100 Self-Driving Cars Tackle Rush Hour Congestion

Researchers deployed 100 reinforcement learning (RL)-trained autonomous vehicles (AVs) onto a highway during rush hour to mitigate congestion and reduce fuel consumption. These AVs learned to smooth traffic flow, maximizing energy efficiency while maintaining throughput and safe operation around human drivers. The experiment demonstrated that even a small percentage of well-controlled AVs significantly improves traffic flow and fuel efficiency for all road users. This large-scale experiment provides valuable insights into deploying AVs to improve traffic conditions.