Autonomous vehicles (self-driving cars), with vehicle-to-vehicle and vehicle-to- infrastructure communication, have the potential to significantly reduce congestion by allowing for closer following distances and centrally controlled vehicle routing. The speed profile of traffic would also be altered because the concept of sight-distance and reaction time would no longer apply, and stopping distances and speeds could be altered based on road-surface conditions so that the 0.35g deceleration currently assumed in highway design would likely be significantly higher in most instances (see Chapters 2 and 3).
In terms of the impact that autonomous vehicles would have on route selection and performance, one could imagine at least two possible outcomes:
1. In all likelihood, the centrally controlled vehicle routing would be system optimal to minimize total vehicle-hours or total person-hours. There would also be the potential to price differentially so that those wishing to travel faster would pay a fee for preferential routing.
2. The route performance function would likely change in two ways. First, the free-flow travel time would likely be reduced because vehicles will be able to travel faster and maintain a high level of safety. Second, the effect of increasing traffic demand on route travel time should be substantially less than current assumptions since vehicles will be able to maintain uniform speeds and closer following distance (the spacing between vehicles as traffic demand increases will depend on the efficiency of the technology and not on the reaction times and driving behavior of humans).