Implementación de una infraestructura de computación de borde (Edge Computing) basada en Arduino UNO Q para la gestión de un instrumento audiovisual interactivo

Abstract

This degree project implements and evaluates an edge computing infrastructure based on Arduino UNO Q to manage an interactive audiovisual instrument. The proposal addresses the need to process several interaction inputs locally, without relying on cloud services. The system integrates computer vision with MediaPipe, electrical signal acquisition from a plant, motion capture through a portable module, and audience interaction through a mobile web interface. These inputs are coordinated by a Python orchestrator running on Linux Debian 13 and are converted into sound and light responses through Pure Data and WS2812B LED strips. The evaluation was carried out with all modules active, considering processor usage, memory consumption, latency, communication, and operational stability. The results showed a total processor usage of 24.6 %, a memory consumption of 982 MB, and a computer vision latency of 212.7 ms under full load. In addition, communication between the ESP32 nodes and the Arduino UNO Q remained below 50 ms. These results indicate that the platform is viable for mediumcomplexity interactive applications, although the vision module requires further optimization in higher-load scenarios.