Techniques for autonomous vehicle accident mitigation
Techniques are described for scenario data generation. An example method can include processing location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network. The method can further include determining a location-specific element based on the location-specific sensor information. The method can further include determining a location-specific contributing factor of the traffic accident based on the location-specific element. The method can further include generating a location-specific model for simulating the traffic accident based on the location-specific contributing factor. The method can further include determining control instructions for a location-specific driving maneuver for an autonomous vehicle to mitigate a future traffic accident based on the location-specific model. The method can further include causing the control instructions to be transmitted to an actuator of the autonomous vehicle to perform the location-specific driving maneuver.
1 . A method, comprising:
processing, by a computing system, location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network, the sensor having a field of view based on being arranged on the infrastructure, and the location-specific sensor information based on the field of view of the sensor;
determining, by the computing system, a location-specific element based on the location-specific sensor information;
determining, by the computing system, a location-specific contributing factor of the traffic accident based on the location-specific element;
generating, by the computing system, a location-specific model for simulating the traffic accident based on the location-specific contributing factor;
accessing, by the computing system, the location-specific sensor information based at least in part on generating the location-specific model;
determining, by the computing system, a fidelity of the location-specific model to the location-specific sensor information;
determining, by the computing system, an acceptability of the location-specific model based at least in part on the fidelity;
incorporating, by the computing system, a representation of a first hypothetical autonomous vehicle into the location-specific model;
determining, by the computing system, control instructions for a location-specific driving maneuver for a second autonomous vehicle to mitigate a future traffic accident based on the location-specific model, the control instructions determined based on incorporating the representation of the first hypothetical autonomous vehicle into the location-specific model; and
causing, by the computing system, the control instructions to be transmitted to an actuator of the second autonomous vehicle to perform the location-specific driving maneuver thereby controlling the second autonomous vehicle at the location based on the location-specific model.
2 . The method of claim 1 , wherein the location-specific element is a first location-specific element, and wherein the method further comprises:
receiving real time location-specific sensor information from the sensor; and
determining a second location-specific element indicating a traffic accident based on the real time location-specific sensor information, wherein the control instructions are transmitted based on the determining the second location-specific element indicating the traffic accident.
3 . The method of claim 1 , wherein the method further comprises;
generating synthetic information associated with the autonomous vehicle, wherein the synthetic information incorporates a representation of the autonomous vehicle into the location-specific model, and wherein the control instructions are determined based on incorporating the representation of the autonomous vehicle into the location-specific model.
4 . The method of claim 1 , wherein the location-specific sensor information is associated with a stationary space based on the sensor being a stationary sensor, and wherein the method further comprises:
determining a transformation function for converting the location-specific sensor information from a stationary space representation to a moving space representation, and wherein the location-specific model is determined based on the moving space representation.
5 . The method of claim 1 , wherein the location is a first location along a route, wherein the location-specific sensor information is first location specific sensor information, wherein the infrastructure is a first infrastructure, and wherein the method further comprises:
identifying second location-specific sensor information from a second sensor arranged on a second infrastructure at a second location along the route, wherein the location-specific model is determined based on the first location specific sensor information and second location-specific sensor information.
6 . The method of claim 5 , wherein the method further comprises:
determining a route of the autonomous vehicle based on the road network;
determining that the second location is on the route;
determining that a third sensor is located at a third location outside of the route; and
determining to use a second location-specific element associated with the second location to determine the location-specific model instead of a third location-specific element associated with the third location based on the second location being on the route and the third location being located outside of the route.
7 . The method of claim 1 , wherein the location-specific element is determined based on information from a driver assistance system.
8 . The method of claim 1 , wherein the location-specific element is a road element describing a characteristic of a road upon which the traffic accident occurred.
9 . The method of claim 8 , wherein the method further comprises;
determining a vehicle element based on the location-specific sensor information, the vehicle element describing a characteristic of a vehicle involved in the traffic accident, wherein the location-specific model is determined based on the vehicle element.
10 . The method of claim 1 , wherein mitigating the future traffic accident comprises avoiding the future traffic accident.
11 . The method of claim 1 , wherein the method further comprises:
receiving an identification of an element not described in the location-specific sensor information; and
generating a location-specific synthetic element representing the element, and wherein the location-specific model is determined based on the location-specific synthetic element.
12 . A computing system, comprising:
one or more processors; and
one or more non-transitory, computer-readable media including instructions that, when executed, cause one or more processors to:
process location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network, the sensor having a field of view based on being arranged on the infrastructure, and the location-specific sensor information based on the field of view of the sensor;
determine a location-specific element based on the location-specific sensor information;
determine a location-specific contributing factor of the traffic accident based on the location-specific element;
generate a location-specific model for simulating the traffic accident based on the location-specific contributing factor;
access the location-specific sensor information based at least in part on generating the location-specific model;
determine a fidelity of the location-specific model to the location-specific sensor information;
determine an acceptability of the location-specific model based at least in part on the fidelity;
incorporate a representation of a first hypothetical autonomous vehicle into the location-specific model;
determine control instructions for a location-specific driving maneuver for a second autonomous vehicle to mitigate a future traffic accident based on the location-specific model, the control instructions determined based on incorporating the representation of the first hypothetical autonomous vehicle into the location-specific model; and
cause the control instructions to be transmitted to an actuator of the second autonomous vehicle to perform the location-specific driving maneuver thereby controlling the second autonomous vehicle at the location based on the location-specific model.
13 . The computing system of claim 12 , wherein the location-specific element is a first location-specific element, and wherein the instructions that, when executed, further cause the one or more processors to:
receive real time location-specific sensor information from the sensor; and
determine a second location-specific element indicating a traffic accident based on the real time location-specific sensor information, wherein the control instructions are transmitted based on determining the second location-specific element indicating the traffic accident.
14 . The computing system of claim 12 , wherein the instructions that, when executed, further cause the one or more processors to:
generating synthetic information associated with the autonomous vehicle, wherein the synthetic information incorporates the autonomous vehicle into the location-specific model, and wherein the control instructions are determined based on incorporates the autonomous vehicle into the location-specific model.
15 . The computing system of claim 12 , wherein the location-specific sensor information is associated with a stationary space based on the sensor being a stationary sensor, and wherein the instructions that, when executed, further cause the one or more processors to:
determining a transformation function for converting the location-specific sensor information from a stationary space representation to a moving space representation, and wherein the location-specific model is determined based on the moving space representation.
16 . The computing system of claim 12 , wherein the location is a first location, wherein the location-specific sensor information is first location specific sensor information, wherein the infrastructure is a first infrastructure, and wherein the instructions that, when executed, further cause the one or more processors to:
identify a second location-specific element from a second sensor arranged on a second infrastructure at a second location on the road network, wherein the location-specific model is determined based on the second location-specific element.
17 . The computing system of claim 16 , wherein the instructions that, when executed, further cause the one or more processors to:
determine a route of the autonomous vehicle based on the road network;
determine that the second location is on the route; and
determine to use the second location-specific element to determine the location-specific model instead of third location-specific element associated with a third sensor that is outside of the route based on the second location being on the route.
18 . The computing system of claim 12 , wherein the location-specific element is determined based on information from a driver assistance system.
19 . One or more non-transitory computer-readable media having stored thereon a sequence of instructions which, when executed, cause one or more processors to:
process location-specific sensor information of a traffic accident at a location on a road network, the location-specific sensor information received from a sensor arranged on an infrastructure at the road network, the sensor having a field of view based on being arranged on the infrastructure, and the location-specific information based on the field of view of the sensor;
determine a location-specific element based on the location-specific sensor information;
determine a location-specific contributing factor of the traffic accident based on the location-specific element;
generate a location-specific model for simulating the traffic accident based on the location-specific contributing factor;
access the location-specific sensor information based at least in part on generating the location-specific model;
determine a fidelity of the location-specific model to the location-specific sensor information;
determine an acceptability of the location-specific model based at least in part on the fidelity;
incorporate a representation of a first hypothetical autonomous vehicle into the location-specific model;
determine control instructions for a location-specific driving maneuver for a second autonomous vehicle to avoid a future traffic accident based on the location-specific model, the control instructions determined based on incorporating the representation of the first hypothetical autonomous vehicle into the location-specific model; and
cause the control instructions to be transmitted an actuator of the second autonomous vehicle to perform the location-specific driving maneuver thereby controlling the second autonomous vehicle at the location based on the location-specific model.
20 . The one or more non-transitory, computer-readable media of claim 19 , wherein the location-specific element is a first location-specific element, and wherein the instructions that, when executed, further cause the one or more processors to:
receive real time location-specific sensor information from the sensor; and
determine a second location-specific element indicating a traffic accident based on the real time location-specific sensor information, wherein the control instructions are transmitted based on determining the second location-specific element indicating the traffic accident.