Projects I have worked on
Technologies: Python (OpenCV2, Tensorflow/Keras, numpy/scikit, pybluez), Java (JavaFX), C/C++ (libbluetooth, pthread, socket), Raspberry Pi, Microsoft Kinect 1 & 2
January 2016 - December 2017
Received Warren-Batts VIP Innovation Award for significant contribution towards the team upon nomination by my research advisor.
I was a member of the EnerCage VIP Research Project at Georgia Tech under the Image Processing and Microcontrollers subteams.
Under the Image Processing subteam, I worked on training Deep Learning Convolutional Neural Network models to accurately predict and classify a lab rat’s behavior from a color video and depth top-down feed. This is done in combination with various preprocessing steps to represent the data in different formats before feeding them into the training models.
Under the Microcontroller subteam, I worked with the Raspberry Pi to establish a hub between a rat-mounted microchip and a researcher’s computer using Bluetooth and TCP connections respectively. I also helped design the GUI for interfacing with the Raspberry Pi Hub that would be used by the researchers on their desktops as well as implement the functionality to connect and log data from multiple hubs simultaneously.
Technologies: C/C++ (Arduino, Raspberry Pi, CANBUS, libbluetooth), Android (Bluetooth)
August 2015 - June 2016
Ranked 6th in the competition from 40 universities
I was a member of the Georgia Tech HyTech Racing team where we design a single-seat Electric Vehicle to compete against teams from up to 40 other universities at the Formula Hybrid competition and FSAE Electric.
I was responsible for linking the CANBUS network of the car and programming the various microcontrollers for functions ranging from pedal/brake control to controlling and monitoring the various EV units such as the Motor Controller, Accumulator, and Low Voltage Boards. I also installed a steering wheel-mounted smartphone that displayed gyroscope stabilized car diagnostics data such as speed, temperatures, charge, etc. via a Bluetooth connection to the car.
Technologies: Python (Django, PostgreSQL), HTML/CSS, Heroku, Raspberry Pi, Analog Design (200A CT Sensors, MCP 3008)
January 2015 - December 2017
I served as a member of an independent, multi-disciplinary team of students who hosted our first energy savings competition in one of the on-campus freshmen housing buildings. Our project won funding from the Greenovation 2017 Fund at Georgia Tech. We designed our own monitoring system to collect live energy usage data from each of the floors in the building and log it to a remote database. We then provided statistics updated daily on the energy usage and comparative reduction for each of the floors through the competition website with the intent of fostering constructive competition and to raise awareness with regards to energy conservation.
Technologies: Android (Picasso, SQLite3, GraphView), Adobe Illustrator
July 2015 - August 2015
Created and published an elegant and resourceful Rubik’s Cube app for Android. The app has since been download 5,700+ times and received an average rating of 4.65/5 across 133 reviews. Features of the app include:
Technologies: Python (Google App Engine, Google Calendar API), HTML/CSS
December 2013 - August 2014
Created a web app aimed at faculty and students at my highschool (ISKL) to allow for easy scheduling/unscheduling of recurring events based on our school’s custom calendar which follows an 8-day cycle. Has been used by over 50 different people to schedule 23,500+ events.
Received 1st place in the competition from over 20 teams
My team of five and I participated in the Google Games Competition in Atlanta which involved various programming, analytical, and puzzle solving challenges.
Ranked 3rd in the competition from over 100 participants
I participated in the Freshmen Hackathon which tested problem solving skills using programming.
Technologies: Android (Bluetooth, Firebase), Myo Gesture Control Armband Android SDK
For our HackGT 2015 project, my team of four designed an Android App that synced with an arm-worn smart band with 3-axis accelerometer sensors. This app would be installed and two or more individuals devices. Based on the accelerometer data and time, it would recognize when two individuals were shaking hands and would automatically exchange their contact information (with prior consent) via a Firebase database.
Technologies: C, GBA Emulator
For my CS 2110 class, I designed the Bomberman game to be run on the GBA emulator. This involved various low-level concepts such as direct memory management, display buffers, bit-hacking, timing simulation, and direct game-state tracking through various C structs.