Future directions in traffic signal control

Michael G.H. Bell

Abstract

This paper reviews in outline the principal developments in traffic signal control since its inception, and explains the current interest in decentralised systems and the discrete time, rolling horizon approaches to real time control. It is argued that signal control alone is insufficient to solve growing problems of urban congestion, and that recent technological advances in Road Transport Informatics are making integrated systems more attractive. One potential area of integration is automatic debiting and signal control. Through automatic debiting for road use and parking it is possible to confront the motorist with the full marginal cost of his activity, inclusive of the environmental impact, leading to more socially efficient decision-making behaviour. Signal control could then be oriented to cost minimisation. A second area of integration is in-vehicle information and signal control. Efficient decision-making requires well-informed drivers. The systems have shared information requirements and interact with each other strongly. A third area of integration is public transport management and signal control. In general, priority is required for public transport and the emergency services on designated lanes and at junctions, without disadvantaging other road users unnecessarily. As with the other two areas of integration, there is considerable scope for information exchange, leading to improved state estimation, and joint optimisation.