California Partners for Advanced Transportation Technology

This study reports on a method to accomplish sensor validation and fusion in Intelligent Transportation Systems (ITS). The method is based on probabilistic and fuzzy techniques that express a confidence in the sensor data and take into account environmental factors and the state of the system. Sensor data fusion uses the confidence assigned to each sensor reading and integrates them into one reading. Noise and failure are filtered from the data and lead to a safety improvement in ITS.