Berkeley MPC Lab has developed Iterative Learning Model Predictive Control (LMPC) forecasting to plan the vehicle trajectory looking a few seconds into the future. Here is a closed-loop trajectory (in black) after 5 laps of learning. The LMPC plans an open-loop trajectory (in red) to minimize lap time and land in the safety set (in blue).

The team will put LMPC to the test at the Indianapolis Motor Speedway!