Excavator rollovers are a major safety concern in construction sites, often caused by excessive loads, steep slopes, or improper operation. Current systems mainly provide post-accident alerts, lacking preventive measures. This study presents a real-ti...
Excavator rollovers are a major safety concern in construction sites, often caused by excessive loads, steep slopes, or improper operation. Current systems mainly provide post-accident alerts, lacking preventive measures. This study presents a real-time rollover risk prediction system using IMU sensors and Zero Moment Point (ZMP) theory. A scaled excavator model with multiple IMU sensors (MPU6050, MPU9250) was tested under flat ground, inclined slopes (10°, 20°, 30°), and varying bucket loads. Sensor data were integrated with RecurDyn and MATLAB to calculate ZMP relative to the support polygon. Results showed that ZMP deviations outside the support area effectively indicate instability, enabling rollover risk prediction. Calibration coefficients were introduced to reduce slope-induced errors, improving accuracy. The proposed system classifies stability into safe, caution, and danger zones, providing intuitive real-time warnings. This research demonstrates the potential of sensor-based monitoring to enhance excavator safety and prevent rollover accidents.