Raphael Trumpp

Papers

Total Citations

H-Index

About

Raphael Trumpp is a researcher at the forefront of autonomous systems and reinforcement learning, with a focus on safe, multi-agent decision-making under constraints. His work bridges the gap between theoretical control and practical deployment, particularly in high-stakes environments like autonomous racing. Trumpp’s most cited paper, “RaceMOP: Mapless Online Path Planning for Multi-Agent Autonomous Racing using Residual Policy Learning” (2024, 5 citations), introduces a novel approach to interactive decision-making that enables vehicles to safely overtake opponents without a pre-built map—a critical challenge for real-world autonomy. This work’s insights extend beyond racing to general multi-agent systems. His second notable paper, “Learning to Generate All Feasible Actions” (2024, 2 citations), tackles the problem of enforcing hard safety and operational constraints in reinforcement learning for complex cyber-physical systems. By ensuring agents only explore viable actions, Trumpp advances the reliability of data-driven control. Though early in his career, his contributions are already shaping safer, more efficient autonomous navigation, with potential impacts on self-driving cars, robotics, and industrial automation.