I am currently working as a software engineer at Optimus Ride in the Planning and Controls team where I focus development efforts on modern control techniques that provide rigorous safety guarantees for self-driving vehicles.
Before joining the industry, I completed my PhD from the University of Illinois, where I was advised by Prof. Naira Hovakimyan. Broadly, my research was at the intersection of robotics, control theory, and machine learning, where I focused on safe motion planning for robots under uncertainty. For a deeper overview of my work during the PhD, head over to the research page. While at UIUC, I also spent a few summers working outside of academia towards robotics related problems at Facebook Reality Labs, Occipital, and Qualcomm.
Ph.D., Mechanical Engineering, University of Illinois Urbana-Champaign. 2021.
M.S., Aerospace Engineering, University of Illinois Urbana-Champaign. 2016.
B.Tech., Mechanical Engineering, VIT University, India. 2014.
Produced a design document rigorously detailing the safety requirements and the associated assumptions for an autonomous vehicle from a controls perspective.
Documented the design of a model predictive control algorithm with computational, stability, and feasibility guarantees.
Performed system identification for a controlled hardware device.
Implemented a disturbance observer-based control augmented with a baseline PID control to compensate for disturbances injected into the system while accurately tracking reference signals.
Designed a C++ motion planning library and implemented different types of planning algorithms for wheeled robots mounted with Structure sensors.
Implemented a computationally efficient distance transform of an occupancy map for fast collision checking and distance-based prioritization when planning.
Designed an obstacle avoidance controller for the Snapdragon Flight board (since discontinued) for assistive collision prevention using noisy vision-based range information.
Developed sampling-based motion planning algorithms to generate distance-optimal collision-free paths for the vehicle from a 3D occupancy map.