Fighting Chaotic Road Congestion Second PhD Annual Progress Report

Road congestion is a common problem worldwide. In developed countries, automated congestion detection techniques are used in road travel assisting applications. But these techniques are mostly inapplicable in developing regions due to high cost and their assumptions of orderly traffic. We present two techniques for effective congestion detection in chaotic traffic of developing regions. Our first system is RoadSoundSense, an acoustic sensing based road congestion monitoring technique. We detail the design of a prototype which, deployed by the roadside, processes road noise and sends various metrics to a remote server. Data from deployment of this prototype on six Mumbai roads, validated against manually observed ground truth, shows feasibility of per minute congestion monitoring. K-means clustering gives 90% accuracy on average, in training our system to work on a new road. Deployment data from one road for six days shows the temporal variation in traffic state for that road. We also present WirelessAcrossRoad, which exploits the variation in wireless link characteristics when line of sight conditions between a wireless sender and receiver vary. The system comprises of a wireless sender-receiver pair across a road. The sender continuously sends packets. The receiver measures metrics like signal strength, link quality and packet reception. These metrics show a marked change in values depending on traffic-state on the road in between. From about 15 hours of data on two different Mumbai roads, we show that we can classify traffic states as free-flowing and congested using a decision tree based classifier with 97% accuracy.