A multiple linear regression approach to extimate lifted load from features extracted from inertial data.

Area of Science:

• Occupational Health
• Biomechanics
• Ergonomics

Background:

• Work-related musculoskeletal disorders are a significant occupational health issue.
• Heavy and repetitive manual lifting tasks are primary contributors to low back complaints.

Purpose of the Study:

• To investigate the correlation between kinematic features from acceleration and angular velocity signals and the load lifted.
• To assess the feasibility of a multiple linear regression model for predicting lifted load during lifting tasks.

Main Methods:

• Acquired acceleration and angular velocity data using an inertial sensor on the chest during lifting tasks (0-18 kg).
• Extracted time-domain features (RMS, Std Dev, MinMax) from sensor signals.
• Performed Pearson correlation analysis and developed a multiple linear regression model.

Main Results:

• Identified six key features (3 from z-axis acceleration, 3 from y-axis angular velocity) with strong correlation (r > 0.7) to lifted load.
• The multiple linear regression model achieved a high predictive accuracy (R-square > 0.9).

Conclusions:

• A combination of kinematic features and a multiple regression model effectively estimates lifted load.
• This methodology shows potential for indirectly monitoring worker load exposure in real-time.