Design and fabricate a portable  device that actuates user’s elbow via BLDC motor, a two stage 19:1 planetary gearbox, and cable dual pulley system for assisting stroke survivors perform activities of daily living (ADLs) 

Accepted Publication in: IEEE Int. Conf. on Rehabilitation Robotics (ICORR)

Used Solidworks to design dual pulley actuator, transmission, and idler tensioner system

Used FDM & SLA 3D printers to manufacture transmission, and dual-pulley system, and tensioner idler. 

Used Simulink, MATLAB, Copley Nano motor driver, and dSPACE to design for proportional, derivative (PD) control system with feedback from CUI 2048 count encoders and experiment testing

Minimum No jerk 90-degree trajectory position tracking, 6 trails, 8.28 degrees RMSE (comparable to other devices)

Frequency Response: -3 dB drop at 2 Hz (Within device bandwidth for assistive devices

Weighs less than 5 kg (0.87 kg on the arm)

Employing machine learning to predict patient movements for assistance during daily living activities in real-time for control use for elbow rehabilitation

Used 3 surface electromyography (semg) sensors, and raw inerital measurement unit (IMU) readings for 10 data acquisition logs and testing

Leveraged a convolutional neural network (CNN) machine learning model to develop a 8 layer simple CNN classification algorithm using python and pytorch

Used Arduino Mega 2560 for sensor integration and serial communication

Raspberry Pi 3B+ for algorithm deployment and real-time classification

 100% confusion matrix accuracy

Able to classify 80% of total elbow movements in during water drinking activity of daily living, live.

Could be improved for a faster response

Proof of concept of classifying movements using machine learning and sensor fusion

An RC Robot Mini Car that is fully autonomous that can avoid obstacles through an obstacle course

Used Texas Instruments TM4C12 microcontroller and C code to control robot

Used ultra sonic sensor, a servo for obstacle avoidance with a 90 degree range view range of the robot

Used interrupts and timers for avoidance detection and distance measurement

Used four dc motors for torque vectoring robot during operation

 Successfully able to navigate obstacle without disturbances

Design and assembled a quad rotor drone for ASME IAM3D competition that is required to pick up a 1x1x1 inch payload

Used SolidWorks to reverse engineer off-the-shelf components such as the frame, brushless DC motor, and propellers using company data sheets

Designed and assembled a rack and pinion actuator to pick up payload

Designed a drone with all the necessary parts to function and pick up a payload