An IoT ML-Based system for Energy Efficiency in smart homes

Internet of Things (IoT) technologies are developing rapidly, creating new challenges, especially on the energy saving side of its resources. This work presents an IoT-based management system for automatically controlling air conditioning, TV and lightning in smart homes. In this system, sensors collect data regarding temperature, humidity, luminosity and motion in monitored room. These data are parameters to perform interventions in smart homes in order to avoid keeping them turned on when the environment is not being used. This paper also presents the use of Machine Learning techniques in the analysis of data gathered from the sensors to feed it to a predictive model to determine when the appliances and lights should be turned on or off.

I Introduction

Worldwide energy consumption is showing a huge increase in the last decade. Several reasons are behind this such as the increase of appliances that use electrical energy, the introduction of the internet of things which implies a growth in energy consumption by sensors, home appliances and people unawareness of the amount of the consumed energy due to their bad habits which is one of the paramount problems in IoT Home automation field. With IoT appearance, energy saving and comfort in smart homes are one of the major topics that interest researchers. With this, we are proposing this system for household appliances that can contribute to energy saving by using machine learning (ML) algorithm to track inhabitant's usage of two appliances (Air conditioner, TV) and lights to reduce energy consumption. The system shows great prediction results that will help in reducing energy consumption by turning the appliances and lights off when they should not be used.