Accident risk model determination using autonomous vehicle operating data

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:
• present virtual test sensor data to the autonomous or semi-autonomous vehicle technology implemented 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 vehicles 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, the actual accident data comprising data collected by a vehicle sensor;
• and
• 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.

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:

• present virtual test sensor data to the autonomous or semi-autonomous vehicle technology implemented 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 vehicles 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, the actual accident data comprising data collected by a vehicle sensor;
• and
• 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.