A Neural Networks Approach to Determine Factors Associated With Self-Reported Discomfort in Picking Tasks

Objective A neural networks approach has been proposed to handle various inputs such as postural, anthropometric and environmental variables in order estimate self-reported discomfort picking tasks. An input reduction method proposed, reducing the minimum data required with similar accuracy network fed all variables. Background Previous works have attempted explore relationship between several factors using observational methods. The results showed that this was not a simple linear relationship. Another study used model function returning reported according static posture, age, anthropometrics demonstrated model’s ability predict discomfort. But available were design network. Method Eleven subjects carried-out tasks masses (0, 1, 3 kg) imposed duration (5, 10, or 15 s). Continuous REBA score, computed, subjects’ collected. set of work consisted computed continuous anthropometric, collected Results correlation estimated experimental tested equal 0.775 when 14 After reduction, only 6 left, very close performance predicting Conclusion Application This potential support ergonomists workstation designing processes, by adding prediction virtual manikins’ behaviors simulation tools.