The 'Arm Force Field' method to predict manual arm strength based on only hand location and force direction

Area of Science:

• Biomechanics
• Ergonomics
• Human Factors Engineering

Background:

• Estimating manual arm strength (MAS) is crucial for ergonomic assessments.
• Current methods for MAS prediction have limitations in accuracy and generalizability across diverse conditions.
• The need for a more robust and reliable strength prediction model is evident.

Purpose of the Study:

• To develop and validate a novel computational method, the Arm Force Field (AFF), for predicting manual arm strength.
• To incorporate the effects of body orientation, hand location, and force direction into MAS predictions.
• To account for the contribution of arm weight in strength estimations.

Main Methods:

• Development of an artificial neural network (ANN) model to predict MAS.
• Integration of hand location and force direction variables within the ANN.
• Inclusion of a sub-model to estimate the influence of arm weight on strength.
• Validation of the AFF model using both internal and external datasets.

Main Results:

• The AFF method demonstrated high accuracy in predicting MAS, with an R-squared value of 0.97 and a Root Mean Square Deviation (RMSD) of 5.2 N for the internal dataset (n=456).
• The model maintained good generalizability with external test data, achieving an R-squared of 0.842 and an RMSD of 13.1 N (n=80).
• The AFF method proved to be a robust, reliable, and valid approach for estimating arm strength capabilities.

Conclusions:

• The developed Arm Force Field (AFF) method provides a highly accurate and generalizable prediction of manual arm strength.
• The AFF method can be seamlessly integrated into Digital Human Modeling (DHM) ergonomics software.
• This novel approach offers a superior alternative to existing methods for estimating arm strength in ergonomic evaluations.