Project description

Intersection control is currently done using traffic lights which are timed based on limited sensor data generated by inductor loops located under the lane surface near the intersection. The traffic lights have limited information regarding the size of the queue and demand in either direction. This limited information makes it difficult to come up with accurate measurements of demand and design the traffic light cycles to optimize mobility. Currently available sensors such as vision and radar which can provide additional and more accurate information are costly to install and maintain and the infrastructure simply does not have the additional funds to carry out such significant changes on a big scale. On the other hand, car manufacturers have been producing vehicles with the latest in sensor technology and have capabilities that if used properly in a cyber based system can replace traffic lights by creating a connected and highly coordinated environment based on software and intelligence. The purpose of this project is to develop an intersection control system where vehicles approaching the intersection will be assigned paths and speed trajectories so that no two vehicles will be put in a colliding path. We assume that vehicles can communicate their location, speed, dynamical characteristics, and direction of travel to a virtual central coordinator which will issue them instructions to cross (green to procced) following a reserved path at a certain desired speed based on certain protocols. The vehicles are expected to follow the instructions of the coordinator and use their on-board sensors to cross the intersection safely in case unexpected obstacles are in their way providing another layer of safety.

The proposed system takes advantage of the current technologies that are already in most vehicles to create a connecting environment where vehicles can communicate with a central coordinator and possibly with other vehicles to improve mobility and enhance safety. The importance of the approach if successful is that hardware and physical elements such as traffic lights and inductor loops will be replaced with software and intelligence in a cyber space Intersection Control of Connected Vehicles for Mobility and Safety which is much easier to maintain and at a much lower cost. In addition, there is a strong potential for significant improvements in mobility as proper dynamic routing based on demand from each vehicle or group of vehicles will eliminate some of the deficiencies and dead periods where vehicles are waiting for green when no vehicles are crossing from the other direction. Safety is also expected to improve by having two layers of safety, one on the infrastructure coordination level and one on the vehicle level using its own on-board sensors and collision avoidance systems to confirm that the path assigned by the coordinator is safe to follow.

In this project we plan to design a traffic light-free intersection controller and quantify its benefits with respect to mobility and safety by comparing it with current traffic lights. In addition, we will examine the evolution of such a highly coordinated system as it transitions from current situation where some vehicles will participate in the coordination, and some will not as they may not have the adequate technologies. In such an environment the integration of the current traffic lights with the proposed central coordinator will be investigated. The expected outcomes of the project are the design and evaluation of a virtual environment for controlling traffic at intersections by replacing current traffic lights and expensive hardware which is costly to buy and maintain with less costly software and intelligence by taking advantage of current and upcoming technologies on vehicles.