Sistema integral de control y seguimiento personalizado para gimnasios, basado en inteligencia articial

Abstract

The sustained growth of fitness centers in Ecuador has motivated the design of a customized intelligent control and monitoring system for gym users. This system integrates an ESP32 board as the processing and wireless communication core, along with an ECG sensor responsible for capturing cardiac signals during physical activity. Power is managed by a 7.4V Li-Po battery, stabilized by an LM7805 regulator to ensure a constant supply to the components. In addition, a Dahua HC4 camera is incorporated, enabling real-time video capture to validate user entry through facial recognition, ensuring secure and automated access. Advanced image processing runs on an NVIDIA Jetson Nano, selected for its high performance in computer vision tasks. The collected data is transmitted via Wi-Fi to a mobile app developed in Flutter, which allows the user to view personalized metrics, log routines, and access recommendations based on their physical performance. During development, Google Colab was used as the training platform for the YOLOv8-based model, and Roboflow was used to manage and label the visual data. This technological solution represents a breakthrough in the automation and personalization of sports environments, offering an intelligent and adaptive experience that improves user performance, safety, and motivation. The system is presented as an innovative tool in the context of data-driven wellness and fitness.