Sudhir Yadav

Developed custom MoveIt! Pro behaviors and a differential IK planner plugin, and optimized ROS2-based mapping for marine sonar on robot arms. Built Isaac Sim and Gazebo simulations for deformable manipulation and agricultural environments, and explored OpenVLA-based approaches for task-level robot learning.

Developed a Unity simulation workflow to collect visual and LIDAR data for a desert environment and built a custom 3D-LiDAR sensor in Unity. Supported dataset generation using multiple sensing setups including motion capture and IMU.

Developed vision-based navigation for GPS-denied environments and demonstrated the approach on a quadcopter. Built reinforcement-learning-based algorithms for learning multiple vision-based manipulation tasks on robotic arms.

Developed a ROS interface for the TAL BRABO industrial robot and created a pipeline to demonstrate correct operation by making the arm follow a green-colored ball. Delivered an end-to-end integration flow to validate the robot interface.

Grade: CGPA: 10.0/10.0

Pursuing a Ph.D. in Robotics and Mobility System at IIT Jodhpur since July 2023 (CGPA: 10.0/10.0).

Grade: CGPA: 8.64/10.0

Activities and societies: Learning based control (manipulation, locomotion), 3D vision, and simulation.

Completed a B.Tech. in Electrical Engineering at IIT Jodhpur (2014–2018), with CGPA 8.64/10.0.