Development of an Edge-Enabled IoT Smart Energy Meter with Artificial Intelligence (AI)-Based Load Prediction for Device-Level Monitoring

• Adekunle O. ADEWOLE

  Department of Electrical/Electronics Engineering, Abraham Adesanya Polytechnic, Ijebu-Igbo, Ogun State, Nigeria

  Author

• Ayodeji O. ARIYO

  Department of Computer Engineering, Abraham Adesanya Polytechnic, Ijebu-Igbo, Ogun State, Nigeria

  Author

The growing demand for intelligent energy management has accelerated the integration of the Internet of Things (IoT), edge computing, and Artificial Intelligence (AI) in smart metering. This paper presents the development of an edge-enabled IoT smart energy meter with AI-based load prediction for device-level monitoring. The system employs a PZEM-004T sensor for measurement of voltage, current, power, energy, and frequency, while a Raspberry Pi serves as the edge device for local processing and storage. A machine learning framework was trained on three months of data and evaluated using k-fold cross-validation. Results show that Linear Regression achieved the highest accuracy (R²: 0.993±0.001, MAE: 0.041, RMSE: 0.051) with minimal training (0.0017s), inference time, and model size (0.05 MB). Random Forest also performed well (R²: 0.990) but required higher computation, while KNN (R²: 0.920) and LSTM (R²: 0.602) were less efficient. SHAP-based analysis confirmed that temporal and electrical features were the most influential. The best-performing model was deployed on the Raspberry Pi and integrated with a Django-based dashboard for real-time monitoring and predictive analytics, providing a practical and efficient solution for energy management.

• FJET_12_40_33

Copyright (c) 2025 FUDMA Journal of Engineering and Technology

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.