Panos Michalopoulos, Professor, Civil Engineering
Researchers have developed a myriad of work-zone intelligent transportation systems (WZITS) to warn drivers of dangerous traffic conditions within work zones. Unfortunately, when actual traffic conditions don't correspond with sensor information, the effectiveness of a WZITS is diminished and false warnings can occur; false warnings confuse drivers and reduce the credibility of the system, which leads to its being ignored. The occurrence of these false positives could be reduced if we could accurately detect the location of queue tails and the speeds of vehicles proximal to the work zone. The goal of this project is to develop a video-based queue detection system that is capable of this. The system will be low cost and portable to work zones. Researchers will accomplish this goal by modifying an existing machine-vision-based device used for collecting traffic data. The new device will wirelessly transmit data to a remote, upstream location, where roadside emergency warning devices will be triggered to warn drivers.