Research Team: Pravin Varaiya (lead) and Alex Kurzhanskiy
UC Campus(es): UC Berkeley
Problem Statement: Multimodal intersections present a very demanding environment for all parties involved. Challenges arise from simultaneous interactions among pedestrians, bicycles and vehicles; complex vehicle trajectories; absence of lane markings to guide vehicles; signals that do not fully reveal who has the right of way; invisible vehicle approaches; obstructions; and illegal movements. It is often the case that on-board sensing equipment is not sufficient for a vehicle’s autopilot to steer the vehicle through an intersection safely and efficiently. In such situations infrastructure-to-vehicle (I2V) technology can provide additional information enhancing the vehicle’s awareness about real-time intersection dynamics.
Project Description: This project seeks to remove one important cause of intersection accidents: drivers, pedestrians and cyclists make mistakes because they lack sufficient information about the movement of others as they proceed through an intersection. This missing information can be supplied by an ‘intelligent intersection’. An intelligent intersection describes the signal from all approaches; predicts when the signal phase will change; uses sensor data to determine which blind spots are occupied; and predicts red light violations before they occur. The intelligent intersection broadcasts this information via radio and can be received by a connected vehicle or indeed anyone in the intersection with a smartphone or bluetooth device, so most intersection users will get this information. The objective of this research is to design intelligent intersection infrastructure and evaluate its performance in terms of safety and mobility benefits.