Evaluation and Application of Automated Traffic Sensor Data under Indian Conditions

The basic requirement of any Intelligent Transportation Systems (ITS) application is the real time capturing of traffic system characteristics. With the advancement in intelligent transportation sector, several types of traffic sensors based on different sensing technology are available for vehicle sensing. However, most of the available sensors are developed for traffic conditions that exist in western countries where, only one or two classes of vehicles move in a lane disciplined fashion. However, traffic conditions in India are different from that, with heterogeneity and lack of lane discipline. Under such conditions, the traffic detectors that are proven to be working well under lane disciplined and homogeneous traffic conditions may not work well. In this paper, four popular sensor technologies - one video image processing software, one video sensor, one radar based and one infra-red based detection technologies, are evaluated for their performance under Indian traffic conditions. The evaluation was carried out for commonly used parameters, volume, classification and speed, based on statistical analysis. Use of the data from best performing sensors for selected ITS application is also demonstrated. The application demonstrated is on real time congestion information by estimating traffic density from the location based data collected using these sensors.