Pranay Dugar

Papers

Total Citations

H-Index

About

Pranay Dugar is a robotics researcher advancing the frontier of humanoid locomotion through reinforcement learning. His primary research areas center on sim-to-real transfer, robust control, and reward function design for bipedal robots. Dugar’s most impactful work, “Revisiting Reward Design and Evaluation for Robust Humanoid Standing and Walking” (2024, 18 citations), systematically investigates how different reward structures affect a humanoid’s ability to reject natural disturbances during standing and walking. This contribution is critical because it moves beyond ad-hoc reward tuning toward principled evaluation, enabling more reliable and transferable locomotion policies. By dissecting the nuances of reward shaping, Dugar’s research helps bridge the gap between simulation-trained controllers and real-world robustness—a key bottleneck in deploying humanoid robots outside the lab. His work has already garnered attention from the sim-to-real community, and its emphasis on rigorous benchmarking sets a new standard for future locomotion studies. For students and researchers entering legged robotics, Dugar’s findings offer both a practical guide and a cautionary tale: reward design is not just an engineering detail but a core determinant of robot resilience.