My Projects
Deep Reinforcement Learning (DRL) on Legged Robots
Northeastern University Silicon Synapse Lab Graduate Research
As the RL team lead at the lab, I work with and guide students interested in the field by defining projects and maintaining timelines. Together, we work on developing policies for our multimodal legged robots that are founded in controls (Linear Inverted Pendulum (LIP), Model Predictive Control (MPC), Raibert Heuristic) as well as whole-body control policies for off-the-shelf robots like the Unitree Go2, G1, and Boston Dynamics Spot with arm.
In-hand Manipulation of a Cube
Assistive Robotics Course Project
My course project for an assistive robotics class where we developed a force controller and RL controller to manipulate the pose of a cube using a 4-fingered dexterous hand.
Centroidal Model Predictive Control on Quadruped Robot
Legged Robotics Course Project
My course project where I developed a swing/stance controller using the Raibert Heuristic and MPC on a quadruped robot. The robot was capable of tracking the swing leg targets and apply correct balancing forces to keep the base stable during locomotion.
Autonomous Navigation and Target Tracking on a Mobile Robot
UMass Amherst Senior Capstone Project
For my capstone project, I was the sensing and navigation lead. I developed the autonomous navigation system and sensor fusion for odometry and localization for our mobile robot. Along with my peers, we were able to develop a fully functional mobile robot that avoided dynamic obstacles while tracking ArUco Markers.
