We want to use SODA.Sim, which was recently integrated into VisionPilot stack, for qualitatively testing the performance of current VisionPilot's longitudinal control.
-
Get SODA.Sim up and running with the latest VisionPilot.
-
Implement NCAP scenarios for collision avoidance and AEB, we can start with 3 or 4 basic scenarios like this:
- Scenario 1: The ego car approaches a static vehicle in its lane and has to come to a safe stop.
- Scenario 2: The ego car approaches a slower moving vehicle in its lane and has to slow down and follow at the correct speed.
- Scenario 3: The ego car is driving normally and another vehicle cuts into the lane, causing the ego car to slow down.
- Scenario 4: There is no car in the lane and the ego car drives normally.
-
Right now, the control layer isn't integrated, so VisionPilot is not sending control signals to SODA.Sim, we need to implement a controller interface as well. The controller interface will map the VisionPilot control outputs to what is expected by SODA.Sim.
We want to use SODA.Sim, which was recently integrated into VisionPilot stack, for qualitatively testing the performance of current VisionPilot's longitudinal control.
Get SODA.Sim up and running with the latest VisionPilot.
Implement NCAP scenarios for collision avoidance and AEB, we can start with 3 or 4 basic scenarios like this:
Right now, the control layer isn't integrated, so VisionPilot is not sending control signals to SODA.Sim, we need to implement a controller interface as well. The controller interface will map the VisionPilot control outputs to what is expected by SODA.Sim.