A Record of Four Contests Entered in 2026 — Two Accepted, Two Rejected
Introduction
In the first half of 2026, I applied to four contests and grant programs using ReLow60 L-HE as my subject. The results were two accepted and two rejected.
Each application was, at its core, “a story about the ReLow60 keyboard,” yet what each program was actually asking was surprisingly different. One evaluated the polish of what I had built; another probed whether the project was viable as a business; another asked what questions I was genuinely pursuing; and another examined how I was collaborating with AI. I ended up re-explaining the same product four times in completely different terms.
The process was a significant learning experience, so I’m documenting it here — rejected applications included. The details of each contest are covered in individual articles.
Overview
| Contest / Program | Organizer | Result |
|---|---|---|
| TiB FAB Makers Challenge 2026 | Tokyo Innovation Base (Tokyo Metropolitan Government) | Accepted |
| AI Utilization Promotion Grant Program for Learning and Research | Chiba Institute of Technology | Accepted |
| DigiKey Make ONE Challenge 2026 | DigiKey | Rejected (first round) |
| QWS Challenge | SHIBUYA QWS | Rejected |
Individual Articles
The weaknesses I self-reported on the application sheet, and the content and Q&A from the second-round presentation
An unusual selection process where the dialogue with AI itself served as the application
The development log submitted to ProtoPedia — comparing three sensor types and selecting capacitors
The full text of the 'question' I wrote on a form that asked nothing about specifications
Retrospective Across All Four Applications
Even with the same product, what each program asked was entirely different
This is the single strongest impression that came out of doing all four.
| Contest | What was being asked |
|---|---|
| DigiKey | What I built and how I built it |
| QWS | What questions I was pursuing |
| TiB | How I planned to sell it and sustain it as a business |
| AI Utilization Grant | How I was collaborating with AI |
I had the same ReLow60 in hand every time, yet I had to reframe my project from a completely different angle for each application. The measurement data I wrote up for DigiKey would have landed flat at QWS, and the philosophy I articulated for QWS would have gotten a “so what did you actually build?” at DigiKey.
A single project can be explained in multiple ways — I hadn’t truly felt that until I went through this process.
The rejected applications supported the accepted ones
The writing I did for the two rejected applications was not wasted.
The question I articulated for QWS — “Is the form of that tool truly ‘optimal’?” — appeared almost verbatim at the opening of my TiB presentation, and I also used it to explain my design philosophy during the AI grant dialogue. When the dialogue ended with “If you were to start over, what would you change?”, my answer was “I’d lead with the QWS application text and my design philosophy.”
The ProtoPedia page I wrote for DigiKey has continued to serve as a reference ever since.
The ideas I put into words for rejected applications ended up supporting the accepted ones. Since QWS and DigiKey came first and TiB and the AI grant came second, the first two effectively served as drafts for the latter two.
My own weaknesses became visible
What the AI grant dialogue exposed most sharply was that I’m good at breaking work into fine-grained tasks, but poor at placing those tasks realistically within a timeline and resource constraints. I had failed to do the basic back-calculation that, given PCB manufacturing lead times, there would be no room for a second order.
The TiB application sheet also required me to list four known weaknesses and disadvantages of my project at this stage. Working through that — covering mass production, quality assurance, market fit, and funding — made what I’m lacking quite clear. An application is written to get accepted, but it is simultaneously a forced audit of where you currently stand.
Looking Ahead
I’m now working on an improved version and a new product for an event to be held in Tokyo in September. The TiB program runs through 2027 and the AI utilization grant through May 2027, so I plan to write up the outcomes of both in future articles.
Low-profile magnetic switch keyboard