Next-generation traffic signals using artificial intelligence-based video analytics for safe, efficient and green intersections

Road traffic crashes result in 1,200 fatalities and another 36,700 injuries on Australian roads each year. Signalised intersections represent a high-risk node in a transportation network, but the current signal designs only consider efficiency but not safety. This project aims to unleash the power of artificial intelligence (AI) and integrate with the advanced extreme value models for proactive and efficient detection of crash risk in real-time. Its innovations lie on 1) deriving the fundamental relationship between crashes and conflicts by harnessing the data mining and analytical capabilities of AI; 2) developing a novel traffic signal control system optimising safety, efficiency and sustainability of signalised intersections in real-time. The proposed real-time traffic signal system will fundamentally transform the intersection operation across the world and lead to reductions in road fatalities, injuries and emissions in Australia.