FURI | Spring 2020

Blame All vs. Blame Me in AVs

Due to the increasing desire to make the roads safer with the innovation in autonomous vehicles (AVs), it is imperative to understand what information should go into a vehicle’s control decisions. Currently, the research is numerically evaluating a game between two AVs with information asymmetry at an unregulated intersection in order to plot the cost functions of each AV as well as their estimation for the cost function of the other AV to find if behaviors converge based on two different control strategies (Blame-me and Blame-all). It is hypothesized that the Blame-all strategy would lead to convergence of behavior.