I’m an undergraduate studying computer science in the College of Engineering at Cornell University. I enjoy math and physics and spend my free time flying drones, taking pictures, and throwing pottery. I’m the lead for the computer science team of the Cornell University Autonomous Drone Project Team (CUAD) which competes in the annual International Aerial Robotics Competition.
I wrote an evolutionary algorithm in Python to optimize the antenna shape for communications antennas, the final design is to be used on F-5 Navy training jets. The plate antenna operates from 108-400MHz, the evolutionary algorithm designed antenna is about twice as good as the old design and exceeds the specification by 25%. Here is the paper I wrote on the process.
Learn more
As part of my work for CUAD doing motion planning, I implemented and benchmarked a few different implementations of A* and related algorithms in Java and C.
Learn more
Inspired by a Ted Talk by Stephen Wilkes, I created a version of his photos in Python using PIL, FFmpeg, multiprocessing and NumPy.
Learn more
I created a visual demonstration of the Fourier transform in the complex plane. The code takes hand drawn paths and constructs a visual representation of the frequency deconstruction.
Learn more
To learn a bit about Python's pandas library, I created some visualizations of historical weather data from OpenWeather.
Learn more
I wanted to create a place for new members to review notes on onboarding topics. I was inspired by the course website for one of my classes and decided to make a similar website that would display markdown files. I coded the website from scratch using a variety of libraries to display the files how I wanted. Hosting the page on GitHub Pages made editing as easy as changing the source files and pushing to the repo.
Check it out here
Over the summer, I worked in VCU's OpenCyberCity Testbed on autonomous drone control in a 1:12 scale model. I worked on installing and setting up a Vicon motion capture system that used IR cameras to get high accuracy position information for tracking markers that were attached to small drone called a Crazyflie. We used the Vero cameras and Tracker 3. Additionally, I helped to network the computers and feed the Vicon tracking data into the drone for use in pose estimations to aid in stable flight. I also wrote code in Python to allows for easy high level control of the drone using waypoints. Finally, I build the infrastructure to enable testing of various traffic management systems. One of the main goals of the lab is to test proposals for a unmanned aircraft traffic management (UTM) system using scale models. I wrote code to allows Linux machines controlling the drones to communicate with each other and a central traffic management system. This simulates how real-world drones will decide how to avoid obstacles and other drones ans will allow future work testing various implementations for the UTM.
I worked under Misha Padidar in Cornell's Applied Math Department on optimizing the magnetic confinement coils’ geometry for stellarator fusion reactors. The process involved finding optimal magnetic surfaces to confine the plasma and the coil shapes that make those fields. I’ve built optimizers in Python using gradient descent, Gauss-Newton, linear and logistic regression, and LaGrange multipliers. Additionally, we are worked on code to compute Green's residue criterion and its derivative with respect to coil parameters using the simsopt framework.
As part of the tribology lab in the University of Nevada, Reno’s Mechanical Engineering department, I helped automate data-analysis and the production of ionic liquids using Python. I wrote code that processed wear-track depth data to determine characteristics like wear volume and surface roughness. Additionally, I helped in the effort to use machine learning to extrapolate data beyond the capabilities of the testing equipment. Finally, I helped design and develop an Arduino controlled synthesizer of ionic liquid. I coded an image processing algorithm to determine the layer boundary between two liquids, this helped automate the refinement of ionic liquid.