About Me
I'm Ph.D. student at the University at Buffalo working with Dr. Karthik Dantu at Distributed RObotics and Networked Embedded Sensing (DRONESLab). Currently, I am also teaching the CSE568 course at UB.
My research interests lie at the intersection of motion planning, control and robot learning. I am working on unifying these by leveraging the knowledge from classical techniques as priors for learning based approaches.
You can find my resume here
My research interests lie at the intersection of motion planning, control and robot learning. I am working on unifying these by leveraging the knowledge from classical techniques as priors for learning based approaches.
You can find my resume here
Selected Publications
KFC: Kinematics Only Differential Flatness based Control for F1Tenth Autonomous Racing
2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
In this work, we show the effectiveness of Differential Flatness based control for high-speed trajectory tracking for car-like robots. We demonstrate a 15% increase in trajectory tracking performance compared to MPC while reducing the required compute by more than 50%, both in simulation and on a real 1:10 scale race-car.
SAGAF1T: Surface Adaptive Grip Aware Trajectory Generation for F1Tenth
Submitted to RAL
Identifying a raceline, is a non-trivial problem in motor-sports. Professional racers spend countless hours in simulations and on the real track to identify a path that lets them push the limits of the car to their maximum. Our paper presents a systematic, optimization based approach to attain time-optimal racing trajectories that incorporate vehicle grip for a car-like robot to operate at the limits of handling under diverse conditions. We compare against more commonly used approaches such as minimum curvature. Our raceline generation algorithm outperforms minimum curvature by 20% in simulation and 15% in experiments.
[Paper (Coming Soon)]
[Project Page (Coming Soon)]
Current Research
Differential Flatness based Trajectory Generation for Autonomous Racing [Project Page (Coming Soon)]
Extending visual planning to 3D spaces using Gaussian Splatting and Imitation Learning[Project Page (Coming Soon)]
Excavators, earth-movers, and large construction vehicles have been instrumental in propelling human civilization forward at an unprecedented pace. Recent breakthroughs in computing power, algorithms, and learning architectures have ushered in a new era of autonomy in robotics, now enabling these machines to operate independently. To this end, we introduce EARTH (Excavation Autonomy with Resilient Traversability and Handling), a groundbreaking framework for autonomous excavators and earth-movers. EARTH integrates several novel perception, planning, and hydraulic control components that work synergistically to empower embodied autonomy in these massive machines. This three-year project, funded by MOOG and undertaken in collaboration with the Center for Embodied Autonomy and Robotics (CEAR), represents a significant leap forward in the field of construction robotics.
Education
Jan. 2024 - Present
University at Buffalo
PhD, Computer Science and Engineering
Distributed RObotics and Networked Embedded Sensing (DRONES) Lab
Advised by Dr. Karthik Dantu
Distributed RObotics and Networked Embedded Sensing (DRONES) Lab
Advised by Dr. Karthik Dantu
Aug. 2022 - Dec. 2023
University at Buffalo
M.S. in Robotics
Aug. 2016 - May 2020
Mumbai University
B.E. in Computer Science