Abstract:

The increasing need for sustainable energy management has expedited the implementation of smart automation technologies in contemporary infrastructures. This study introduces a monitoring and control system based on IoT aimed at improving energy efficiency via real-time data collection, predictive analytics, and responsive load management. The system combines various sensors, intelligent meters, and communication devices to gather energy consumption data, which is analyzed at both the edge and cloud levels for insights and decision-making. A forecasting model driven by machine learning, utilizing Long Short-Term Memory (LSTM) networks, anticipates short-term energy consumption and facilitates proactive control strategies for high-energy-use devices including HVAC, lighting, and connected loads. A 30-day deployment of experimental evaluation shows the system’s effectiveness, realizing energy savings of 18–27% in key load categories. Performance metrics such as MAPE (1.83%), RMSE (0.09 kW), and MAE (0.073 kW) confirm the excellent precision of the forecasting model. The findings validate that the suggested IoT-oriented framework greatly enhances energy use efficiency, facilitates automated decision-making, and establishes a basis for scalable smart-building and smart-grid systems. Future efforts will focus on enhancing the control intelligence via reinforcement learning and increasing interoperability with renewable energy systems