We just released the first version of the simulated autonomous vehicles (AVs) involved in the crash database. Thanks to our undergraduate student researcher James Holland, for his effort.

By gathering AV accidents’ data and simulating them, it is possible to identify the fundamental causes of crashes and test and verify AVs ahead of time. The proposed framework creates a crash scenario generator that works based on actual AV crashes. The database includes recent crashes related to autonomous vehicles with their locations and descriptions. The virtual scenarios of accidents are available for further research-related purposes. It also includes the location of test tracks for testing and verifications of AVs.

The database can be found here:

https://www.rancs-lab.com/AMI/