Managing embedded linux and ROS2 infrastructure teams for development of next-gen iRobot product line.

Technical lead of team improving ROS2 infrastructure, architecture, and testing at iRobot.

Core contributor to the ros-planning/navigation2 repository for the Robot Operation System 2.0 (ROS2) framework, aimed at the port, re-design, and enhancement of the ROS navigation stack for the mobile navigation of autonomous mobile robots. Continuous contributions in design and architecture, with key focuses on mapping and ROS2 parameter support

- Developed online Bayesian learning algorithms for control of complex robotic systems- Publication accepted in IEEE Robotics and Automation Letters (RA-L) journal for paper Sparse, Online, Locally Adaptive Regression using Gaussian Processes for Bayesian Robot Model Learning and Control

• Translated test design language into executable simulation scenarios in car simulation environment • Designed system-level requirements to address autonomous vehicle performance and safety core functionality• Assisted in bringing up hardware-in-the-loop for communication between simulated virtual sensors and actuators• Proposed C++ system architecture for autonomous vehicle trajectory and behavioral planning

Received Peter Griffith Prize for Outstanding Experimental Project from MIT MecheE Department• Bachelor’s Thesis project: Analysis of Human Arm Impedance and Control for Constrained Motions• Design and run experiments on human subjects to measure arm impedance and control using robotic arm• Process and analyze subject data in order to test hypotheses of human motor control under constraints

Intern with the Bioengineering Systems and Technologies Group:• Coordinated data collection and analysis for Biomechanics Lab using three different types of motion capture systems: passive marker-based, marker-less, and inertial motion capture suit• Supported US Military sponsors’ projects with goals of high volume data analysis and targeted reports• Developed application platform for quick data processing (ambiguous to motion capture system) and intuitive simulation and graphical visualization tools

• Capture and analyze motion and force data of human subjects on lab treadmill to assess the performance of exoskeleton devices

• • Supported and extended a doctoral candidate’s thesis of full body motion simulation by creating a successful process for arm motion simulation• Created a process that models, analyzes, and improves (control design) the simulation for the motion of a multi-body system through the application of the Port-Hamiltonian approach to systems modeling• Generated real motion trajectories from the motion capture of a human subject via lab cameras to be used as a reference motion for the simulation

Project: Small-scale Impedance-controlled robot • Led in designing and performing experiments to study human motor behavior in catching a ball• Measured EMG signals from joint muscles (elbow, forearm, wrist) and used pressure sensors and goniometer to acquire data about muscle activation reaction time, angle deflection over time, and mechanical parameters of arm. • Created the machine and control design of a robot to perform as prescribed by simulations of the dynamic model of the arm in Matlab