This is Gait AI, a smart insole 3 classmates and I built in 24 hours for the 2026 Babson Buildathon, winning the AI-Enhanced Athletic Performance track!
The system is made up of 4 force-sensitive resistors attached to the bottom of a typical running shoe insole (RIP the other two resistors).
These sensors were wired to an ESP-32 microcontroller and allowed us to examine in realtime whether a runner is striking with their toes or their heel while running!
My main role was prototyping sensor placement, wire management, and housing/strap prototyping and CAD.
We needed to build fast and ship a working prototype, and wanted to build something genuinely useful for such a huge audience of users (when refined, of course).
The Gallery
Our Testing Setup
Our Battery Solution
A bike inner tube was a great leg strap!
Saving Print Time by Reducing Walls!
