Optimised Home Electricity Management using Machine Learning

This project presents an integrated system that interfaces Arduino with machine learning algorithms, vibration and PIR sensors to enhance home automation and safety. The system is designed to monitor and respond to vibrations and motion within a house. Vibration and PIR sensors are employed to detect any unusual activity, and this data is transmitted to a machine learning module. The machine learning model analyses the sensor data and, based on learned patterns, triggers the appropriate actions. For instance, if suspicious motion is detected, it can turn on lights and fans for added security or energy efficiency. Simultaneously, the DHT11 sensor records environmental data, and this information is uploaded to a web server via a NodeMCU module, allowing users to remotely monitor their home's conditions.

Furthermore, an ultrasonic sensor is integrated into the system to measure distances within the house. If it exceeds a predefined threshold, it activates a DC pump, potentially preventing flooding or other water-related issues. Additionally, a gas sensor is included to detect hazardous gas levels, and in the event of a gas leak, the system triggers a GSM module to send an SMS alert to the homeowner or designated contacts, ensuring immediate awareness and action. This comprehensive home automation system leverages Arduino, machine learning, and various sensors to enhance security, comfort, and safety within a residence. It combines sensor data analysis, remote monitoring, and timely alerts to create a more intelligent and responsive living environment.

Keywords—Arduino, Nodemcu, PIR, web server, DHT11, Ultrasonic, Machine learning.