BIOGRAPHY

Welcome to my website!

Here is my Resume!

I am Sahil Chaudhary, a Master of Science in Mechanical Engineering graduate from Carnegie Mellon University, Pittsburgh. My specialization lies in the vast and exciting field of Robotics. In particular, I am interested in all things at the intersection of Autonomous Decision-Making, Planning and Control, particularly for autonomous systems and field robotics. I invite you to explore my work and experiences that have shaped and defined my passion for the field.

My Skillset

• Programming Languages: C++, Python, MATLAB, Julia
• Tools and Software: Isaac Lab, Gazebo, Matplotlib, PyTorch, NumPy, SKLearn, SolidWorks, Ansys, Fusion 360
• Platforms: ROS, Git, Linux, Docker
• Knowledge Areas: Planning, Reinforcement Learning, Controls, SLAM, Robot Dynamics, Machine Learning, CAD, Mechanical Design, Rapid Prototyping

My Work and Projects

• As a Graduate Research Assistant in the Biorobotics Lab at CMU, I am working on developing algorithms and features for Multi-Agent Planning and Controls for a heterogenous system of robots, along with Comms-Aware Planning.

  • Spearheaded the development of a MANET framework using DDS and ROS to ensure communication fidelity in heterogeneous robot convoys, implement network topology repair and recovery behaviors, and enforce a communication boundary.
  • Comms-Aware Planning: Designed a novel algorithm attaining 99% success rate to maintain communication fidelity over radio among robots in convoy, by formulating a modified Max-Min Spanning Tree, and validated the algorithm through extensive hardware testing.
  • Heterogeneous Convoy Framework: Developed a decentralized convoy framework integrating RC cars and quadrupeds like Boston Dynamics’ Spot, and an algorithm enabling rendezvous at intersections and coordinated return as a convoy, as part of exploratory features.
  • Enhanced the operational efficiency of the Local Planner by up to 29%, through waypoint optimization and trajectory smoothing, reducing unnecessary deceleration between waypoints and improving overall robot speed and motion continuity.
  • Payload Redesign: Engineered a modular, serviceable payload for RC cars and quadruped robots, accomplishing a 7% weight reduction and lowering the center of gravity while ensuring optimal sensor field-of-view and accessibility, and incorporating sensors such as LiDAR, IMU, cameras, onboard computer, motor controller, and circuit boards.

• My prior work and professional experiences include:

  • Worked as a Course Assistant for the Machine Learning and Artificial Intelligence for Engineers (24787) course in Carnegie Mellon University’s College of Engineering, in which I assisted Professor L. Burak Kara in teaching core concepts to graduate students.
  • Worked at ArcelorMittal Nippon Steel (AM/NS) India Limited, as a Graduate Engineer Trainee.

• My coursework and projects cover a wide variety of domains, contributing to my development as a holistic Robotics Engineer. Here are some of the domains in which my knowledge and expertise lies:

  • Controls: PID, LQR, MPC, and MPPI controllers, and have also worked on non-linear traj-opt methods like iLQR.
  • Planning: A* and its variants, D* Lite, PRM, RRT and its variants, and have hence developed a keen understanding of these algorithms. Other domains include:
  • SLAM: State estimation methods like EKF, and LiDAR-IMU time syncing.
  • Machine Learning: Traditional supervised and unsupervised learning techniques, and implementation using PyTorch and SKLearn libraries.
  • Reinforcement Learning: General RL algorithms. In terms of experience, I have trained a PPO-agent to play a game of Pinball!, and also developed and trained policies for last-mile manipulation tasks in Isaac Lab.
  • Hardware: Worked on several hardware projects over the ears, including designing payloads for robots. I am a MechE afterall!

Thank you for taking the time to go through my portfolio. I look forward to any opportunities that come my way!