Green IoT: AI-Powered Solutions for Sustainable Energy Management in Smart Devices

In response to global social and environmental challenges, cities worldwide increasingly adopt sustainable infrastructure strategies. This paper presents the architecture and results of implementing IoT-based Smart Green Energy (IoT-SGE) solutions to enhance energy management in urban settings. Key strategies include sustainable mobility policies, energy-efficient building updates, renewable energy production, improved waste management, and ICT integration. A key focus is on the development of smart city energy systems through mixes of on-site and off-site energy sources, where IoT technologies have a key role in monitoring and control. In this paper, it is proposed a technique that utilizes IoT sensors and deep reinforcement learning to predict energy demand and optimize consumption. This comprises various aspects of the architecture, including IoT devices for data collection, machine learning algorithms for predictive analytics, and best practices in management for sustainable energy. Testing results are presented, showing that the IoT-SGE solutions significantly improve energy efficiency and sustainability. In particular, the performance of this synthetic dataset using an even-thoroughly-tuned XGBoost model was moderate, with a Mean Squared Error of 9028.58 and R2 of 0.22.