Title: Ergonomic risk assessment based on a Bayesian-optimised expert system

Authors: Chris Dansie; Richard Sesek; Donald S. Bloswick

Addresses: Department of Mechanical Engineering, University of Utah, 50 S Central Campus Dr RM 2110, Salt Lake City, UT 84112, USA ' Department of Industrial and Systems Engineering, Auburn University, 3312 Shelby Center, Auburn Alabama 36849, USA ' Department of Mechanical Engineering, University of Utah, 50 S Central Campus Dr RM 2110, Salt Lake City UT 84112, USA

Abstract: Improper application of an ergonomic analysis tool increases the likelihood of high-risk jobs not being detected, thus jeopardising worker's health. Likewise, significant time and cost may be incurred by redesigning jobs improperly identified as high risk. Utah Intelligent Data Driven Ergonomic Assessment System (IDDEAS) is an expert system that aggregates the outputs of multiple analysis tools to create a more predictive ergonomic analysis tool. Rules in the expert system were optimised by processing health outcome data from hundreds of jobs and analysing the resulting relationship between the system's prediction and the known health outcomes. IDDEAS appears to improve in predictive ability with addition of expert knowledge and to be more accurate at predicting the level of risk than individual ergonomic analysis tools used alone. The study also provides insight into methods of quantifying health outcomes and analysis tool outputs for use in software systems that integrate or compare ergonomic information.

Keywords: data driven ergonomic assessment; Bayesian optimisation; ergonomic analysis; ergonomics; expert systems; intelligent ergonomic assessment; health outcomes; human factors.

DOI: 10.1504/IJHFMS.2013.055782

International Journal of Human Factors Modelling and Simulation, 2013 Vol.4 No.1, pp.23 - 34

Received: 25 Oct 2011
Accepted: 28 Nov 2012

Published online: 18 Jul 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article