Joe Eappen

Papers

Total Citations

H-Index

About

Joe Eappen is a researcher working at the intersection of deep reinforcement learning, robot control, and safety-critical systems. His work addresses one of the most pressing challenges in modern AI-driven robotics: ensuring that learning-based controllers remain reliable, safe, and robust in real-world deployments. Eappen's most notable contribution, "Model-free Neural Lyapunov Control for Safe Robot Navigation" (2022, 10 citations), tackles the fundamental tension between the flexibility of model-free deep reinforcement learning and the safety guarantees required for autonomous navigation. By integrating neural Lyapunov methods with DRL, his work provides a principled framework for certifying safety in controllers that operate under unknown dynamics — a significant step forward for practical robotics applications. His earlier work, "Robustness to Adversarial Attacks in Learning-Enabled Controllers" (2020, 3 citations), examines vulnerabilities in cyber-physical systems, demonstrating how learning-enabled controllers can be compromised through state perturbations and exploring defenses against such threats. Together, these contributions position Eappen as an emerging voice in trustworthy autonomous systems, bridging the gap between the expressive power of deep learning and the rigorous safety demands of real-world robotic and cyber-physical applications.