Fully Automatic Gym Exercises Recording: An IoT Solution
Sizhen Bian, Alexander Rupp, Michele Magno
TL;DR
The authors address the challenge of automatic gym exercise recording by deploying ultra-low-power BLE beacons with inertial sensors on equipment, which broadcast machine type and repetition data to nearby devices. An ESP32C3 gateway collects advertisements and pushes updates to a Firebase cloud, while users’ smartwatches and smartphone apps record activity based on RSSI signals, culminating in a privacy-preserving, scalable ecosystem with a dashboard for managers. Real-world testing yielded 94.6% accuracy, demonstrating feasibility and highlighting system sensitivities such as gateway scanning windows and beacon orientation. The approach promises low cost, long battery life, and easy deployment, with future work targeting larger-scale validation, data history, free-weight exercises, and weight-tracking capabilities to broaden utility.
Abstract
In recent years, working out in the gym has gotten increasingly more data-focused and many gym enthusiasts are recording their exercises to have a better overview of their historical gym activities and to make a better exercise plan for the future. As a side effect, this recording process has led to a lot of time spent painstakingly operating these apps by plugging in used types of equipment and repetitions. This project aims to automate this process using an Internet of Things (IoT) approach. Specifically, beacons with embedded ultra-low-power inertial measurement units (IMUs) are attached to the types of equipment to recognize the usage and transmit the information to gym-goers and managers. We have created a small ecosystem composed of beacons, a gateway, smartwatches, android/iPhone applications, a firebase cloud server, and a dashboard, all communicating over a mixture of Bluetooth and Wifi to distribute collected data from machines to users and gym managers in a compact and meaningful way. The system we have implemented is a working prototype of a bigger end goal and is supposed to initialize progress toward a smarter, more efficient, and still privacy-respect gym environment in the future. A small-scale real-life test shows 94.6\% accuracy in user gym session recording, which can reach up to 100\% easily with a more suitable assembling of the beacons. This promising result shows the potential of a fully automatic exercise recording system, which enables comprehensive monitoring and analysis of the exercise sessions and frees the user from manual recording. The estimated battery life of the beacon is 400 days with a 210 mAh coin battery. We also discussed the shortcoming of the current demonstration system and the future work for a reliable and ready-to-deploy automatic gym workout recording system.