Attending the 48 hour long “Umma Hüsla Hackathon“, Susanne Kiesenhofer and I created a push-up detector equipped with a 9-DOF sensor, a micro processor, machine learning algorithms and a visualization app.

The tracking device transmits sensor data via Wifi to the host computer, where a trained machine learning model evaluates the data with the help of Wekinator.

The custom-made visual application counts the number of push-ups and gives feedback on their quality.

Check out the source code here.