An edge-enabled multimodal smart home energy management system using deep learning

The rapid proliferation of Internet of Things (IoT) technologies has significantly transformed smart home environments. Heterogeneity of various IoT applications generally leads to interoperability requirements that needs to be fulfilled along with ensuring federated device configuration. In this paper, we propose a deep learning-based Smart Home Energy Management System (SHEMS) that integrates multimodal sensor data collected from diverse sources to resolve the management of heterogeneous IoT devices in smart homes. Here, all computations are done over the edge, and online servers are used for managing the whole system remotely. A prototype of the proposed system is built, and a comparative analysis is done for predicting the energy consumption of the smart devices using EMA, LSTM and ARIMA models over different Linux based SoCs which are used as local server. The system uses EMA for predicting the next day’s power consumption, ARIMA for short term and LSTM for long term power consumption prediction. MQTT protocol is used as a performance metric for evaluating the reliability, speed and robustness of the proposed model. Experimental results demonstrate that the proposed system not only ensures robust and efficient energy prediction, but also facilitates scalable and secure smart home automation. The integration of edge computing with multimodal learning makes it a promising solution for future green and intelligent living environments.

Text 127606.pdf - Published Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (2MB) Official URL or Download Paper: https://link.springer.com/10.1007/s43926-026-00345...

Actions (login required)

View Item