Intelligent urban solid waste management system based on IoT and Deep Learning for Arequipa, Peru

Paredes Rodriguez J.; Revilla M.M.; Talavera Suarez J. · Przeglad Elektrotechniczny (2026) · DOI: 10.15199/48.2026.09.40

Resumen

Solid waste management in Latin American cities suffers from operational flaws due to the lack of live data and reliance on fixed routes that do not account for the spatiotemporal variability in waste generation. This article presents the design and implementation of an intelligent waste management system for the city of Arequipa, Peru. The system combines embedded sensors in containers, a low-power, long-range communications network, and machine-learning algorithms designed to predict fill levels and actively optimize collection routes. The proposed platform collects and processes live telemetry, models generation patterns using deep-learning time-series techniques, and resolves routing using an adaptive heuristic with continuous event-based updates. In controlled pilot tests, the system showed potential reductions in kilometres travelled and service time, as well as decreased overflows and better operational punctuality. The results indicate that integrating IoT and deep learning into municipal waste collection enables evidence-based decision-making, increases efficiency, and contributes to urban sustainability, delivering a practical modernization path for local governments.

Instituciones: Universidad Nacional de San Agustín de Arequipa