Accident risk model determination using autonomous vehicle operating data
Methods and systems for evaluating the effectiveness of autonomous operation features of autonomous vehicles using an accident risk model are provided. According to certain aspects, an accident risk model may be determined using effectiveness information regarding autonomous operation features associated with a vehicle. The effectiveness information may indicate a likelihood of an accident for the vehicle and may include test data or actual loss data. Determining the likelihood of an accident may include determining risk factors for the features related to the ability of the features to make control decisions that successfully avoid accidents. The accident risk model may further include information regarding effectiveness of the features relative to location or operating conditions, as well as types and severity of accidents. The accident risk model may further be used to determine or adjust aspects of an insurance policy associated with an autonomous vehicle.
1. A computer-implemented method of evaluating effectiveness of an autonomous or semi-autonomous vehicle technology, the method comprising:
implementing, by one or more processors, the autonomous or semi-autonomous vehicle technology within a virtual test environment configured to simultaneously test multiple autonomous or semi-autonomous vehicle technologies;
presenting, by the one or more processors, virtual test sensor data to the autonomous or semi-autonomous vehicle technology implemented within the virtual test environment, wherein the virtual test sensor data simulates sensor data for operating conditions associated with a plurality of test scenarios within the virtual test environment;
generating, by the one or more processors, test responses of the autonomous or semi-autonomous vehicle technology implemented within the virtual test environment in response to the virtual test sensor data;
generating, by the one or more processors, an accident risk model indicating one or more risk levels for vehicle accidents associated with the autonomous or semi-autonomous vehicle technology based upon the test responses;
receiving, at the one or more processors, actual accident data associated with accidents involving vehicles using the autonomous or semi-autonomous vehicle technology in a non-test environment;
adjusting, by the one or more processors, the accident risk model based upon the actual accident data by adjusting at least one of the one or more risk levels of the accident risk level;
identifying, by the one or more processors, a customer vehicle having the autonomous or semi-autonomous vehicle control technology; and
generating or updating, by the one or more processors, an insurance policy associated with the customer vehicle based upon the adjusted at least one of the one or more risk levels of the adjusted accident risk model.
2. The computer-implemented method of claim 1 , wherein:
generating the test responses includes generating test responses relative to additional test responses of another autonomous or semi-autonomous vehicle technology; and
the one or more risk levels of the accident risk model are generated based in part upon compatibility of the test responses of the autonomous or semi-autonomous vehicle technology with the additional test responses of the other autonomous or semi-autonomous vehicle technology.
3. The computer-implemented method of claim 2 , wherein the compatibility of the test responses and the additional test responses is determined for a plurality of versions of the other autonomous or semi-autonomous vehicle technology.
4. The computer-implemented method of claim 1 , wherein generating the accident risk model includes determining the one or more risk levels based upon an effectiveness metric associated with the autonomous or semi-autonomous vehicle technology calculated from the test responses.
5. The computer-implemented method of claim 1 , further comprising:
causing, by the one or more processors, information regarding all or a portion of the insurance policy to be presented to a customer associated with the customer vehicle via a display of a customer computing device for review.
6. The computer-implemented method of claim 1 , wherein the virtual test sensor data includes virtual test communication data simulating autonomous vehicle-to-vehicle communication data.
7. The computer-implemented method of claim 1 , wherein the autonomous or semi-autonomous vehicle technology involves at least one of a vehicle self-braking functionality or a vehicle self-steering functionality.
8. The computer-implemented method of claim 1 , wherein the operating conditions are associated with one or more of the following: a road type, a time of day, or a weather condition.
9. A computer system for evaluating effectiveness of an autonomous or semi-autonomous vehicle technology, comprising:
one or more processors;
one or more program memories coupled to the one or more processors and storing executable instructions that, when executed by the one or more processors, cause the computer system to:
implement the autonomous or semi-autonomous vehicle technology within a virtual test environment configured to simultaneously test multiple autonomous or semi-autonomous vehicle technologies;
present virtual test sensor data to the autonomous or semi-autonomous vehicle technology implemented within the virtual test environment, wherein the virtual test sensor data simulates sensor data for operating conditions associated with a plurality of test scenarios within the virtual test environment;
generate test responses of the autonomous or semi-autonomous vehicle technology implemented within the virtual test environment in response to the virtual test sensor data;
generate an accident risk model indicating one or more risk levels for vehicle accidents associated with the autonomous or semi-autonomous vehicle technology based upon the test responses;
receive actual accident data associated with accidents involving vehicles using the autonomous or semi-autonomous vehicle technology in a non-test environment;
adjust the accident risk model based upon the actual accident data by adjusting at least one of the one or more risk levels of the accident risk level;
identify a customer vehicle having the autonomous or semi-autonomous vehicle control technology; and
generate or update an insurance policy associated with the customer vehicle based upon the adjusted at least one of the one or more risk levels of the adjusted accident risk model.
10. The computer system of claim 9 , wherein:
the executable instructions that cause the computer system to generate the test responses cause the computer system to generate test responses relative to additional test responses of another autonomous or semi-autonomous vehicle technology; and
the one or more risk levels of the accident risk model are generated based in part upon compatibility of the test responses of the autonomous or semi-autonomous vehicle technology with the additional test responses of the other autonomous or semi-autonomous vehicle technology.
11. The computer system of claim 10 , wherein the compatibility of the test responses and the additional test responses is determined for a plurality of versions of the other autonomous or semi-autonomous vehicle technology.
12. The computer system of claim 9 , wherein the executable instructions that cause the computer system to generate the accident risk model further cause the computer system to determine the one or more risk levels based upon an effectiveness metric associated with the autonomous or semi-autonomous vehicle technology calculated from the test responses.
13. The computer system of claim 9 , wherein the executable instructions further cause the computer system to:
communicate to a customer computing device, via a communication network, information regarding all or a portion of the insurance policy to be presented to a customer associated with the customer vehicle for review via a display of the customer computing device.
14. The computer system of claim 9 , wherein the virtual test sensor data includes virtual test communication data simulating autonomous vehicle-to-vehicle communication data.
15. A tangible, non-transitory computer-readable medium storing executable instructions for evaluating effectiveness of an autonomous or semi-autonomous vehicle technology that, when executed by at least one processor of a computer system, cause the computer system to:
implement the autonomous or semi-autonomous vehicle technology within a virtual test environment configured to simultaneously test multiple autonomous or semi-autonomous vehicle technologies;
present virtual test sensor data to the autonomous or semi-autonomous vehicle technology implemented within the virtual test environment, wherein the virtual test sensor data simulates sensor data for operating conditions associated with a plurality of test scenarios within the virtual test environment;
generate test responses of the autonomous or semi-autonomous vehicle technology implemented within the virtual test environment in response to the virtual test sensor data;
generate an accident risk model indicating one or more risk levels for vehicle accidents associated with the autonomous or semi-autonomous vehicle technology based upon the test responses;
receive actual accident data associated with accidents involving vehicles using the autonomous or semi-autonomous vehicle technology in a non-test environment;
adjust the accident risk model based upon the actual accident data by adjusting at least one of the one or more risk levels of the accident risk level;
identify a customer vehicle having the autonomous or semi-autonomous vehicle control technology; and
generate or update an insurance policy associated with the customer vehicle based upon the adjusted at least one of the one or more risk levels of the adjusted accident risk model.
16. The tangible, non-transitory computer-readable medium of claim 15 , wherein:
the executable instructions that cause the computer system to generate the test responses cause the computer system to generate test responses relative to additional test responses of another autonomous or semi-autonomous vehicle technology; and
the one or more risk levels of the accident risk model are generated based in part upon compatibility of the test responses of the autonomous or semi-autonomous vehicle technology with the additional test responses of the other autonomous or semi-autonomous vehicle technology.
17. The tangible, non-transitory computer-readable medium of claim 16 , wherein the compatibility of the test responses and the additional test responses is determined for a plurality of versions of the other autonomous or semi-autonomous vehicle technology.
18. The tangible, non-transitory computer-readable medium of claim 15 , wherein the executable instructions that cause the computer system to generate the accident risk model further cause the computer system to determine the one or more risk levels based upon an effectiveness metric associated with the autonomous or semi-autonomous vehicle technology calculated from the test responses.
19. The tangible, non-transitory computer-readable medium of claim 15 , further storing executable instructions that, when executed by at least one processor of the computer system, cause the computer system to:
cause information regarding all or a portion of the insurance policy to be presented to a customer associated with the customer vehicle via a display of a customer computing device for review.
20. The tangible, non-transitory computer-readable medium of claim 15 , wherein the virtual test sensor data includes virtual test communication data simulating autonomous vehicle-to-vehicle communication data.