Decentralized Traffic Flow Optimization Through Intrinsic Motivation

Traffic congestion poses a significant challenge, with its impact increasingly felt on the global economy. Among the various solutions for traffic management, centralized approaches [1]-[3] are typically restricted to areas with high-quality infrastructure, enabling near real-time connectivity to central servers and neighboring vehicles. This is a limiting assumption, in terms of location and of the number of cars.