Portfolio

Automated 3-Phase 480V Test Equipment


Led development, localization, and integration of custom automated test systems for active harmonic filters used in data center applications. Collaborated closely with an international supplier in India to design and build four automated testers, then traveled on-site to oversee development, ensure proper localization for the US market, and resolve critical design issues. Successfully managed the deployment process, including the NRTL field evaluation in the US to achieve full regulatory compliance after identifying multiple compliance gaps upon arrival. Assisted in the development of the test software and performed Gage R&R studies to validate measurement system accuracy and repeatability. The automated testers perform a complete production validation sequence:

  • HiPot and ground bond safety testing to meet UL specifications
  • Controlled ramp-up to 480V using a computer-controlled adjustable AC source
  • Switch-matrix routing of current transformers (CTs) for open loop and closed loop active harmonic distortion correction
  • Evaluation of unit performance using Yokogawa power meters
  • Extended burn-in testing to detect early-life failures

Key Result: Reduced product cycle time by 33% while significantly improving test consistency, safety verification, and early failure detection.

Unconventional Data Transmission


This project explores several different forms of unconventional data transmission. While most electronic data is relayed using a wired connection, WiFi, or Bluetooth, this demonstration displays unique ways that data can flow from one computer to another. This includes an array of solenoids tapping out a full byte in binary, a photodetector decoding flashes from an LED, a rotary encoder detecting and translating the rotation of a motor, sound waves transmitted through a can and string, and even an interactive capacitive touch model. This project was exhibited at Open Sauce 2025 in San Francisco and allows guests to type messages and draw images on one laptop, and transmit it to another through the unconventional method of their choice.

This project was created using LabView, the Arduino platform, 3D printing, and several custom PCBs.

Smart Torque Drivers


During my employment at Schneider Electric, I implemented two different systems of intelligent torque drivers (Cleco and Mountz). The torque drivers monitor rotation angle and converted torque to ensure every fastener is secured correctly. This system interlocks with our site's manufacturing execution system (MES) which enables us to collect detailed data for every product produced. The smart torque drivers communicate wirelessly to their controller which updates MODBUS registers on the line's PLC.

This project utilized LabView, Schneider Electric PLCs, MODBUS, and programming of our MES.

Autonomous Robotics Platform


My master's research required development of a reliable robotics platform for data collection and processing. We built upon the modular platform developed by the US Army Combat Capabilities Development Command (DEVCOM), called the GVR bot. Using custom 3D printed fixtures, I integrated a VLP-16 3D LIDAR, an IMU, an NVIDIA Jetson, and networking equipment. The project required me to create a custom power distribution system and write software to drive it with a PS4 controller. You can read more about my research with ROS on my resume.

PrimeMind: Neural Network Classification of Amazon Prime


In this project, we utilize BERT to classify Amazon products into 112 unique sub-categories and 20 main categories based on textual descriptions and titles. The dataset selected was the 2023 Amazon Product Sales Dataset, a comprehensive collection of over 500,000 product listing with details such as names, categories, descriptions, brands, and prices. BERT-base achieved an accuracy of 89.5% after only one one epoch. When we extended the training of BERT-base to 15 epochs, we did not see a significant improvement of more than 1%. Most of the incorrect classifications involved placing a product into an adjacent category. This suggests that the model struggles with distinguishing related sub-categories where product titles share common keywords and concepts.

Magnetic Sand Plotter


Using a magnetic polar gantry this device drags a ball bearing through sand to plot beautiful patterns. It features an ESP32 to receive commands wirelessly and can plot vector images. This project was the 1st place winner of the 2021 MasterpiECE Mania design contest at Michigan Tech. It was selected for presentation at Open Sauce 2024 in San Francisco.

6DOF Virtual Reality IR Tracker


This device was designed emulate and abilities of the HTC Vive Tracker 3.0 enabling 6DOF positioning using the SteamVR Tracking System. I created custom PCB sensors utilizing the TS4231 and an IR photodiode. An array of these sensors connected to an ESP32 along with an IMU to determine position and rotation of the device in 3D space. The device demodulates the 6-12MHz IR signal and multiplexes the 5+ input channels. The pose of the device is wirelessly transmitted to a PC using Bluetooth.

Infinity Mirror


This device utilizes two parallel mirrors to create the infinity effect using individually addressable LEDs. It features capacitive touch buttons for selecting brightness, color, and modes. This project was the 2nd place winner of the 2020 MasterpiECE Mania design contest at Michigan Tech.

Portfolio Website


This website was created to show off my abilities as a front-end developer. It doubles as my digital business card and features my projects, experience, and education. It was developed using GitHub for version control and was hand written in VS Code without the use of generative AI.