During the development of modern cities, there is a strong demand articulated for the sustainability of progress. Since transportation is one of the main contributors to greenhouse gas emissions, the modernization and efficiency of transportation are key issues in the development of livable cities. Increasing the number of lanes does not always provide a solution and often is not feasible for various reasons. In such cases, Intelligent Transportation Systems are applied primarily in urban environments, mostly in the form of Traffic Signal Control. The majority of modern cities already employ adaptive traffic signals, but these largely utilize rule-based algorithms. Due to the stochastic nature of traffic, there arises a demand for cognitive decision-making that enables event-driven characteristics with the assistance of machine learning algorithms. While there are existing solutions utilizing Reinforcement Learning to address the problem, further advancements can be achieved in various areas. This paper presents a solution that not only reduces emissions and enhances network throughput but also ensures universal applicability regardless of network size, owing to individually tailored state representation and rewards.

Efficient traffic flow management on highway scenarios is crucial for ensuring safety and minimizing emissions through the reduction of so-called shockwave effects. In this paper, we propose a novel approach based on cooperative Multi Agent Reinforcement Learning for optimizing traffic flow, utilizing Variable Speed Limit Control in dynamic simulation environments with random anomalies. Our method leverages Reinforcement Learning to adaptively adjust speed limits on distinct road sections in response to alternating traffic conditions, thereby improving not only general traffic flow parameters, but also reducing sustainability measures overall. Through extensive simulations in a Simulation of Urban MObility environment, we demonstrate the superiority of our approach in enhancing traffic flow efficiency and robustness compared to alternative solutions found in literature. Our findings reveal an enhanced performance of RL-based VSL control over traditional approaches due to its generalizability, which contributes to the progression of Intelligent Transportation Systems by presenting a proactive and adaptable resolution for highway traffic management within dynamic real-world contexts.