Projects

Finley: 3D Printed Humanoid

Under Pradeep Radhakrishnan, PhD

This project aimed to aid assistive care workers in performing simple at-home tasks. The Finley robot is now capable of pick-and-place and human-robot interaction. Finley also features a modular wrist connection and human interaction utilizing artificial intelligence. Finley showcases a versatile two finger gripper with one rigid and one flexible finger. A human study was also performed to evaluate the need for humanoid robots. The team redesigned Finley to reduce the number of 3D printed parts and fasteners which reduced assembly time and improved customizability for Finley.

• Senior Design Project leveraging the open-source Poppy Humanoid

• Designed and assembled a humanoid, printed entirely from resin and PLA, for at-home medical assistive care

• Introduced pick and place system, sensor integration, human-robot interaction, and end effector modularity

• Integrated AI model for human-robot conversation, function invocation, and speech to motion capabilities

Mass General Brigham Hospital Website

Under Wilson Wong, PhD

• Served as Scrum master on an 11-person team to apply Agile development methodologies and software design patterns in the creation of a full-stack web application for hospital pathfinding and integrated service requests

• Developed using React.js, Typescript, Shadcn/ui, Express, Prisma ORM, and PostgreSQL

• Deployment on the cloud was accomplished using AWS EC, a Docker container, and AWS RDS to run the PostgreSQL server

TurtleBot SLAM

Under Greg Lewin, PhD

Explored navigation, including mapping, localization, and path planning, using a TurtleBot3. Mobility was provided through smart motors, configured for differential drive; distance sensing is provided by a LiDAR; communication and control were provided by a Raspberry Pi. Navigation, position estimation, and mapping. Concepts of forward and inverse kinematics, dead reckoning, and noise filtering were implemented through Robot Operating System (ROS), Ubuntu 20.04, and Python.

• Implemented differential drive kinematics, probabilistic state estimation (Kalman and Particle Filters), and optimal path planning for safe autonomous navigation

• Utilized ROS, C++, Ubuntu, and Gazebo for communication, simulation, and physical navigation

3DOF Robotic Arm

Under Siavash Farzan, PhD

Coordinated motion of multiple actuators to execute complex manipulation tasks in the physical space. Studied transformations along with position and velocity kinematics, and fundamental concepts of robot dynamics and control, trajectory planning, and computer vision. Applied theoretical methods learned in the classroom during practical laboratory sessions, which culminated in the construction and programming of a vision-guided, multi degree-of-freedom robotic manipulator. MATLAB and Python was used to introduce the necessary concepts for robot programming.

• Performed forward kinematics, inverse kinematics, and trajectory planning for a 3 DOF RRR manipulator

• Explored robotic vision, camera calibration, and feedback control in MATLAB to identify and move objects

Solar Panel Delivery Robot

Introduced effective conversion of electrical power to mechanical power, and power transmission for purposes of locomotion, and of payload manipulation and delivery. Concepts of energy, power and kinematics were applied. Concepts from statics such as force, moments and friction were used to determine power system requirements and structural requirements. Simple dynamics relating to inertia and the equations of motion of rigid bodies were considered. Power control and modulation methods through software control of existing embedded processors and power electronics.

• Constructed, programmed, and tested the operation of a mobile robotic system capable of locomotion and payload manipulation for model solar panel delivery

• Designed 4-bar lifting mechanism, performed power control and modulation of system using embedded C/C++