A description is given of a control experiment aimed at the development of a computer methodology for improving the flow of traffic through the Lincoln Tunnel of New York City. Some remarks are given on the potential spin-offs from this experiment which may influence the development of other traffic control systems.
