Simulating Operations Applications of Vehicle Infrastructure Integration

Abstract

Vehicle infrastructure integration (VII) is receiving significant attention because of its potential to provide a wide range of benefits. However, to inform design, deployment, and operations decisions best, it will be necessary to evaluate potential VII-enabled applications fully. A research effort investigated a VII-enabled traffic monitoring application. First, the VII simulation environment developed to support the research was described. Next, traffic monitoring coverage as a function of VII roadside unit (RSU) configuration (i.e., number of units and placement) was investigated. The results of the study indicated that, on the basis of current guidance in the VII architecture, roughly 55% of the sections in the network would be within the direct range of an RSU. Finally, the accuracy and the coverage of monitored links were evaluated. The results of the analysis demonstrated that only 35% of RSU-covered sections were monitored at 1% penetration rates, but the combined effect of direct RSU coverage and indirect coverage through vehicle storage was quite high (around 60%) for a heavily traveled urban network. The results also illustrated that at low levels of vehicles equipped with VII (1% to 10%), the mean absolute error of speed estimation was about 4 mph. However, as the penetration rate increase to 30%, the error dropped to even lower values, roughly 2.5 mph. Finally, it was found that with a default buffer value of 30 snapshots of vehicle storage, the greatest error in indirectly covered sections for mean freeway section speed was 6 mph at 1% penetration rate.