Predictive model for automated vehicle recommendation system

Details Predictive model for automated vehicle recommendation system Predictive model for automated vehicle recommendation system

1998-08-18

2001-05-15

Cuchlinski, Jr., William A. (Department: 3661)

Data processing: vehicles, navigation, and relative location

Vehicle control, guidance, operation, or indication

Traffic analysis or control of surface vehicle

C701S200000, C701S208000, C701S214000, C340S990000, C340S988000, C340S907000

Reexamination Certificate

active

06233517

ABSTRACT:

TECHNICAL FIELD
This invention relates to vehicle tracking and automatic vehicle location systems. Specifically, the present invention relates to computer-aided dispatch systems.
BACKGROUND ART
Vehicle tracking using an automatic vehicle location (AVL) system is well known in the art. In a typical AVL system, each of a plurality of vehicles communicates its position to a base station. The position of each of the vehicles is commonly superimposed over a simplified map to provide a graphical display of the position of the vehicles with respect to the underlying simplified map. An AVL system can also be used to dramatically enhance the function of a computer aided dispatch (CAD) system.
In one use of a CAD system, a system operator dispatches emergency vehicles to a reported event. For example, a 911 operator using a CAD system receives a phone call reporting a medical emergency. The CAD operator automatically receives information indicating the geographic position of the caller. (Often the event reported by the caller occurs nearby the location from which the caller is calling.) The geographic position of the caller is obtained, for example, by an automatic number identification/automatic location identification (ANI/ALI) system used in conjunction with a geo-file. By accessing an AVL system, the CAD operator is able to visually determine which of the displayed vehicles is positioned near or nearest to the location of the caller. However, an AVL system does not necessarily indicate which of the vehicles can respond most quickly to the location of the event reported by the caller. That is, although a first vehicle may appear to be physically closer than a second vehicle to the location of the event, a vehicle impeding barrier may exist between the first vehicle and the location of the reported event. The vehicle impeding barrier can be, for example, a building, a river, a one way street, and the like. Thus, even though the first vehicle appears to the CAD operator to be closer to the reported event, the second vehicle is able to more quickly respond to the event. Therefore, although an AVL system can enhance the functionality of a CAD system, an AVL system can not reliably determine which of a plurality of vehicles can respond most quickly to an event. The shortcomings of a combined AVL and CAD system are especially prevalent in an environment containing complex geographic features. Such environments include, for example, large metropolitan areas such as San Francisco, New York, Los Angeles, Boston, and the like.
As an additional drawback, prior art CAD systems or combined AVL and CAD systems require the CAD operator to decide which of a plurality of available vehicles can respond most quickly to a reported event. That is, even when the CAD operator is aware of vehicle impeding barriers, the CAD operator must consider the barriers when determining which vehicle to dispatch to the event. Furthermore, often the CAD system operator must choose which vehicle to dispatch while under the extreme pressure associated with handling life threatening events commonly reported to CAD system operators. As a result, CAD system operators may make errors even when the operators are aware of geographic and or vehicle impeding barriers present at or near the reported event. To make matters even worse for CAD system or combined AVL and CAD system operators, dispatch decisions must be made under rigorous time constraints. Hence, even well trained CAD operators may mistakenly dispatch the wrong vehicle or a slower responding vehicle to a reported event.
An automatic vehicle recommendation (AVR) system which swiftly and automatically recommends which of a plurality of vehicles can respond most quickly to an event has been developed and previously disclosed. The AVR system can function in conjunction with an existing AVL or CAD system, and reduces CAD system operator decision making processes.
A potential drawback to an AVR system is the inability to accurately locate the plurality of vehicles within the visual display of the AVR system. That is, there is a resolution associated with the visual display. The smallest resolved component of the visual display may actually encompass a relatively large geographic area, such as a city block or perhaps larger. within that resolution, a vehicle would appear to be in the center of a geographic area, where in reality it may be up against one boundary of the geographic area. Thus, the vehicle may actually have to traverse the width of the geographic area in response to an emergency call, adding to its response time. This additional time would not be recognized by an AVR system or its operator unless the location of the vehicle can be more precisely located. Hence, the AVR system may recommend and/or the operator may select a slower responding vehicle.
Another potential drawback to an AVR system pertains to the methodology employed to generate predicted response times from one geographic position to another. A methodology that uses the crow flies distance provides a reasonable far field approximation for computing response times to distant locations, but inherently includes large uncertainties because it doesn't account for vehicle impeding boundaries or the actual distance to be traveled. Other methods can be used to provide a more accurate prediction, but these methods also can introduce large and unsatisfactory uncertainties into the AVR system.
Finally, there are potential drawbacks associated with the database of predicted vehicle response times. The database would need to be very large if a predicted response time is generated for every possible combination of travel routes from one geographic region to another. Therefore, it may not be practical to provide a predicted response time for every combination. However, a combination that is not considered may turn out to be eventually needed. Hence, the AVR system would not be able to model that combination and thus may recommend one vehicle when in actuality another vehicle is better able to respond. In addition, as described above, the predicted response times may incorporate uncertainties, either because of the methodology employed to make the prediction, or because of factors not recognized during the calculation. Without recognition of how the predicted response time compares to the actual time it takes to respond, the AVR system may not provide the proper recommendation.
Thus, a need exists for an AVR system which can accurately and automatically position vehicles within the monitored geographic region. Such an AVR system should be able to predict with reasonable accuracy the response time for each of a plurality of vehicles for all potential dispatch destinations. Also, such a system should ideally become more accurate through use.
DISCLOSURE OF THE INVENTION
The present invention provides an AVR system accurately and automatically position vehicles within the monitored geographic region. The present invention further provides an AVR system which predicts with reasonable accuracy the response time for each of a plurality of vehicles for all potential dispatch destinations. Also, the present invention provides a system which becomes more accurate through use.
Specifically, in one embodiment, the present invention divides a geographic region into a plurality of geographic cells. The present automatic vehicle recommendation system then calculates a response time for each of the plurality of vehicles to travel from each of the geographic cells to every other of the geographic cells. The position of each of the plurality of vehicles is monitored by an automatic vehicle location (AVL) system. The present automatic vehicle recommendation system then determines which of the plurality of vehicles is able to respond most quickly to an event reported to a computer aided dispatch (CAD) system. The event is located within one of the geographic cells. The present invention then communicates to the CAD system which of the plurality of vehicles is able to respond most quickly to the reported event. Thus, the present

Affiliated with

Also associated with

Rating

Predictive model for automated vehicle recommendation system does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Predictive model for automated vehicle recommendation system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Predictive model for automated vehicle recommendation system will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2470211