Designing a Safety Condition Monitoring System for Body Harness Installation Using the Decision Tree Method

Work safety in high-altitude activities remains a major issue, particularly due to body harness installation errors such as hooks not being attached to anchor points or excessive tension on safety ropes. This study aims to develop a sensor-based body harness safety system and Decision Tree algorithm for real-time automatic monitoring. The system uses inductive proximity sensors to detect hook attachment and MPX5010 pressure sensors to measure excessive pressure on the harness rope. Sensor data is processed by a Decision Tree algorithm embedded in an Arduino Uno microcontroller to classify safe and unsafe conditions. In unsafe conditions, the system sends an emergency notification in the form of a misscall via the SIM800L GSM module. Sensor testing results show a 100% success rate for proximity and pressure sensors. Overall system testing using a confusion matrix results in 88% classification accuracy without false negatives. The proximity sensor functions as hook attachment verification, while the pressure sensor is the main parameter with a threshold of 3.425 kPa. The results of the study show that the system is capable of improving work safety at heights through effective early warning.