U.S. patent application number 15/924002 was filed with the patent office on 2019-02-28 for systems and methods for detecting patterns in spatio-temporal data collected using an rfid system.
This patent application is currently assigned to Mojix, Inc.. The applicant listed for this patent is Mojix, Inc.. Invention is credited to Christopher Richard Jones, Ramin Sadr.
Application Number | 20190069054 15/924002 |
Document ID | / |
Family ID | 44787813 |
Filed Date | 2019-02-28 |
United States Patent
Application |
20190069054 |
Kind Code |
A1 |
Sadr; Ramin ; et
al. |
February 28, 2019 |
Systems and Methods for Detecting Patterns in Spatio-Temporal Data
Collected Using an RFID System
Abstract
Systems and methods are described that collect spatio-temporal
data using an RFID system that is capable of locating the spatial
position of a sensor, which is typically unaware of its location.
Such systems and methods can be contrasted with conventional RFID
systems in that they are able to determine the location of sensors
in space as opposed to with respect to read zones related to the
underlying RFID reader infrastructure. One embodiment includes an
RFID system having a plurality of read zones and configured to
obtain the spatio-temporal state of sensors within a sensor cloud,
where the spatio-temporal state of each sensor includes sensor
information, a time stamp, and a spatial location specified
independently of the read zones of the RFID system, a
spatio-temporal database configured to store spatio-temporal state
of a plurality of sensors over time, and an application server
configured to trigger events based upon the detection of at least
one condition by applying at least one filter to the data within
the spatio-temporal database.
Inventors: |
Sadr; Ramin; (Los Angeles,
CA) ; Jones; Christopher Richard; (Pacific Palisades,
CA) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Mojix, Inc. |
Los Angeles |
CA |
US |
|
|
Assignee: |
Mojix, Inc.
Los Angeles
CA
|
Family ID: |
44787813 |
Appl. No.: |
15/924002 |
Filed: |
March 16, 2018 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
15674908 |
Aug 11, 2017 |
9924244 |
|
|
15924002 |
|
|
|
|
13087354 |
Apr 14, 2011 |
9762976 |
|
|
15674908 |
|
|
|
|
61324282 |
Apr 14, 2010 |
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H04Q 2209/47 20130101;
H04Q 9/00 20130101; H04Q 2209/823 20130101; G06Q 10/087 20130101;
H04Q 2209/753 20130101; H04Q 2209/886 20130101 |
International
Class: |
H04Q 9/00 20060101
H04Q009/00 |
Claims
1. A method for monitoring sensors using an RFID system having a
plurality of read zones, comprising: obtaining a state of a
plurality of sensors using an RFID system, wherein the state of
each sensor comprises sensor information, a time stamp, and a
spatial location that is a representation of a single unique
position in space in which the sensor is located, which is
specified independently of the read zones of the RFID system,
wherein obtaining the state comprises: determining, using the RFID
system, the spatial location of a sensor independent of the read
zones of the RFID system based on a set of properties of signals
received from the sensor when the sensor is illuminated, wherein
the set of properties comprises at least one of a phase and a
direction of arrival of the signals; and storing the state of the
plurality of sensors within a database, wherein the stored state of
at least one sensor comprises a first spatial location within a
particular read zone with a first time stamp and a different,
second spatial location within the particular read zone with a
different, second time stamp; determining whether the stored state
of the plurality of sensors satisfies at least one spatial
condition based on the first and second spatial locations of the at
least one sensor within the particular read zone; and triggering at
least one event based upon determining that the stored state
satisfies the at least one spatial condition.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The current application claims priority as a continuation of
U.S. patent application Ser. No. 15/674,908 filed Aug. 11, 2017,
entitled "Systems and Methods Systems and Methods for Detecting
Patterns in Spatio-Temporal Data Collected Using an RFID System",
which is a continuation of U.S. patent application Ser. No.
13/087,354 filed Apr. 14, 2011 and issued as U.S. Pat. No.
9,762,976 on Sep. 12, 2017, entitled "Systems and Methods Systems
and Methods for Detecting Patterns in Spatio-Temporal Data
Collected Using an RFID System", which claims priority under 35
U.S.C. .sctn. 119(e) to U.S. Provisional Patent Application No.
61/324,282 filed Apr. 14, 2010, entitled "Systems and Methods for
Event Generation Based Upon Spatio-Temporal Data Collected Using an
RFID System" the disclosure of which is incorporated by reference
herein in its entirety.
BACKGROUND
[0002] The present invention relates generally to radio frequency
identification (RFID) and more specifically to interpreting data
generated by RFID systems.
[0003] A conventional RFID system involves locating RFID tags
within read zones of RFID readers. Each of the RFID tags typically
includes an identification sequence enabling identification of the
tag, when it responds within the read zone of an RFID reader. The
precision with which the RFID system can locate an RFID tag is
determined by the size of the read zones of each of the RFID
readers. A typical RFID system builds a data set from RFID tag
reads, where each read indicates a unique identifier, has an
associated time stamp, and indicates a read zone. When the read
zones of RFID readers overlap, analysis of information such as the
read rate of a specific RFID tag by different readers can be used
to determine the zone in which the RFID tag actually resides.
[0004] Many RFID systems are deployed in support of real world
business processes. The term business process is used here to
describe a sequence of activities that are performed in response to
predetermined conditions in a real world environment. Examples of
business processes include, but are not limited to, assembly,
packaging, sorting, shipping and receiving. As assets move through
a supply chain, they are typically the subject of numerous business
processes. Significant amounts of data can be generated by RFID
systems related to movement of assets during the performance of
business processes. The data can be collected and rules and/or
queries (collectively filters) can be applied to the collected
data. When location precision is limited to zones defined by the
placement of RFID readers, placement of the RFID readers becomes
important to obtaining information that is relevant to coarse asset
movements. For example, determining whether goods are within a
specific staging area typically involves placing RFID readers in
such a way that the read zones of a set of readers only cover the
staging area. Such a system tends to be static. Therefore, changing
a business process can necessitate moving the RFID infrastructure
(i.e. the RFID readers and any associated cabling and sensor
hardware) in order to obtain information that is useful to support
the new business process. In addition, the information concerning
motion of assets is limited to dwell times within and transitions
between read zones.
SUMMARY OF THE INVENTION
[0005] Systems and methods are described that collect
spatio-temporal data using an RFID system that is capable of
locating the spatial position of a sensor, which is typically
unaware of its location, using an RFID reader system. Such RFID
systems and methods can be contrasted with conventional RFID
systems in that they are able to determine the location of sensors
in space as opposed to with respect to read zones related to the
underlying RFID reader infrastructure. In addition, such systems
are able to apply filters to the spatial locations of sensors
collected over time (i.e. spatio-temporal data) that involve
spatial conditions (i.e. allowed spatial locations) that are
defined independently of the read zones of the underlying RFID
reader infrastructure. In many embodiments, RFID systems utilize
event driven actions and data driven actions to enhance real world
business processes occurring within a sensor cloud.
[0006] One embodiment includes obtaining the spatio-temporal state
of a plurality of sensors within the sensor cloud using the RFID
system, where the spatio-temporal state of each sensor includes
sensor information, a time stamp, and a spatial location specified
independently of the read zones of the RFID system, storing the
spatio-temporal state of the plurality of sensors within a
spatio-temporal database, applying at least one filter to the
spatio-temporal data to detect the presence of at least one
condition, and triggering at least one event based upon the
detection of the at least one condition.
[0007] In a further embodiment, the spatial location is specified
as a spatial co-ordinate.
[0008] In another embodiment, the spatial coordinate specifies a
unique position in 3-dimensional space.
[0009] In a still further embodiment, the spatio-temporal state of
at least one sensor also includes at least one value selected from
the group consisting the identity of a device within the RFID
system that was activated to obtain the sensor data, the phase of
the signal from the sensor received by a RFID receiver, and the
direction of arrival of the signal received from the sensor by a
RFID receiver.
[0010] In still another embodiment, the calculation of the spatial
location of each sensor does not rely upon the receipt by the
sensor of a signal generated by a device that does not form part of
the RFID system.
[0011] In a yet further embodiment, the filter applied to the
spatio-temporal data to detect the presence of at least one
condition comprises at least one spatial condition, where each
spatial condition specifies at least one spatial location
independently of the read zones of the RFID system.
[0012] In yet another embodiment, the filter detects whether a
sensor is present within the at least one spatial location
specified by the at least one spatial condition.
[0013] In a further embodiment again, the filter also detects
whether the spatio-temporal data collected concerning a sensor
having a spatio-temporal state that satisfies the at least one
spatial condition also satisfies a temporal condition.
[0014] In another embodiment again, the filter also detects whether
the spatio-temporal state of a sensor that satisfies the at least
one spatial condition includes sensor data that satisfies a data
condition.
[0015] In a further additional embodiment, the filter applied to
the spatio-temporal data to detect the presence of at least one
condition comprises a sequence of spatial conditions, where each
spatial condition specifies at least one spatial location
independently of the read zones of the RFID system.
[0016] In another additional embodiment, the filter detects whether
a sensor moves between the spatial locations specified by the
sequence of spatial conditions.
[0017] In a still yet further embodiment, the filter also detects
whether the spatio-temporal data collected concerning a sensor
having a trajectory that satisfies the sequence of spatial
conditions also satisfies at least one temporal condition.
[0018] In still yet another embodiment, the filter also detects
whether the spatio-temporal state of a sensor having a trajectory
that satisfies the sequence of spatial conditions includes sensor
data that satisfies a data condition.
[0019] In a still further embodiment again, the filter applied to
the spatio-temporal data to detect the presence of at least one
condition detects motion of a sensor.
[0020] In still another embodiment again, the event triggered by
the filter is the reading of a sensor that is determined to be in
motion by an RFID system to obtain the spatio-temporal state of the
sensor.
[0021] A still further additional embodiment also includes applying
a time based filter to condition the spatio-temporal data within
the spatio-temporal database.
[0022] In still another additional embodiment, the event is
selected from the group consisting of sending a notification,
initiating a sensor read using the RFID system, activating a
device, and initiating a process.
[0023] In a still yet further embodiment again, the sensor cloud
comprises sensors selected from the group consisting of binary
detectors, passive RFID tags, active RFID tags, and RFID tags that
include transducers.
[0024] Another further embodiment includes an RFID reader system
having a plurality of read zones and configured to obtain the
spatio-temporal state of sensors within a sensor cloud, where the
spatio-temporal state of each sensor includes sensor information, a
time stamp, and a spatial location specified independently of the
read zones of the RFID reader system, a spatio-temporal database
configured to store spatio-temporal state of a plurality of sensors
over time, and an application server configured to trigger events
based upon the detection of at least one condition by applying at
least one filter to the data within the spatio-temporal
database.
[0025] In another further embodiment again, the application server
is configured to apply filters including at least one spatial
condition, where each spatial condition specifies at least one
spatial location independently of the read zones of the RFID reader
system.
BRIEF DESCRIPTION OF THE DRAWINGS
[0026] FIG. 1 illustrates a process involving collection of
spatio-temporal data from a sensor cloud and applying filters to
the spatio-temporal data to provide asset visibility and to trigger
actions in accordance with an embodiment of the invention.
[0027] FIG. 2 conceptually illustrates spatial constraints defined
within the coverage area of an RFID system in accordance with an
embodiment of the invention.
[0028] FIG. 3 is a flow chart illustrating a process for applying a
presence detection filter to spatio-temporal data in accordance
with an embodiment of the invention.
[0029] FIG. 4 is a flow chart illustrating a process for applying a
motion based filter to spatio-temporal data in accordance with an
embodiment of the invention.
DETAILED DESCRIPTION OF THE INVENTION
[0030] Turning now to the drawings, RFID systems in accordance with
embodiments of the invention are described that are configured to
build spatio-temporal databases of sensor information and to detect
and respond to patterns of data within the spatio-temporal
database. In many embodiments, the RFID system is capable of
locating the spatial position of sensors within a sensor cloud. The
term sensor cloud refers to all of the sensors that are visible
within the coverage area of an RFID system. As information is
collected from the sensors, the RFID system can build a
spatio-temporal database that describes the spatio-temporal
trajectories of the sensors. Filters can be applied to the data
within the spatio-temporal database based upon a priori knowledge
of real world actions or business processes that are likely to be
applied to assets within the sensor cloud.
[0031] The ability of an RFID system in accordance with an
embodiment of the invention to locate the spatial position of
sensors at a given time as opposed to simply identifying the
presence of a sensor within a read zone enables the RFID system to
respond to the spatial location and/or the trajectory of sensors
instead of in terms of the presence or absence of sensors within
read zones determined by the underlying RFID reader infrastructure.
The RFID system can react to information obtained via application
of a filter to spatio-temporal data in any of a variety of ways
including but not limited to tracking the occurrence of an event,
sending an alert message, sending a signal to activate one or more
devices, and/or initiating a process. Furthermore, modification of
real world business processes need not involve modification of the
underlying RFID reader infrastructure. As real world business
processes change, a modified set of filters can be defined and
enforced upon the spatio-temporal data collected from the sensor
cloud by the RFID system. In many embodiments, event based filters
are applied to a spatio-temporal database of sensor information. In
several embodiments, time based filters are applied to a
spatio-temporal database of sensor information. Systems and methods
for collecting spatio-temporal data from the sensor cloud and
applying filters and/or constraints are discussed further
below.
The Sensor Cloud
[0032] RFID systems in accordance with embodiments of the invention
rely on the presence of a sensor cloud in order to build a
spatio-temporal database of sensor information. The term sensor is
used here to describe any device, which measures a physical
quantity and provides information to the RFID system. Sensors can
be as simple as a binary detector, such as a motion detector.
Passive RFID tags are devices that harvest energy from
electromagnetic waves and when powered by an incident signal, the
tag responds with the information embedded in the device, typically
preprogrammed with an identification sequence. More sophisticated
RFID devices may embed sensor circuits including but not limited to
transducers to convert various forms of energy, such as energy
arising from chemical interaction, electromagnetic or thermal
couplers to sense temperature, or piezoelectric transducers to
measure pressure. The term sensor also encompasses sensors with an
embedded battery, such as active RFID tags. As can be readily
appreciated from the above description, the term sensor is not
restricted to obtaining any particular type of information or usage
of any specific energy source.
Collection of Spatio-Temporal Data from the Sensor Cloud
[0033] Collection of spatio-temporal data from the sensor cloud
involves use of an RFID system that is capable of obtaining the
spatial location of a sensor. A distinction can be made between an
RFID system obtaining the spatial location of a sensor, and a
sensor obtaining its spatial location (i.e. independently
collecting information from which spatial location can be
determined). The RFID system is considered to obtain the spatial
location of a sensor, when the calculation of the spatial location
of the sensor does not rely upon the receipt of a signal by the
sensor generated by a device that does not form part of the RFID
system. Many classes of sensor include the capability of obtaining
their own spatial location, (e.g. devices equipped with Global
Position System (GPS) receivers or that are able to determine
location based upon RF sources other than the RFID system). Such
devices enable the calculation of spatial location based upon
signals received from devices. When the sensor independently
collects information from a source other than the RFID system that
is used in determining the spatial location of the system, the
sensor is considered to have obtained its spatial location
independently of the RFID system.
[0034] An RFID system that can only express the location of an RFID
tag in terms of read zones is typically incapable of generating
spatio-temporal data (i.e. the data generated is specified in terms
of the location of the underlying RFID infrastructure instead of
the spatial location of the sensor). In several embodiments, an
RFID system is used to collect spatio-temporal data that is similar
to the RFID system described in U.S. patent application Ser. No.
12/423,796 entitled "Radio Frequency Identification Tag Location
Estimation and Tracking System and Method" to Sadr et al., filed
Apr. 14, 2009, the disclosure of which is incorporated by reference
herein in its entirety. In other embodiments, any RFID system that
is capable of determining the spatial position of a sensor can be
utilized to collect spatio-temporal data from a sensor cloud.
Definitions of Spatio-Temporal State
[0035] In many embodiments, the spatio-temporal data collected by
the RFID system is referred to as the spatio-temporal state of the
sensor and includes at least sensor information, a time stamp, and
a spatial co-ordinate (e.g. x, y, z) or some other representation
of a unique position in space. As noted above, sensor information
can take any of a variety of forms including, but not limited to,
an EPC identifier and/or a transducer measurement. The
spatio-temporal state of a sensor can also include other data
collected by an RFID system. In a number of embodiments, the
spatio-temporal state of a sensor includes data derived from low
level information collected by the RFID reader infrastructure such
as, but not limited to, the identity of the exciter that was
activated to obtain the sensor data, the phase of the received
sensor signal and/or the direction of arrival of the signal
received from the sensor. As can be readily appreciated, the
additional data (if any) collected in order to record the
spatio-temporal state of a sensor in a spatio-termporal database
can be determined based upon the requirements of a specific
application.
Applying Spatio-Temporal Filters
[0036] When sensors such as RFID tags are fixed to assets, an RFID
system in accordance with embodiments of the invention can build a
database describing the spatio-temporal trajectory of the assets as
they move within a facility. The spatio-temporal trajectories
provide real-time visibility of both inventory and asset location.
A priori knowledge of the business processes employed within a
sensor cloud can be used to define and apply filters to the
collected spatio-temporal database. The RFID system can respond to
information obtained by applying filters to the collected
spatio-temporal data. In many embodiments, the RFID system can
activate devices, initiate processes, and/or provide alerts that
enable the modification of real world business processes. The
collection of spatio-temporal data, the application of filters to
the collected spatio-temporal data and responses to information
obtained by specific filters in accordance with embodiments of the
invention are discussed further below.
[0037] A process for building a spatio-temporal database, and
applying filters based upon knowledge of real world business
processes in accordance with an embodiment of the invention is
illustrated in FIG. 1. The process of collecting and analyzing
spatio-temporal data 10 concerning assets within a facility
involves collecting the spatio-temporal state of sensors within a
sensor cloud 12 using an RFID receiver system, where the sensor
cloud 12 includes sensors fixed to the assets, and building a
spatio-temporal database 14 describing the spatio-temporal
trajectories of sensors within the sensor cloud. As noted in the
illustrated embodiment, assets bearing sensors including but not
limited to passive RFID tags can enter and leave the sensor
cloud.
[0038] A set of filters 16 are defined based upon a priori
knowledge of the business processes occurring in the real world
within the sensor cloud. As data is being collected in real time,
these filters can be applied to the spatio-temporal database by an
application server (not shown) configured to communicate with the
spatio-temporal database. In many embodiments, the application of
filters provides real-time asset visibility within the sensor cloud
and enables the RFID system to respond to observed patterns within
the spatio-temporal data to intervene and/or initiate business
processes related to the assets within the sensor cloud. In a
number of embodiments, the RFID system responds to information
derived using filters by sending notifications, activating or
otherwise controlling devices configured to communicate with the
RFID system via some form of machine-to-machine communication
system, and/or initiating a process including but not limited to
modifying the manner in which the RFID system collects information
from within the sensor cloud. Processes for building
spatio-temporal databases, applying filters to provide asset
visibility, and responding to conditions detected using filters are
discussed further below.
Using Filters to Detect Conditions within Sensor Cloud
[0039] RFID systems in accordance with embodiments of the invention
can observe conditions within a sensor cloud by applying a filter
or set of filters to a spatio-temporal database. The RFID system
can respond to the observation of the condition or a set of
conditions by invoking one or more actions. In several embodiments,
the actions can be event driven, and/or data driven. Event driven
filters can be applied to invoke actions that are triggered based
upon the spatio-temporal trajectories of sensors within the sensor
cloud. Data driven filters can be used to invoke actions that are
triggered in response to the non-spatio-temporal components of the
states of one or more sensors. Filters or sets of filters can also
be constructed that invoke actions based upon a combination of
spatio-temporal trajectories of sensors within the sensor cloud,
and non-spatio-temporal data.
Event Driven Filters
[0040] RFID systems in accordance with embodiments of the invention
enable the definition of event driven filters specified in terms of
the trajectory in two or three dimensional Euclidean space of one
or more sensors within the sensor cloud. Such conditions are not
bound to the underlying RFID reader infrastructure and allow the
triggering of actions in response to a wide array of motions. As is
discussed further below, simple filters can be generated involving
presence detection or motion detection. More complex motion filters
detect the movement of a sensor between a sequence of locations.
The sophistication of a presence detection filter or a motion
filter can be further increased by placing temporal constraints on
the amount of time at which a sensor can dwell at a specific
location.
[0041] In order to illustrate the operation of various event driven
filters, a pair of spatial conditions (i.e. sets of allowed spatial
locations) that define two dimensional areas within Euclidean space
are conceptually illustrated in FIG. 2. A number of RFID exciters
or readers 30 can be distributed throughout an area, which can be
read in read zones 32 defined by the underlying RFID exciter/reader
infrastructure. Due to the ability of an RFID system in accordance
with embodiments of the invention to collect spatio-temporal data,
the spatio-temporal trajectory of RFID tags through the coverage
area can be determined.
[0042] As can be appreciated from the illustrated embodiment,
spatial conditions 34, 36 for event based filters can be defined
independently of the boundaries of the read zones of the RFID
exciters/readers. In the illustrated embodiment, two rectangular
spatial conditions 34, 36 are defined so that each overlay portions
of the excitation zones of a number of RFID exciters/readers.
Although rectangular two dimensional spatial conditions are shown,
any of variety of spatial conditions, including but not limited to,
three dimensional spatial conditions that are appropriate to a
specific application can be utilized in accordance with embodiments
of the invention. As is discussed further below, once spatial
conditions are defined, various categories of event driven filters
can be applied to a spatio-temporal database generated by an RFID
system in accordance with an embodiment in order to trigger actions
under appropriate conditions.
Presence Detection
[0043] In its simplest form, an event driven filter can simply
detect whether a particular sensor is located at a specific spatial
location or within a set of allowed spatial locations (i.e. a
spatial condition). Such a filter can be referred to as a presence
detection filter. A presence detection filter contrasts with simply
detecting whether a sensor is within a read zone of a conventional
RFID receiver, because a presence detection filter utilizes the
spatial location of the sensor obtained by the RFID system to
determine whether a spatial condition defined independently of the
read zones of the RFID system is satisfied. Accordingly, multiple
overlapping or non-overlapping presence detection filters can be
defined within a single read zone. As is discussed further below,
the ability to define multiple allowed spatial locations
independently of read zones/the underlying RFID system
infrastructure can be useful not only in presence detection, but in
a wide array of event based filters.
[0044] A process for applying a presence detection filter to
spatio-temporal data for a number of sensors N in accordance with
an embodiment of the invention is illustrated in FIG. 3. The
process 40 involves progressively testing the spatial location
information for each sensor against one or more spatial conditions
and triggering events based upon the spatial conditions. In the
illustrated embodiment, a counter is initialized 42 and the spatial
location of each sensor is retrieved 44. A determination (46) is
made concerning whether the spatial location for each tag satisfies
a spatial condition. If a sensor's spatial location satisfies a
spatial condition, then an event is triggered (48). In many
embodiments, additional data or temporal conditions are applied to
a sensor satisfying a spatial condition. In the event that the
spatial condition and the data and/or temporal conditions are
satisfied, then an event is triggered. The counter continues to
increment (50) until a determination (52) is made that the spatial
location of each sensor has been compared to the spatial condition
or set of spatial conditions. Typically, the presence detection
filter continuously processes the spatio-temporal data to determine
whether one or more spatial conditions are satisfied by any of the
sensors in real time.
[0045] Applications of presence detection filters similar to the
presence detection filter shown in FIG. 3 include virtual fencing,
which involves defining presence detection filters that associate
sensors to a particular area or lane. For example, virtual fences
can be defined using a presence detection filter enabling the
triggering of a notification message in the event that perishable
goods are placed in an unsafe area (e.g. removed from a
refrigerated area). Virtual fencing has applications in a variety
of areas including security monitoring and Electronic Article
Surveillance (EAS). Security applications include defining spatial
conditions that trigger an event in response to an unauthorized
asset and/or person from being present within a restricted area. In
EAS applications, spatial conditions can be defined that detect the
unauthorized removal of an asset. Virtual fencing can be
particularly useful in logistic intensive security applications,
including but not limited to the tracking of airport luggage and
other types of freight including parcels and mail. In a retail
setting, application of presence detection filters to
spatio-temporal data can be used to identify the presence of a
specific person or customer. Once the presence of the customer is
detected, events can be triggered that notify sales staff and/or
provide customer specific shopping experiences including
controlling display of advertising based upon information related
to the customer.
[0046] Presence detection filters can also be applied to
spatio-temporal databases to automatically aggregate assets that
satisfy a spatial condition including but not limited to for the
purpose of manifest generation. Additional conditions can then be
applied to the aggregated assets. In many embodiments, spatial
conditions can be defined that enable a determination of whether a
correct group of assets is located on a pallet or within a staging
area. When incorrect assets are present (i.e. either a required
asset is absent or an undesired asset is present), an event can be
triggered. In a retail context, similar presence detection filters
can be used to detect the presence of goods within a lane adjacent
a point of sale and to group the assets that are present as a
single transaction. The transaction data can be compared against
information collected at the point of sale. Alternatively, the
asset data retrieved by the RFID system based upon the group of
assets determined to be present can be used to conduct the
transaction at the point of sale. Other applications of presence
detection filters to aggregate assets that can be useful in a
retail and/or warehouse setting include shelving applications
involving counting the number of assets on a specific shelf or
within a specific staging area. Extensions of such filters include
the remote monitoring of display placements. In an occupational
health and safety context, presence detection filters can be used
for a variety of purposes including but not limited to managing
evacuation and mustering employees in safe areas.
Sequences of Spatial Conditions
[0047] Motion filters can be utilized to detect an allowed sequence
of spatial conditions. In many embodiments, motion filters are
defined in such a way that a spatio-temporal condition C specifies
an allowed sequence of spatial conditions:
C = U i c i ##EQU00001##
[0048] where c.sub.i is a spatial condition in a sequence of
spatial conditions and C is the union of each spatial condition
c.sub.i.
[0049] A process for applying a motion filter to spatio-temporal
data collected with respect to a specific sensor in accordance with
an embodiment is illustrated in FIG. 4. The process 60 can be
performed in real time or applied to spatio-temporal data that has
already been collected concerning a sensor by an RFID system in
accordance with an embodiment of the invention. The process 60
commences by initializing (62) a counter i that stores a value of
the current spatial condition in the sequence of spatial conditions
defined by the motion filter. The process retrieves (64) the
spatial location for the sensor and determines (66) whether the
spatial location satisfies the current spatial condition in the
sequence of spatial conditions defined in the motion filter. In the
event the spatial condition is satisfied, then the counter is
incremented (66) and comparisons are performed between the spatial
location of the sensor over time and subsequent spatial conditions
in the sequence of spatial conditions. When all of the spatial
conditions in the sequence of spatial conditions have been
satisfied, an event is triggered (70). In many embodiments,
additional data and/or time conditions are applied to the sensor
and the event is triggered when these additional conditions are
also satisfied.
[0050] Referring back to FIG. 2, by way of example, a motion filter
can be defined that specifies that once a sensor has entered a
first area 34 defined in two dimensional space, the sensor should
not enter a second two dimensional area 36. Therefore, an event is
triggered in the event that a sequence is detected in which an
asset bearing a sensor enters the first area, and then the second
area. Examples of this type of motion filter include but are not
limited to motion filters that can trigger events based upon
movements within a read zone, such as an item being removed from a
shelf and placed back in a different location, triggering a
notification of the need to re-shelve the item. Similar motion
filters can be used to activate a device or machine or commence a
process or operation based upon the detected movement of an asset.
In a number of embodiments, motion filters include but are not
limited to motion filters that monitor sequences of spatial
conditions with respect to multiple sensors. Examples include but
are not limited to monitoring the sequence with which different
tools are used in the completion of a project and/or ensuring
appropriate inventory levels or availability of assets for a
specific process or sequence of operations.
[0051] Although the motion filters discussed above involve
detection of specific sequences of motion, many motion filters in
accordance with embodiments of the invention simply detect motion.
In many embodiments, events are triggered when a sensor moves. The
triggering of events in response to the motion of a sensor can be
useful in improving the accuracy with which an RFID system is
capable of obtaining the spatial location for sensors. In many
instances, RFID readers can interfere with each other by
interrogating sensors at the same time. Therefore, motion filters
can be used to restrict the interrogation of sensors to regions in
which sensors are currently in motion. Furthermore, when a sensor
is in motion the effects of RF degradation due to the environment
are diminished thanks to spatial diversity (i.e. the sensor is
transmitting from different positions as time progresses).
Therefore, spatial location estimates are also more reliable when a
sensor is in motion.
[0052] While specific motion filters and specific applications of
motion filters are discussed above, any of a variety of motion
filters and applications for motion filters can be utilized to
trigger events based upon spatio-temporal data collected by an RFID
system in accordance with embodiments of the invention.
Time Based Filters
[0053] While spatio-temporal data naturally lends itself to
filtering based upon allowed or prohibited motion, the temporal
characteristics of the motion can also be utilized as part of the
filtering process. In many embodiments, time based filters are
defined that trigger actions based upon allowed or prohibited dwell
time of goods (i.e. the amount of time in which the goods have been
stationary) and/or allowed or prohibited speed of movement of
goods. In several embodiments, time based filters involve adding a
dwell time condition to a presence detection filter or a motion
filter. In a number of embodiments, time based filters are utilized
in the storage and handling of perishable goods. For example, a
time based filter can trigger an event in response to use of a
perishable good in a real world business process after the
expiration date of the perishable good. Time based filters can also
be useful in detecting errors within the spatio-temporal data
collected by the RFID system.
[0054] Location errors can occur for a variety of reasons including
reflections of RF signals used by the RFID receiver to locate
sensors or blind spots in the coverage area of the RFID system.
Errors in determining the location for a sensor can result in the
incorrect enforcement of an event action. For example, when a
sensor is incorrectly determined to be outside of a facility, an
asset outflow action can be enforced resulting in the removal of
the asset from the real-time inventory of the facility. Over time,
however, location errors may be detected by analyzing
spatio-temporal data collected by the RFID system. For example, the
source of the error may be temporary or intermittent. In addition,
motion of the sensor may result in the sensor moving out of a blind
spot and/or into a location where the position of the sensor can be
more accurately determined. Therefore, time based filters can be
applied to condition data collected by an RFID system in accordance
with embodiments of the invention. In several embodiments, the time
based filters can be adaptive based upon the spatio-temporal
trajectory of a sensor. When a sensor is stationary, a longer time
base can be applied during the time base filtering. When a sensor
is in motion, a shorter time base can be applied. In a number of
embodiments, the length of the time base that is applied is
determined by the latency tolerance of a specific application. In
many embodiments, time base filters are used to exclude identified
spatio-temporal state data from processes that detect one or more
conditions by applying filters to the spatio-temporal data.
[0055] In several embodiments, blind spots in the coverage area can
be observed over time. In a number of embodiments, spatio-temporal
state estimates can be automatically provided concerning the
location of sensors as they move through blind spots.
[0056] Although specific applications of time based filters are
discussed above, any of a variety of time based filters can be
applied to spatio-temporal databases to trigger events and/or
condition the data in accordance with embodiments of the
invention.
Data Driven Filters
[0057] As can be readily appreciated, data driven actions can be
applied to a database that does not contain spatio-temporal data.
While the condition that triggers a data driven action typically
does not involve spatio-temporal data, the presence of
spatio-temporal data in the database enables a significantly
greater array of actions in response to a trigger event. For
example, a data driven action can prompt the generation of a list
of spatial information concerning sensors that are to be collected
into a single location, such as a list of the location of sensors,
which are attached to goods, that are to be picked and formed into
a pallet. In other words, filters can be applied that trigger
actions designed to produce pre-determined modifications to the
spatio-temporal state of sensors within the sensor cloud.
[0058] As noted extensively above, spatio-temporal conditions can
also be applied in combination with other types of conditions. For
example, entry of a shipping request (i.e. a non-spatio-temporal
condition) can prompt the RFID system to inspect subsequent
spatio-temporal data to confirm that the appropriate quantity and
quality of assets are shipped in a timely manner. As can be readily
appreciated, the ability to define event driven actions using
spatio-temporal conditions alone or in combination with other types
of conditions is almost limitless and a wide variety of event
driven actions appropriate to specific business processes can be
readily defined in accordance with embodiments of the
invention.
Modification of Business Rules without Modifying Infrastructure
[0059] Referring again to FIG. 2, event driven actions can be
enforced in response to the presence of spatial and/or
spatio-temporal conditions. As was noted above, the ability of RFID
systems in accordance with embodiments of the invention to record
the spatio-temporal trajectories of sensors enables the definition
of spatial and spatio-temporal conditions in terms of two or three
dimensional Euclidian space instead of in terms of read zones that
are constrained to the locations of components of the underlying
RFID infrastructure. Consequently, modifications of business
processes can be accommodated by simply modifying the definitions
of the spatial and/or spatio-temporal conditions that form the
basis of the event driven actions enforced by the RFID system.
Modification of the location of the underlying RFID infrastructure
is not required.
Data Mining
[0060] As can be readily appreciated, the collection of
spatio-temporal data by RFID systems in accordance with embodiments
of the invention provides almost limitless possibilities for mining
the spatio-temporal data for information that can enable
improvement of real world business processes. Examples of the ways
in which data mining can be used include but are not limited to
observing and optimizing work flows through the sensor cloud and/or
the location of reused assets within the sensor cloud to increase
the efficiency with which assets are moved through a facility.
Data Aggregation
[0061] RFID systems in accordance with embodiments of the invention
collect spatio-temporal data from sensors within a sensor cloud. In
many embodiments, information can be aggregated across multiple
RFID systems to provide visibility within all of the nodes of a
supply chain. In several embodiments, the spatio-temporal data from
each of the RFID systems is aggregated into a single
spatio-temporal database. In other embodiments, business process
rules are enforced upon the spatio-temporal database to generate
asset visibility data with respect to each facility and the asset
visibility data is aggregated in a single database. Examples of
data indicative of asset visibility include, but are not limited
to, asset inflow data, asset outflow data, and inventories of
specific asset classes.
[0062] Although the present invention has been described in certain
specific aspects, many additional modifications and variations
would be apparent to those skilled in the art. It is therefore to
be understood that the present invention may be practiced otherwise
than specifically described. Thus, embodiments of the present
invention should be considered in all respects as illustrative and
not restrictive.
* * * * *