I build hardware,
then I build the deck.

The first real design decision was architectural: build one sealed airframe that does both jobs, instead of pairing a separate air vehicle with a separate underwater vehicle. A two-vehicle approach is the obvious way to solve transmedium flight, but it's heavier, harder to manufacture, and adds a second full set of failure points. I chose a single-airframe architecture instead, which meant every subsystem — structure, propulsion, avionics — had to survive and function in both domains rather than being optimized for just one.

That drove the rest of the mechanical design. The vehicle is VTOL, so it launches and recovers without a runway, catapult, or net — deployable from a vessel, a shoreline, or open ground. Rotor arms fold for efficient flight and reconfigure as the vehicle enters the water, and I isolated all sensitive electronics in a single dry, pressure-tolerant bay so the same avionics and flight controller govern behavior in both domains instead of handing off between two onboard computers. Propulsion is domain-specific by design — rotors tuned for stable, efficient flight and a dedicated underwater thruster for low-noise submerged movement — with the air-to-water transition designed to be fast and repeatable rather than a fragile one-time event, though that's still a design target to validate once the physical prototype exists.

Navigation has to solve the same domain-switching problem. In air, the vehicle relies on GPS and inertial sensing; the instant it goes underwater GPS disappears, so the plan is to hand off to dead reckoning from its own inertial and motion data, then bring sonar online for object detection and underwater navigation. Making that handoff automatic and continuous is one of the hardest open problems on the project — the autonomy software that will run it is still being architected, and I'm working through that design alongside our software/autonomy engineer rather than calling it a finished capability.

The bag is built around a modular compartment system I designed and filed a provisional patent on in January 2025, aimed at a problem every gymnast has but no backpack solved: chalk from grips and chalk blocks contaminates everything it touches. The design uses a hinged horizontal divider — a "trap door" — that isolates a chalk-only lower compartment from a clean upper one, a fully removable grip pouch that attaches with hook-and-loop so an athlete can carry just their grips to an event, and removable vertical dividers that let the same bag reconfigure between separated, expanded, and fully open modes. PE board reinforcement keeps the structure from collapsing under real gym-bag loads.

Getting there took iteration, not one clean design. I built and photographed a first-generation version in early 2024, tested it with real gymnasts including my own teammates, then redesigned the compartment layout, strap system, and exterior materials into the current 2.0 version before sourcing manufacturing partners to produce it at scale.

On CU Boulder's Formula SAE team I designed the high-voltage junction box that distributes power from the accumulator on our electric race vehicle, coordinating with the electrical and integration teams so my sealing and enclosure choices didn't create problems downstream. My process started from the failure mode rather than the happy path: I researched enclosure waterproofing and sealing approaches for wet operating conditions first, then designed the enclosure around maintaining electrical integrity under those conditions — treating a sealing failure on a live high-voltage system as a safety event, not an afterthought.

Level 10 JE gymnast, Above The Bar Gymnastics Academy, Nov 2014 – Jun 2023. Competed at elite level through Level 10 JE; qualified for JO Nationals and Future Stars Nationals at the U.S. Olympic Training Center; earned Academic All-American honors and qualified for the R3 Elite Team. Nine years of training is where the product insight behind Locked In Athletics came from.