Explainable AI-powered edge computing solution for smart building energy management in green IoT

Today, climate change and global warming are among the most serious problems of humanity. In combating these problems, urgent and serious actions are needed especially in energy preferences, utilization, and management. Especially in the building sector, energy consumption has increased rapidly and today it has reached 40% of total global energy consumption. Therefore, the use of low-cost, eco-friendly, and sustainable green technologies is critical to mitigate the negative environmental impacts of carbon emissions and the depletion of the world's energy resources. In this context, emerging technologies such as Artificial Intelligence (AI), the Internet of Things (IoT), and edge computing are both necessary and promising in terms of ensuring energy efficiency in buildings, grid security, and supplydemand balance. In this paper, we propose a low-cost end-to-end IoT system architecture for energy monitoring and management in smart buildings. In this architecture, we develop explainable ML models that predict building energy consumption based on edge computing. We also develop a mobile application with video call and instant messaging features for monitoring and managing energy consumption for expert users. Experimental and test results show that the proposed system can be used in building energy management in a fast, effective and interpretable way with the support of AI and IoT. With the developed prototype architecture, we present a future projection that the energy management of buildings in Green IoT can be low-cost, transparent, and understandable.