Hey! I'm Nithin, a junior studying Robotics at the University of Michigan. At the moment, I'm deeply interested in robot learning and its applications in self-driving and humanoid robotics.
Will be working on robot navigation and localization for autonomous docking protocols.
Building software for robotic arms that allows for fully autonomous wire harness installation in airplane frames.
Securing funding and sponorships, and coordinating tech talks and design reviews to enrich ARV members.
Researching the conversion of linear motion to legged motion.
Designed computer vision software for ARV's perception stack.
A robotic grasping simulator built in Python using PyBullet. The project trains a Kuka robotic arm with Proximal Policy Optimization (PPO) to pick up objects in a physics-based environment. It includes a custom reinforcement-learning setup, episodic training with reward shaping, and a playback script to visualize the learned grasping behavior.
View on GitHub
A lightweight autonomous driving simulator built in Python using Pygame. The project implements grid-based path planning and a pure-pursuit controller to steer a kinematic car model toward a target while avoiding static and dynamic obstacles. Includes support for manual driving, real-time replanning, and randomized environments.
View on GitHub
A robotic locomotion project built on Michigan Robotics’ MBot platform. The system uses custom-designed, 3D-printed legs mounted to the chassis, with joint angles controlled via MicroPython running on a Raspberry Pi Pico.