Anurag Shetye
I am a robotics researcher working on perception and control for decision making under uncertainty, the parts a robot needs before it can be trusted to act without a person standing behind it. I currently work with Dr. Arpita Sinha at IIT Bombay's Centre for Systems and Control on autonomous manipulation, with Dr. Johannes Köhler in Imperial College London's Autonomous Systems Group on model predictive control for navigation among moving obstacles, and with Dr. Mark Vlutters in Twente's Nakama Robotics Lab on vision for learning from demonstration.
I got my start on real flight hardware in the Robotics and Drones Laboratory at RAIT, DY Patil University, where I completed my B.Tech in Computer Science & Business Systems in June 2026 with a CGPA of 9.48/10. I turned down research offers from ISRO (URSC) and BARC to work full-time at IIT Bombay.
Check out my research. Check out the resources I built along the way.
Research
Extending an MPC framework for cluttered environments to obstacles that move, where the free space a plan was computed against no longer exists by the time the robot gets there.
Autonomous Systems Group, Imperial College London
DetailsA whole-body QP controller that corrects joint velocities from measured base pose, so the arm keeps reaching while the base is still moving.
Centre for Systems and Control, IIT Bombay
DetailsGrasp and contact events from the robot's own proprioception seed segmentation, so the task-relevant objects in a demonstration are labelled without a human.
Nakama Robotics Lab, University of Twente
DetailsAlgebraic inverse kinematics, stochastic planning, reactive obstacle deflection and admittance control, joined into one continuous pipeline.
Centre for Systems and Control, IIT Bombay
DetailsLiDAR altitude estimation, RTAB-Map point-cloud mapping and precision landing for Martian-like terrain. Final Field Round, top 16 of 500+ teams, IRoC-U 2025.
ISRO Robotics Challenge, URSC
DetailsA sequence model for the forward map and a mixture density model for the inverse, learned from motion capture on a soft arm that has no closed-form kinematics.
Centre for Systems and Control, IIT Bombay
DetailsStarted at fourteen after the Nepal earthquake. Ground-penetrating radar, vision, gas sensing and LoRa on a paired aerial and ground platform. Bronze at CASTIC; Indian patent filed.
Independent
DetailsPID stabilization, precision landing, and leader–follower consensus with Lighthouse positioning, flown on real hardware.
Robotics & Drones Laboratory, RAIT
DetailsSetpoint-based autonomous control for a quadruped platform, moving it from teleoperation to commanded waypoints.
Robotics & Drones Laboratory, RAIT
DetailsFollowing natural-language route instructions through a city-scale point cloud, where the map is dense but the instruction is the only goal specification.
Centre for Systems and Control, IIT Bombay
DetailsCargo placement and rearrangement for a space station, where stowage volume is fixed and an item buried behind three others has effectively been lost.
National Space Hackathon, IIT Delhi × ISRO
Details