Our roads may become safer for motorists and pedestrians through a new, innovative artificial intelligence (AI) tech project by the University of Hawaiʻi at Mānoa College of Engineering.
The project, led by Department of Civil, Environmental and Construction Engineering Professor Guohui Zhang, earned a $100,000 award from the U.S. Department of Transportation (DOT). Zhang’s proposed design concept, “Toward Vision Zero: Sensing, Predicting, and Preventing Intersection Collisions,” was one of 15 winners among 120 entries in the U.S. DOT Intersection Safety Challenge.
Creating an AI traffic safety net
The challenge encourages new technologies to improve intersection safety by addressing risks for vehicles, pedestrians, bicyclists and other road users.
Zhang’s project aims to create a safety system for intersections using advanced technologies such as sensors, edge computing, machine learning and wireless communication. The system will focus on preventing crashes between vehicles and pedestrians. Artificial intelligence will be used to identify and track objects such as pedestrians and vehicles in real-time. By predicting collisions and issuing warnings, the system aims to improve intersection safety and reduce costs. Additionally, the technology may have positive effects on traffic systems, including reducing congestion and emissions.
“We are truly honored to receive this exceptionally competitive award, resulting from the hard work and collaboration between UH, Hawaiʻi Department of Transportation (HDOT), the Pacific International Center for High Technology Research, and our industry partner, NEC Corporation of America,” Zhang said. “We also appreciate all the support and leadership from Edwin Sniffen and Robin Shishido at HDOT and Brennon Morioka at the UH College of Engineering for their guidance and contributions. This kind of project will go a long way in helping Hawaiʻi in its efforts towards reducing accidents and saving lives on Hawaiʻi’s roads by utilizing today’s cutting-edge innovation and technology.”
Zhang’s project will now advance into the next phase of the challenge, where teams are expected to develop, train and improve algorithms for the detection, localization and classification of vulnerable road users and vehicles using U.S. DOT-supplied sensor data collected at a controlled test roadway intersection.