Projects and Leadership

In this project, fellow students and I investigate and evaluate optimal control methods in the cartpole environment. We focus on this system since it has nonlinear dynamics and is noisy, like real world systems. We find that since all methods successfully stabilize the system, it is sufficient to just use simpler ones. Additionally, it is essential to design the costs well since these methods are sensitive to the cost matrices.

The objective of this project is to apply reinforcement learning algorithms to the puzzle video game "Baba Is You". I build off of work already done that emulates this game. Reward tampering in reinforcement learning refers to when the agent can actively change the reward function of the environment. This is an interesting problem in reinforcement learning and has applications in safety.

Sunstone is an internal research and development project at Raytheon that focuses on providing navigational information in GPS-denied environments. I was responsible for the software organization and infrastructure of the project. I also used computer vision algorithms and generated point clouds.

RIGEL uses computer vision with high resolution lunar images to identify geological features and help lunar astronauts better understand the environment. I proposed and worked on this as a capstone with a fellow student.

Inspired by Iron Man's JARVIS, VEGA is leverages conversational AI methods to provide the user with hands-free access to any necessary information. I proposed and worked on this project as a capstone with a fellow student.

ATLAS is an AR system for the Microsoft Hololens, designed to be a tool for astronauts on lunar expeditions. During an EVA, they are given navigation assistance, access to vital information, and tools to aid in sample collection.

A group of space enthusiasts and I submitted to the Mars City State Design Challenge. We named our city state Adrestia, and considered topics such as economy, government, society, and technology.

In this project, a fellow student and I investigate methods to predict whether or not society finds a member to be in the wrong and what factors contribute to this verdict. The results of this project will help better understand societal norms.

The project's goal is to predict statistics about the area by analyzing aerial images of cities. A team of computer vision students and I accomplish this by analyzing features such as lighting, land usage, streets, and other relevant features.

Voiced GIMP is a conversational AI project that integrates with the GIMP program to streamline image editing. Users can talk or type image processing commands without having to navigate complicated menus. This project was my MDE (major design experience) and done with a group of fellow students.

I interned with the AI/ML research group in Dulles and worked on R&D projects. For my primary task, I was the technical lead of a project that aimed to find the least cost path through a point cloud. My work resulted in two IP (intellectual property) disclosures, the combination of which is under patent review

The goal of MEISSA is to use modern augmented reality technology to improve how astronauts communicate with their tasks, ground support crew, and biometric data during a spacewalk. This system was tested with EVA specialists at the Johnson Space Center in Houston, TX.

TBEx is a pair of satellites funded by NASA that launched to study the variability of the upper atmosphere. I worked on software for the satellite that is responsible for sending packets of data from the ground to the satellite. I also helped others with software for commanding the satellite.

I worked with the CLAWS team and faculty advisor to develop a LEGO outreach activity that demonstrates to students the difficulties of extra-vehicular activities in space. We organized this activity initially at Detroit Public Schools.

GPS OCX is the next generation GPS operation control system. This improved system will increase accuracy and have capabilities to fly twice as many satellites. Additionally, this system will provide coverage to hard to reach areas. As an intern, I was responsible for the infrastructure on the software side.

The mission objective of the MARIO satellite is to characterize the performance of piezoelectric actuators and health monitoring systems in low Earth orbit conditions.

LITE, or LED Initial Test, is a prototype mission to test LED visibility and tracking capabilities. The mission consisted of a payload with two bright red LED panels and accompanying sensors. This payload flew on a high altitude balloon. I worked on the on-board multi-threaded software suite for this mission.

Project Strato is the high-altitude balloon team in MXL. The team pursues new technologies to improve flight and serves as a testbed for MXL's satellites. My tasks involved calibrating sensors, end to end testing, and writing code.