Eduardo Candela

Papers

Total Citations

H-Index

About

Eduardo Candela is a leading researcher at the intersection of autonomous driving and multi-agent reinforcement learning (MARL), with a particular focus on bridging the critical gap between simulation and real-world deployment. His most-cited work, "Transferring Multi-Agent Reinforcement Learning Policies for Autonomous Driving using Sim-to-Real" (2022, 38 citations), addresses one of the field's most stubborn challenges: achieving robust coordination among multiple autonomous agents. Candela's key contribution lies in developing transfer learning frameworks that allow MARL policies trained in simulated environments to maintain their effectiveness when deployed on physical vehicles, tackling the notorious "reality gap" that often plagues robotics and autonomous systems. His research demonstrates that complex multi-agent behaviors—such as cooperative merging, intersection negotiation, and platooning—can be successfully learned in simulation and then transferred to real-world platforms with minimal performance degradation. By systematically addressing domain adaptation and policy robustness, Candela has provided a practical pathway for deploying learned coordination strategies in safety-critical autonomous driving applications. His work has been instrumental in advancing the feasibility of multi-agent autonomous systems, earning recognition from both the robotics and reinforcement learning communities for its rigorous experimental methodology and real-world relevance.