Analysis of Kinematic and Kinetic Parameters during Manual Load Lifting Using Markerless Motion Capture and Inverse Dynamics Method

Abstract

Manual material handling activities continue to be a primary cause of musculoskeletal disorders in industrial environments. This study aimed to perform a biomechanical moment analysis of symmetric load lifting using kinematic data obtained via markerless motion capture technology without the use of physical sensors. Ten healthy male volunteers performed squat lifting tasks in the sagittal plane under external loads of 2.5 kg, 5.0 kg and 7.5 kg. Joint angle, velocity, and acceleration data were acquired using the MediaPipe Holistic algorithm. These data were combined with anthropometric constants to calculate joint moments using the inverse dynamics method via Newton-Euler equations. The results revealed that increased external load significantly elevated peak moment values in the hip, shoulder, and elbow joints. Specifically, as the load increased from 2.5 to 7.5 kg, normalized peak hip moments rose from 1.94 to 2.33 Nm/kg, shoulder moments from 0.30 to 0.58 Nm/kg, and elbow moments from 0.11 to 0.26 Nm/kg. In contrast, ankle and knee joint moments showed relatively small changes. Additionally, participants anthropometric differences influenced their lifting strategies, resulting in notable variations in shoulder and elbow kinematics during the load-centering and stabilization phases. In conclusion, this methodology demonstrates that ergonomic risks and mechanical joint stresses during manual material handling tasks can be practically analyzed in a non-invasive manner, independent of laboratory constraints.