Evaluating Infrastructure-Less Cooperative Parking Guidance via Agent-Based Simulation

Efficient parking management in urban environments remains a critical challenge, exacerbating traffic congestion and emissions. This paper presents an agent-based simulation study, conducted in SUMO, to evaluate the feasibility of intelligent vehicles cooperating to assist drivers in locating available parking spots without relying on external infrastructure. Agents infer parking availability by observing vehicle locking and unlocking events, dynamically updating a shared real-time heatmap of potential parking areas. Through comprehensive simulations, we explore the minimal adoption threshold required for the system to deliver tangible benefits in terms of reduced search times and improved urban mobility. Results indicate that even moderate participation significantly enhances parking efficiency, demonstrating the practicality and effectiveness of distributed, infrastructure-free agent cooperation for smart urban mobility solutions.