About the Project

Outcomes:

Phantom traffic jams - the ones that seemingly occur without an obvious cause like a bottleneck or incident - can be created by the collective human driving behavior alone. Since automated vehicles can take over some driving tasks such as speed control from humans, they may be able to reduce the occurrence of these jams, if those vehicles are properly designed. This project explored the possibility of automated vehicles to reduce the presence of phantom traffic jams in settings when only as few as 5% of the vehicles are automated, and the rest remain under human control. The project delivered new mathematical models and control algorithms that were demonstrated in theory, computer simulations, and field experiments to eliminate phantom traffic jams. Using a real automated vehicle and more than 20 human drivers, field experiments were conducted that validated the concept that automated vehicles can in fact smooth traffic flow. The main findings of the project are as follows.

• A small fraction of automated vehicles can dramatically reduce the presence of phantom traffic jams. A video of the main experiment is available here: https://youtu.be/2mBjYZTeaTc. A video explaining the research and the experiment is here: https://youtu.be/CKo-v_qwJwo.
• Compared to when the phantom jam was present, when the autonomous vehicle removed the jam, the total fuel consumed by all vehicles in the experiment was reduced by approximately 40%. The finding demonstrates that the benefits of automated vehicles may begin to occur even before all vehicles are automated.
• Based on fuel consumption models applied to the vehicle trajectories it was found that the wave-dampening by the automated vehicle can reduce the vehicles’ emissions of harmful nitrogen oxides by up to to 70%.
• Additional experiments were conducted to assess the jam-absorbing potential of current adaptive cruise control systems on commercial 2018 model year vehicles. The experiments identified that current systems are not able to dampen phantom jams, suggesting more improvements are needed in the design of commercial systems before the potential benefits to traffic are realizable.

Funding:

This material is based upon work supported by the National Science Foundation under Grants No. CNS-1446715 (Piccoli), CNS-1446690 (Seibold), CNS-1446435 (Lysecky and Sprinkle), and CNS-1446702 (Work) through the project, CPS: Synergy: Collaborative Research: Control of Vehicular Traffic Flow via Low-Density Autonomous Vehicles.