University of California Institute of Transportation Studies

Modern adaptive traffic signals rely on sensors like inductive loop detectors (ILDs) embedded in pavement and cameras mounted above roadways to detect the presence of vehicles and adjust signal timing accordingly. These systems aim to reduce congestion and improve safety. However, bad weather –like rain and fog—can cause sensors to fail or generate false signals, creating unnecessary delays and safety hazards. In addition, an emerging concern is “spoofing”, where individuals intentionally disrupt ILDs by sending false electrical signals. This could cause traffic signals to mismanage flow, increasing congestion and risk of crashes. Sensor fusion, the practice of combining ILD and camera data, is gaining traction as a solution for improving performance. To better understand the vulnerabilities and potential advantages, we tested how sensor fusion performed under a simulated range of traffic and weather conditions and how spoofing attacks affected signal control and object tracking performance.