Gaurav Chaudhary

Gaurav Chaudhary is a Reinforcement Learning Scientist at Neura Robotics, Germany in the PLA Team, where he develops RL policies for robotic manipulation. His work lies at the intersection of reinforcement learning, embodied AI, and robotics, with the goal of enabling robots to learn efficiently and operate reliably in complex real-world environments.

Previously, he was a postdoctoral researcher at Nanyang Technological University, Singapore in the Algorithmic Robotics Group, working with Prof. Yooonchang Sung, where we explored the research direction related to the World-Action Model and Uncertainty estimation in offline Reinforcement Learning.

He received his PhD in Electrical Engineering from the Indian Institute of Technology (IIT) Kanpur under the supervision of Prof. Laximdar Behera, where his research focused on sample-efficient reinforcement learning under sparse rewards and partial observability. His work develops learning frameworks that combine structured generalization, autonomous exploration, and perception-aware control to improve the robustness and scalability of embodied agents.

His research spans the full simulation-to-real pipeline, including algorithm design, large-scale simulation experiments, and deployment on real robotic systems such as Flexiv Rizon-4 and UR10 manipulators. His work has been published in leading machine learning and robotics venues, including TMLR, AAMAS, ICASSP, and IEEE IRC.

Research interests

• Reinforcement Learning for Robotics
• Diffusion Policies and Test-Time Adaptation
• Embodied AI
• Robotic Manipulation
• Sim-to-Real Transfer
• Active Perception and Visual Policy Learning