Worked in the Robot Locomotion Group advised by Russ Tedrake

Developed algorithms for planning collision-free motions to near global optimality using convex optimization (using Graphs of Convex Sets)

Researched learned visual feedback policies from demonstration for robotic manipulators

Studied algorithmic shortfalls for manipulating deformable objects

Rewrote portion of the motion planning pipeline including a custom Time Optimal Path Parameterization

Developed tools for dissecting failures in the planning pipeline that led to collisions

Massively reduced the rate of errors caused by violation of robots dynamic limits

Worked in the Agile Robotics Lab advised by Scott Kuindersma

Researched control algorithms for bipedal locomotion

Setup communication and low-level control protocol for Cassie, a new bipedal research platform

Developed Zero Moment Point and Capture Point based balancing controllers in simulation and deployed on hardware

Researched formulations for collision-free trajectory optimization

Designed removable data storage for Autonomous Underwater Vehicles (AUVs), simplifying mission turnaround

Analyzed the structural integrity of pressure vessels using FEA and redesigned parts for next generation low cost AUVs to improve machinability & reduce cost

Developed actuator allocation Simulink model for controlling new AUV design that will be used to write the sub’s control code & thoroughly documented/cleaned up existing code

Assembled prototype subs for a trade show demonstration of an AUV capable of launching an AUV swarm