Odometry features from on-board sensors are streamed into an artificial neural network, providing real-time haptic feedback for the user to help minimise their risk of injury whilst exercising and generate useful heuristics on the workout.
This repository contains all the code housed on-board the Smartbell:
Code for interacting with the hardware sensors and haptic feedback soldered to the Raspberry Pi.
Neural network implementation. Interprets the data stream from the sensors and outputs appropriate feedback.
Training data for Belle
Code to provide data to an online server, which in turn will relay it to the user's phone app.