* Preconditions: Detections generated by radar detection + tracking algorithm (including ML based post-processing/classification) and sent for triggering camera shall be supervised by a sophisticated mechanism (probably inside the camera) in order to detect anomalies that can overload the overall system. * Input conditions / steps: image from simulator/real-world containing thousands of passes of various vehicles (dense traffic) including radar data with anomalies. * Expected result: The system should detect abnormality in incoming detections (based on estimation of typical traffic/worst traffic scenario) from radar side, cut-out radar and immediately report it.