Models, Simulation, and Performance of Fully Automated Highways

This research seeks to provide an understanding of the functioning of a fully automated highway system (AHS). In order to do that, several detailed questions of AHS design need to be answered with a level of detail sufficient to permit the development of mathematical and computer models of the AHS. This is the modeling stage. The next stage of research is the simulation stage. Its objective is to realize those models in a simulation and visualization program, called SmartPath. SmartPath has undergone several revisions. The latest version, SmartPath 2.Oi, is interactive. The user can sit in front of a workstation and “drive” a vehicle on the AHS, while the rest of the vehicles are automatically controlled. SmartPath has proved to be an effective tool for: testing strategies for the automatic coordination and control of individual vehicles, testing TMC (Traffic Management Center) strategies for diverting AHS lanes when an incident occurs, evaluating AHS capacity under a variety of operating regimes (including ACC and platoon organization) and, not least, offering people a limited visual appreciation of what an AHS might look like and how one might experience driving on it. The third stage of research is the performance evaluation of the AHS under a variety of operating regimes and patterns of demand. Two regimes are investigated in depth: the first assumes ACC (adaptive cruise control), which some experts believe to be the form in which AHS will be realized in the near term; the second assumes the coordinated movement of vehicles organized in platoons. The main performance measures considered are: achievable flows under different patterns of origin and destination, and the ability of the TMC to recover smooth traffic flow following a lane-blocking incident. The research makes three contributions. First, under the idealized assumptions of the simulations, it is established that the platoon organization of vehicles can achieve flows on the order of 6,000 to 8,000 vehicles per lane per hour (compared with a maximum of 2,000 on today’s best highways). Buttressed by other research within PATH, this lends credibility to the claim that automated highway systems can provide dramatic increases in capacity. Second, the simulation program SmartPath creates a useful and versatile synthetic environment for developing, validating and comparing the effectiveness of a variety of strategies for the control of AHS. Third, this and companion research within *Work supported by the PATH program, Institute of Transportation Studies, University of California, Berkeley, and by National Science Foundation under Grants ECS9111907 and IR19120074.