Development of Smart Light Detection and Monitoring System with Machine Learning

In this paper, the smart light detection and monitoring system is presented. The light sensors are used to detect the light temperature value in the room. The Arduino UNO and ESP32 are used as the main embedded systems. The ZigBee transmitters are used as the main transmitter. MySQL is set as the database. Moreover, Machine Learning is used to classify the light temperature value to natural light or fluorescent light. Test result shows that the cross-validation accuracy of KNN is 0.996. This accuracy of KNN is greater than other algorithms. Moreover, this system can be used to examine the status of the light in the room that can solve the problem of forgetting to turn off the light and can analyze data of using light as well.