and/or after the vehicle 104 passes the location
202 associated with the user 110. In the event that the vehicle
computing system 102 is able to identify an available parking
location, the method 1200 can include sending a communication to a
user 110 to indicate that the vehicle 104 has arrived and the
location of the vehicle 104. Once the user has arrived at the
vehicle 104, the vehicle 104 can provide the vehicle service to the
user 110. In some implementations, the vehicle computing system 102
can determine that a parking location that is out of the travel way
600 is unavailable for the vehicle 104.
[0160] At (1210), the method 1200 can include obtaining traffic
data associated with the geographic area that includes the location
associated with the user. The vehicle computing system 102 can
obtain traffic data 500 associated with a geographic area 200 that
includes the location 202 associated with the user 110. The traffic
data 500 can be associated with the vicinity 206 of the location
202 (e.g., a block, neighborhood, etc. where the user 110 is
located) and/or other portions of the geographic area 200. For
instance, the vehicle computing system 102 can obtain, via one or
more sensors 112 of the vehicle 104, sensor data 118 associated
with the surrounding environment of the vehicle 104 that is within
the vicinity 206 of the location 202 associated with the user 110.
The vehicle computing system 102 can determine a level of traffic
based at least in part on the sensor data 118, as described
herein.
[0161] At (1212), the method 1200 can include determining an
estimated traffic impact. For instance, the vehicle computing
system 102 can determine an estimated traffic impact 504 of the
vehicle 104 on the geographic area 200 based at least in part on
the traffic data 500. The estimated traffic impact 504 can be
indicative of an estimated impact of the vehicle 104 on one or more
objects within a surrounding environment of the vehicle 104 in the
event that the vehicle 104 were to stop at least partially in the
travel way 600 (e.g., a current lane 602) within the vicinity 206
of the location 202 associated with the user 110. In some
implementations, to determine the estimated traffic impact 504, the
vehicle computing system 102 can compare the level of traffic
(e.g., determined based at least in part on the sensor data, other
traffic data) to a traffic constraint 506. The traffic constraint
506 can include a traffic threshold indicative of a threshold level
of traffic. In some implementations, the traffic constraint can be
determined at least in part from a machine-learned model, as
described herein.
[0162] At (1214), the method 1200 can include obtaining location
data associated with a user. For instance, the vehicle computing
system 102 can obtain location data associated with a user device
138 (e.g., mobile device) associated with the user 110, as
described herein. The location data associated with the user device
138 can be indicative of one or more locations of the user device
138 associated with the user 110 at one or more times.
[0163] At (1216), the method 1200 can include determining an
estimated time of user arrival. For instance, the vehicle computing
system 102 can determine an estimated time of user arrival based at
least in pan on the location data associated with the user device
138. The estimated time of user arrival can be indicative of an
estimated time at which the user 110 will arrive at the vehicle
104.
[0164] At (1218), the method 1200 can include determining one or
more vehicle actions based at least in part on the estimated
traffic impact and/or the estimated time of user arrival. For
instance, the vehicle computing system 102 can determine one or
more vehicle actions based at least in pan on the estimated traffic
impact 504. The vehicle computing system 102 can determine the one
or more vehicle actions also, or alternatively, based at least in
part on the estimated time of user arrival. The one or more vehicle
actions can include at least one of stopping the vehicle 104 at
least partially in a travel way 600 within a vicinity 206 of the
location 202 associated with the user 110 or travelling along a
second vehicle route 208 (e.g., entering into a vehicle holding
pattern). The second vehicle route 208 can be at least partially
different from the first vehicle route 204. The second vehicle
route 208 can include a route that leads to the location 202
associated with the user 110 (or at least to a vicinity 206 of the
location 202).
[0165] At (1220), the method 1200 can include causing the vehicle
to perform the one or more vehicle actions. For instance, the
vehicle computing system 102 can cause the vehicle 104 to perform
the one or more vehicle actions. To do so, the vehicle computing
system 102 can provide one or more control signals to one or more
systems onboard the vehicle 104 to cause the vehicle 104 to perform
the vehicle action(s) (e.g., to stop in the travel way 600, enter
the vehicle holding pattern).
[0166] At (1222), the method 1200 can include providing a
communication to the user. For instance, the vehicle computing
system 102 can provide a communication to a user device 138
associated with the user 110 that is indicative of a vehicle
action. By way of example, in the event that the vehicle 104 stops
at least partially within a travel way 600 within the vicinity 206
of the location 202 associated with the user 110, the vehicle
computing system 102 can provide, to the user device 138 associated
with the user 110, a communication indicating that the vehicle 104
is stopped (and/or will stop). In response to receiving the
communication, the user device 138 can display a map user interface
1006 that indicates a vehicle location of the vehicle 104 and a
user route 1008 to the vehicle location of the vehicle 104. In the
event that the vehicle 104 travels in accordance with the second
vehicle route 208, the vehicle computing system can provide a
communication to the user device 138 associated with the user 110
indicating that the vehicle 104 is travelling to return to the
location 202 associated with the user 110.
[0167] FIGS. 13A-B depict a flow diagram of an example method 1300
of controlling autonomous vehicles according to example embodiments
of the present disclosure. One or more portion(s) of the method
1300 can be implemented by one or more computing devices such as,
for example, the one or more computing device(s) of the vehicle
computing system 102 and/or other systems. Each respective portion
of the method 1300 (e.g., 1302-1346) can be performed by any (or
any combination) of the one or more computing devices. Moreover,
one or more portion(s) of the method 1300 can be implemented as an
algorithm on the hardware components of the device(s) described
herein (e.g., as in FIGS. 1 and 14), for example, to control an
autonomous vehicle FIGS. 13A-B depicts elements performed in a
particular order for purposes of illustration and discussion.
[0168] At (1302) of FIG. 13A, the vehicle 104 can enter into an
approach operating mode 108D. For example, a vehicle computing
system 102 can obtain data 142 indicative of a location 202
associated with a user 110. The vehicle computing system 102 can
cause the vehicle 104 to travel in accordance with a first vehicle
route 204 to arrive within a vicinity 206 of the location 202
associated with the user 110. The vehicle 104 can enter into the
approach mode 108D, for example, when it is within the vicinity 206
of the location 202. The vehicle 104 can approach the location 202
associated with the user 110 in the approach operating mode 108D.
Moreover, the vehicle computing system 102 can send accurate (e.g.,
localized) approach data (e.g., through a network) to a software
application running on a user device 138 associated with the user
110. The software application can cause the user device 138 to
display a user interface via a display device. The user interface
can display a map interface with the user's position (e.g., based
on GPS) and a precise location approach (e.g., of the vehicle 104),
as well as a target location of the vehicle 104 to meet the user
110. The user interface can also alert the user 110 that the
vehicle 104 is arriving (e.g., "your vehicle is arriving, please
prepare to board")
[0169] In some implementations, the vehicle 104 can each include an
outwardly visible lighting element, such as a number or array of
LED lights capable of producing rapid flash patterns. For example,
the lighting element can be located within the interior and
viewable through the front of the vehicle (e.g., windshield),
and/or can be located on the exterior of the vehicle 104. When the
vehicle 104 is assigned (or accepts) a service request, the
operations computing system 106 can transmit a flash code to the
vehicle 104 and the requesting user device 138. As the vehicle 104
approaches the pick-up location, the vehicle 104 can output the
flash code using the lighting element. The requesting user 110 can
be prompted to hold up the user device 138 so that a camera or the
camera lens of the user device 138 is pointed towards the vehicle
104 and can detect the flash code (e.g., the camera can be pointed
towards the vehicle 104 and the display screen of the mobile device
can display a viewfinder or preview of the imagery detected or
captured by the camera). Upon detecting the flash code from the
vehicle 104, the user device 138 can determine whether the flash
code matches the flash code provided by the operations computing
system 106 (e.g., utilizing a perception algorithm). If so, the
user device 138 can display an indicator, such as a circle or a
highlight for the vehicle 104, so that the requesting user 110 can
readily identify the vehicle 104.
[0170] While in the approach operating mode 108D, the vehicle 104
can search for an available parking location that is out of a
travel way 600 (e.g., out of a current travel lane 602), at (1304).
For example, the vehicle computing system 102 can search for an
out-of-lane parking location when the vehicle 104 is within a
certain distance (e.g., an acceptable walking distance 302) from
the location 202 associated with the user 110. To do so, the
vehicle computing system 102 can utilize the vehicle's sensor(s),
as described herein.
[0171] In the event an out-of-lane parking location is found, the
vehicle computing system 102 can cause the vehicle 104 to park
(e.g., autonomously, without user input), at (1306), and send a
communication to the user, as described herein. In some
implementations, if the estimated time until the user 110 starts
interaction with the vehicle 104 (e.g., the estimated time until
the user 110 starts boarding the vehicle 104) is low less than a
few seconds, 1, 2, 3 s, etc.) the vehicle doors can be unlocked
and/or user specific vehicle settings (e.g., music, temperature,
seat position, etc.) can be implemented, at (1307) (e.g., because
boarding is imminent). If the user 110 arrives within a certain
time frame (e.g., starts and/or completes boarding, item retrieval,
item placement, etc. within "X" time), at (1308), the vehicle 104
can start to provide the user 110 with a vehicle service (e.g.,
transport the user 110 to a destination location), at (1310).
However, if the user 110 does not arrive (e.g., start and/or
complete boarding, item retrieval, item placement, etc.) at the
vehicle 104 within the time frame, the vehicle computing system 102
can again contact the user 110, at (1312). The contact can be
facilitated by the vehicle computing system 102 providing a
communication to the user 110 via a software application on the
user device 138 (e.g., using a suitable communication protocol). If
unsuccessful (e.g., the user 110 is unresponsive, does not travel
to vehicle 104), the vehicle computing system 102 can cancel the
vehicle service, at (1314), as described herein. However, if the
contact is successful, the vehicle 104 can wait another timeframe
"Y" time) for the user 110 to arrive e.g., start and/or complete
boarding, item retrieval, item placement, etc.), at (1316). The
vehicle computing system 102 can cause the vehicle 104 to provide
the vehicle service to the user 110, in the event that the user 110
arrives at the vehicle 104 (e.g., boards the vehicle 104). In the
event the user 110 does not arrive within "Y" time (e.g., start
and/or complete boarding, item retrieval, item placement, etc.),
the vehicle 104 can enter into the holding pattern and travel in
accordance with the second vehicle route 208, at (1318).
[0172] In the event that an out-of-lane parking location is not
found, the vehicle computing system 102 can determine at least one
of an estimated traffic impact 504 of the vehicle 104 on the
geographic area 200 based at least in part on traffic data 500, at
(1320) or an estimated time of user arrival based at least in part
on the location data associated with the user device 110 as
described herein, at (1322). The vehicle computing system 102 can
determine the estimated traffic impact 504 to determine if and how
long to wait in-lane for the user 110 to arrive at the vehicle 104.
As the vehicle 104 moves towards the location 202 associated with
the user 110, the vehicle computing system 102 can determine
whether the vehicle 104 should stop at the location 202 based at
least in part on at least one of the estimated traffic impact 504
or the estimated time to user arrival. For example, the vehicle
computing system 102 can start sensing the traffic presence and
object speed around the vehicle 104 (e.g., in-lane, behind and
ahead of the vehicle 104). Moreover, the vehicle computing system
102 can use RF sensors and/or Bluetooth beacons to determine a user
presence, general distance between the user 110 and the vehicle
104, and/or changes in the distance indicating number of seconds
from user arrival.
[0173] In some implementations, in the event that the estimated
traffic impact 504 is low, the vehicle computing system 102 can
cause the vehicle 104 to stop in the travel way 600, at (1324). If
the user 110 arrives (e.g., start and/or complete boarding, item
retrieval, item placement, etc.) within a certain timeframe (e.g.,
within "Y" time), the vehicle 104 can provide the vehicle service
to the user 110. If the user 110 does not arrive (e.g., start
and/or complete boarding, item retrieval, item placement, etc.)
within the timeframe, the vehicle 104 can enter into the holding
pattern.
[0174] In some implementations, in the event that the estimated
traffic impact 504 is high, the vehicle computing system 102 can
determine the estimated time to user arrival. In the event that the
estimated time to user arrival is low, the vehicle computing system
102 can cause the vehicle 104 to stop within the travel way 600
(e.g., in a travel lane), at (1326), despite potential traffic
build-up. If the user 110 arrives (e.g., starts and/or completes
boarding, item retrieval, item placement, etc.) within a certain
timeframe (e.g., within "Y" time), the vehicle 104 can provide the
vehicle service to the user 110. If the user 110 does not arrive
(e.g., start and/or complete boarding, item retrieval, item
placement, etc. within the timeframe, the vehicle 104 can enter
into the holding pattern.
[0175] With reference to FIG. 13B, the vehicle computing system 102
can cause the vehicle 104 to travel past the location 202
associated with the user 110, at (1328). In some implementations,
this can be a portion of the second vehicle route 208. The vehicle
computing system 102 can search for a parking location out of the
travel way 600 after the vehicle 104 passes the location 202
associated with the user 110, at (1330). This can occur until the
vehicle 104 travels a certain distance past the location 202 (e.g.,
until the vehicle 104 reaches the acceptable walking distance 302).
In some implementations, even after the vehicle travels past the
location associated with the user 110, if the estimated traffic
impact 504 and/or the estimated time of user arrival is low enough
the vehicle computing system 102 can cause the vehicle 104 to
stop.
[0176] In the event that a parking location is found, at (1332),
the vehicle computing system 102 can contact the user 110 via the
user device 138 (e.g., indicating the location of the vehicle 104
and a user route thereto). The contact can be facilitated by the
vehicle computing system 102 providing a communication to the user
110 via a software application on the user device 138 (e.g., "This
is as close as I can get. Please come to me"). In some
implementations, if the estimated time until the user 110 starts
interaction with the vehicle 104 (e.g., the estimated time until
the user 110 starts boarding the vehicle 104) is low (e.g., less
than a few seconds, 1, 2, 3 s, etc.) the vehicle doors can be
unlocked and/or user specific vehicle settings (e.g., music,
temperature, seat position, etc.) can be implemented, at (1333)
(e.g., because boarding is imminent). If the user 110 arrives
(e.g., starts and/or completes boarding, item retrieval, item
placement, etc.) at the vehicle 104 within a certain timeframe
(e.g., with "Z" time), at (1334), the vehicle 104 can provide a
vehicle service to the user 110, at (1336). If the user 110 does
not arrive (e.g., start and/or complete boarding, item retrieval,
item placement, etc.) within the timeframe, the vehicle computing
system 102 can contact the user 110, at (1338). If the contact is
successful, the vehicle 104 can again wait for the user 110 to
arrive (e.g., start and/or complete boarding, item retrieval, item
placement, etc.) to the vehicle 104. If the contact is not
successful, the vehicle computing system 102 can cancel the vehicle
service, at (1340).
[0177] In the event that the vehicle computing system 102 does not
find a parking location after the location 202 associated with the
user 110 and within the vicinity 206 of the location 202 (e.g., an
acceptable walking distance 302), the vehicle computing system 102
can cause the vehicle 104 to implement a vehicle holding pattern
(e.g., enter a holding pattern operating mode), at (1342). As such,
the vehicle 104 can ignore any potential parking locations and
provide a communication to the user device 138 associated with the
user 110. The user device 138 can display a map interface depicting
a location of the vehicle 104. In some implementations, the vehicle
computing system 102 can request a second vehicle route 208 from
the operations computing system 106. The operations computing
system 106 can provide data indicative of the second vehicle route
208 to the vehicle computing system 102. The vehicle computing
system 102 can obtain the data indicative of the second vehicle
route 208 and send one or more control signals to cause the vehicle
104 to travel in accordance with the second vehicle route 208. At
(1344), the vehicle computing system 102 can send a communication
to the user 110 indicating that the vehicle 1.04 is travelling to
return to the location 202 associated with the user 110, as
described herein. As the vehicle 104 returns back toward the
location 202 associated with the user 110, the vehicle 104 can
enter into the approach operating mode 108D again, at (1346). As
such, the process can continue as shown in FIG. 13A.
[0178] The vehicle 104 can continue in the holding pattern until
the vehicle service cancellation threshold 1102 is reached. After
the vehicle service cancellation threshold 1102 is reached, the
vehicle computing system 102 can provide a communication directly
to the user device 138 (e.g., via the vehicle's communication
system 136) to inform the user 110 that the vehicle service has
been cancelled. In some implementations, if another vehicle would
be more appropriate to provide the vehicle service to the user 110
(e.g., another vehicle could arrive to the location 202 associated
with the user 110 quicker than the vehicle 104), the operations
computing system 106 can re-route the other vehicle to the location
202 associated with the user 110.
[0179] FIG. 14 depicts example system components of an example
system 1400 according to example embodiments of the present
disclosure. The example system 1400 can include the vehicle
computing system 102, the operations computing system 106, and a
machine learning computing system 1430 that are communicatively
coupled over one or more network(s) 1480.
[0180] The vehicle computing system 102 can include one or more
computing device(s) 1401. The computing device(s) 1401 of the
vehicle computing system 102 can include processor(s) 1402 and a
memory 1404 (e.g., onboard the vehicle 104). The one or more
processors 1402 can be any suitable processing device (e.g., a
processor core, a microprocessor, an ASIC, a FPGA, a controller, a
microcontroller, etc.) and can be one processor or a plurality of
processors that are operatively connected. The memory 1404 can
include one or more non-transitory computer-readable storage media,
such as RAM, ROM, EEPROM, EPROM, on