Abstract We present a reproducible benchmark for evaluating sim-to-real transfer of Multi-Agent Reinforcement Learning (MARL) policies for Connected and Automated Vehicles (CAVs). The platform, based on the Cyber-Physical Mobility Lab (CPM Lab) (M. Kloock et al ., “Cyber-physical mobility lab: An open-source platform for networked and autonomous vehicles,” in European Control Conference (ECC) , 2021, pp. 1937–1944), integrates simulation, a high-fidelity digital twin, and a physical testbed, enabling structured zero-shot evaluation of MARL motion-planning policies. We demonstrate its use by deploying a SigmaRL-trained policy (J. Xu, P. Hu, and B. Alrifaee, “SigmaRL: A sample-efficient and generalizable multi-agent reinforcement learning framework for motion planning,” in IEEE International Conference on Intelligent Transportation Systems (ITSC) , 2024, pp. 768–775) across all three domains, revealing two complementary sources of performance degradation: architectural differences between simulation and hardware control stacks, and the sim-to-real gap induced by increasing environmental realism. The open-source setup enables systematic analysis of sim-to-real challenges in MARL under realistic, reproducible conditions.
Beerwerth et al. (Fri,) studied this question.