Hydrofoil Watercraft

This was my comprehensive Master's project, where I designed and engineered the mechanical systems, custom circuitry, and firmware from the ground up. To establish a rigorous scientific foundation, I mathematically modeled and verified the thermal cooling, hydrofoil lift, and electrical load requirements prior to physical prototyping.

IoT Enabled Watering System

To automate plant care and learn new integration protocols, I developed a custom watering system. It features integrated soil moisture and ambient light sensors that publish telemetry to a local Home Assistant instance via MQTT. The system monitors environmental conditions and automatically triggers a water pump when required.

The system enclosure, PCB layout, and control firmware were all custom-designed and fabricated.

Hydrofoil Drive PCB

Developed as a core subsystem for the hydrofoil craft, this custom board integrates dual Electronic Speed Controllers (ESCs) running the AM32 firmware. An onboard STM32 microcontroller processes inputs from an RF receiver while an IMU provides active stabilization data. The crystal oscillator layout features a grounded guard ring to minimize electromagnetic interference (EMI).

I conducted extensive mathematical modeling to simulate the board's thermal dynamics prior to fabrication. By calculating heat dissipation across the layer stack and heat sink connection, I verified that the MOSFET temperatures would remain well within safe operating limits, with an expected delta of under 25°C.

7-segment Breakout PCB

To facilitate efficient in-field hardware debugging without relying on a full debugger bridge or serial terminal, I designed this standalone diagnostic board. Powered by an ATmega4809 microcontroller, it displays real-time system states and sensor readings directly on a 7-segment screen.

Hydrofoil Drive Logic

This embedded software acts as the primary system controller for the hydrofoil boat, coordinating telemetry and drive signals. Developed in C via STM32CubeIDE, the firmware handles real-time PPM drive command output, processes RF receiver inputs via UART, and continuously monitors battery voltage to prevent cell degradation.

IoT Watering System Firmware

Architected around a rigid finite state machine (FSM) in embedded C, this firmware manages the telemetry collection, network synchronization, and actuation cycles for an automated plant irrigation system. Running on an Arduino MKR microcontroller, it orchestrates transitions between sensor polling, MQTT communication, Home Assistant command listening, and low-power sleep states.

To facilitate battery-operated deployment, the firmware implements aggressive power management: it cuts the power rail of analog sensors via a dedicated GPIO switch, gracefully disconnects the WiFi radio module, and initiates a low-power deep sleep state for 15-minute intervals. Telemetry measurements are averaged over multiple iterations to filter analog noise before publishing to Home Assistant, which can trigger the 12V pump cycle remotely.

Machine Vision Tabletop Aid

To streamline range calculations and targeting in tabletop games, I developed a computer-vision assistant that tracks real-time hand gestures. Using pinch motions, players can dynamically position and scale virtual area-of-effect templates, which are projected onto the board from overhead. The application features a FastAPI backend, enabling seamless remote control from any browser or mobile app.