Predictive traffic tools to guide AVs and slash congestion

The number of mobility options on offer to city residents and visitors has been growing rapidly, but urban congestion remains a serious problem. It not only exacerbates air pollution and wastes time, but it is also a tremendous drag on local economies. Traffic consultancy INRIX estimates that congestion can cost cities tens of billions of dollars a year.

A new machine learning algorithm could help with that. The US Department of Energy's Pacific Northwest National Laboratory (PNNL) has developed a tool, called TranSEC (Transportation State Estimation Capability), to provide urban traffic engineers with actionable information about traffic patterns in a given city. While other tools on the market can optimise an individual traveller's journey from A to B, city traffic engineers want to ensure all vehicles reach their destinations efficiently. There may be times when a route that is efficient for an individual driver results in too many vehicles accessing a road that was never designed to handle high volumes of traffic.