Development and Performance Evaluation of AVCSS Deployment Sequences to Advance from Today's Driving Environment to Full Automation

This report presents the findings of its investigation into deployment sequences to better understand the paths that could be taken from today's driving environment to vehicle-highway automation. One of the most vexing problems has always been that of determining how to advance from the present-day manually-controlled vehicles to the future fully automated vehicles. Considerable research attention has been devoted to defining the architecture and operating protocols, as well as the technology, of automated highway systems. Rather less attention has been devoted to defining the steps by which we can get there. Initially, targets of opportunity were identified for accelerating progress toward highway automation, taking account of the operational constraints. Next, after reviewing existing literature on automated highway systems deployment, a set of principles to govern the design of deployment strategies is suggested followed by proposed deployment sequences for automated highway systems, beginning with adaptive cruise control and then adding elements of vehicle-vehicle cooperation and lane protection to build toward automated highway systems capabilities within constraints of technological, human factors and economic feasibility. A general deployment staging sequence is then presented along with example deployment "roadmaps" shown for transit buses, heavy trucks and light-duty passenger vehicles. Finally, we discuss the findings of our modeling and evaluation work for the beginning stages of a specific deployment sequence for light-duty passenger vehicles in the setting of a single highway lane. This sequence incorporates the use of cooperative adaptive cruise control along with conventional or autonomous adaptive cruise control and manual-driven vehicles. The evaluation assesses the impact of each of these three operational driving modes on traffic flow dynamics and highway capacity as well as of increasing proportions of both autonomous and cooperative adaptive cruise control vehicles relative to manually driven vehicles. Such effects are difficult to estimate from field tests on highways because of their necessarily low market penetration of these vehicles. Our approach uses Monte Carlo simulations based on detailed modeling work to estimate the quantitative effects of varying proportions of vehicle control types on lane capacity and on queue lengths and wait times at on-ramps.