Evaluating Wearable Sensor Technologies for Predicting Shoulder Endurance.

Area of Science:

• Biomechanics
• Ergonomics
• Wearable Technology

Background:

• Work-related musculoskeletal disorders are a significant concern in occupational health.
• Accurate prediction of physical task endurance is crucial for preventing fatigue and injury.
• Existing biomechanical models often lack real-time, wearable integration for dynamic tasks.

Purpose of the Study:

• To investigate the impact of different wearable sensors on predicting endurance times (ETs) for dynamic shoulder tasks.
• To assess the feasibility of a fully wearable system for real-time worker monitoring using a torque-based biomechanical endurance model.
• To evaluate a novel approach for generating maximum torque inputs within the endurance model.

Main Methods:

• Comparison of three prediction methods: non-wearable (motion capture), and two wearable (inertial measurement units, pressure insoles).
• Integration of sensor data with a torque-based biomechanical endurance model.
• Analysis of prediction accuracy using absolute mean error for endurance time estimations.

Main Results:

• Wearable sensor integration significantly impacted ET predictions.
• Inertial measurement units (IMUs) provided the most precise ET predictions (24.8% absolute mean error).
• Motion capture (30.2% error) and pressure insoles (29.8% error) showed higher errors, with insoles often underestimating ETs.

Conclusions:

• Wearable sensors, particularly IMUs, show strong potential for real-time fatigue prediction in occupational settings.
• Integrating wearable technology with biomechanical endurance models can enhance the prevention of work-related musculoskeletal disorders.
• Further research is needed to optimize real-time prediction accuracy in actual work environments.