Reinforcement Learning AI Shows Promise as a Tool for Adjusting Congestion Charges

Researchers have compiled a review on how artificial intelligence utilizing reinforcement learning can help adjust road and congestion charges to improve traffic flow. The background is the growing demand for mobility: as traffic volumes increase, the 'free' pricing of road use easily leads to traffic congestion, and jams weaken traffic flow. The idea of road pricing is to influence driver behavior by charging for road use. Since the traffic situation is constantly changing, the charges should also be able to adapt according to the study: dynamic pricing means that charges are adjusted based on the time and traffic situation. The review focuses on reinforcement learning, a method where the system learns to make better decisions based on feedback. The basic idea resembles trial and error: the algorithm searches for ways to improve the objective, such as traffic flow. Reinforcement learning has already been used, for example, in traffic light control, and is now increasingly applied to dynamic road pricing as well. The authors reviewed recent solutions and compared how different studies have attempted to address the typical challenges of reinforcement learning. The summary is positive: the presented methods demonstrate the usefulness of reinforcement learning in road pricing and traffic optimization. At the same time, the review notes that some key challenges remain relatively unexplored. Source: Reinforcement learning for road pricing: a review and future directions, Artificial Intelligence Review.