Systems and Methods for Building Road Models, Driver Models, and Vehicle Models and Making Predictions Therefrom
Systems and methods for building road models, driver models, and vehicle models and making predictions may use driving data received from a plurality of vehicles being driving along a path by a driver. The driving data including a minimum of vehicle location preferably supplemented by data related to the time of observations, vehicle dynamics, and various vehicle sensors. The received data may be categorized into a plurality of maneuvers and a plurality of variables that describe the maneuvers and the received driving data may be identified. A road model may then be built for the path based on the identified variables and maneuvers and the behaviors of a particular driver compared to the road model to assess their driving and intervene as appropriate.
1 . A computer implemented invention comprising:
receiving driving data from a plurality of vehicles being driving along a path, each of the vehicles of the plurality being driven by a driver, the driving data including vehicle location information
categorizing the data into a plurality of maneuvers;
identifying a plurality of variables that describe the maneuvers and the received driving data;
building a road model for the path based on the identified variables and maneuvers, the road model including a set of reference trajectories for a subsequent vehicle being driven at an identical or similar location along the path;
receiving driving data from a vehicle separate from the plurality of vehicles as it is being driven on the path by its respective driver;
determining one or more characteristics of the driver of the plurality of drivers based on the driving data received from the vehicle; and
building a driver model for the driver using the driver behavior.
2 . The computer implemented method of claim 1 , further comprising:
comparing the characteristics of the driver to the road model to determine driver behavior;
determining whether driver behavior is consistent with the road model; and
executing a predetermined action in the vehicle responsively to the determination of whether the driver behavior is consistent with the road model, the predetermined action being at least one of providing a message to the driver and automatically adjusting a manner in which the vehicle is driven.
3 . The computer implemented method of claim 2 , further comprising:
assigning a risk score to the driver responsively to the comparison of the driver behavior to the road and the driver models, wherein the predetermined action is responsive to the risk score.
4 . The computer implemented method of claim 3 , wherein the risk score is provided to at least one of the driver and a third party.
5 . The computer implemented method of claim 1 , wherein the driving data includes data collected by at least one of a vehicle mounted radar, a vehicle mounted lidar, a vehicle mounted optical vision system, and a vehicle mounted ultrasonic sensor.
6 . The computer implemented method of claim 1 , further comprising:
determining one or more characteristics of a vehicle of the plurality of vehicles driven by the driver based on the driving data received from the vehicle;
comparing the characteristics of the vehicle to the road model to determine vehicle behavior; and
building a vehicle model for the driver using the driver behavior.
7 . The computer implemented method of claim 6 , further comprising:
receiving vehicle data from a third party; and
updating the vehicle model to include the vehicle data received from the third party.
8 . The computer implemented method of claim 6 , further comprising:
receiving vehicle characteristic data relating to the vehicle; and
updating the vehicle model based on the received vehicle characteristic data relating to the vehicle.
9 . The computer implemented method of claim 6 , further comprising:
comparing the vehicle behavior with the driver model and the road model;
determining whether the vehicle behavior is consistent with the driver model and the road model; and
executing an action in the vehicle responsively to a determination of whether the driver behavior is consistent with the road model and the driver model, the predetermined action being at least one of providing a message to the driver and automatically adjusting a manner in which the vehicle is driven.
10 . The computer implemented method of claim 6 , further comprising:
assigning a risk score to the vehicle responsively to the comparison of the vehicle behavior to the road model, the driver model, and the vehicle model.
11 . The computer implemented method of claim 1 , further comprising:
building a driver model for the driver using the driver behavior.
12 . The computer implemented method of claim 11 , further comprising:
receiving driver data from a third party; and
updating the driver model to include the driver data received from the third party.
13 . The computer implemented method of claim 11 , further comprising:
receiving demographic data about the driver; and
segmenting the driver model based on the received demographic data.
14 . The computer implemented method of claim 11 , further comprising:
comparing the driver behavior with the driver model;
determining whether the driver behavior is consistent with the driver model; and
executing a predetermined action in the vehicle responsively to the determination of whether the driver behavior is consistent with the driver model.
15 . The computer implemented method of claim 1 , wherein the driver is not a human being.