Microcontroller Based Sitting Position Detection System Using Decision Tree Method

The purpose of this research is to use a decision tree to create a system that can identify when a person is seated. The premise behind the sitting position detector is a technology that can identify a sitting position and warn employees if they are sitting in an unsafe manner. The proposed system makes use of the APDS 9960 sensor, the ESP 32 controller, the DFMini Player media player, and the Whatsapp messaging service to alert the user. The individual is asked to adopt a variety of seated postures in order to obtain the data necessary to construct the decision tree. Good position, lean to the right, lean to the left, slightly inclined forward, slightly lean to the right, slightly lean to the left, very lean forward, very lean to the right, very lean to the left, and no one on the chair are all positions the subject may be asked to perform. The information is used to create a decision tree and assess its efficacy. The decision tree has a high degree of accuracy (98.52%). The f1-score, recall, and precision of the decision tree are above 95 %. These results suggest that the decision tree method utilized for system implementation. The system's position detection accuracy is 93.33% on the hardware implementation. The overall results of the testing indicate that the sitting position detecting system is functional.