A new scoring system for the Rapid Entire Body Assessment (REBA) based on fuzzy sets and Bayesian networks

Abstract

Traditional methods for assessing the risk of work-related musculoskeletal disorders (WMSDs) have a low sensitivity to changes in input variables. Using them, it is possible to obtain the same risk score for totally different postures, and in some cases, the effectiveness of ergonomic interventions cannot be demonstrated. This study aimed to develop a new scoring system for REBA, FBnREBA, using fuzzy sets and the Bayesian network (BN) approach to cover the drawbacks of the traditional REBA. First, the risk factors of WMSDs were defined in terms of fuzzy membership sets. Next, a BN model was developed based on REBA. Fourteen different postures were assessed using FBnREBA, and the results were compared with those of the original REBA. Lastly, a case study was performed to demonstrate how the new scoring system can be used to rank various interventions based on their effectiveness. FBnREBA is a BN model with 26 nodes and is based completely on REBA, but its results differ from those of REBA for identical postures. A comparison of the results of FBnREBA with those of REBA indicated that FBnREBA is more sensitive to changes in WMSDs risk factors than REBA. A case study was conducted using FBnREBA, and the effectiveness of modifying each body segment was determined and ranked. FBnREBA is more sensitive to changes in input variables so that it is unlikely to obtain the same risk score for different body postures. The introduced methodology can be used to modify the scoring systems of other similar methods.