In-field instrumented ergonomic risk assessment: Inertial measurement units versus Kinect V2

Abstract

Musculoskeletal disorders are a major category of nonfatal work-related diseases in the industry. Therefore, various risk factor assessment methods, such as Rapid Upper Limb Assessment (RULA), were proposed for preventing musculoskeletal disorders in industrial environments. Nevertheless, RULA measurements based on workers' self-report or external rater observation are subjective and suffer from low repeatability. Thus, the objective of this paper is to investigate the accuracy and reliability of (1) a wearable technology (using inertial measurement units) and (2) a marker-less optical technology (using Kinect V2) for RULA risk assessment tool against motion-capture cameras (reference) for different manual material handling tasks. The obtained RULA scores were compared against the reference system using the proportion agreement index (Po) and Cohen's Kappa coefficient (κ). The wearable technology showed a “substantial” agreement (κ > 0.6) with the motion-capture cameras. The marker-less technology showed inconsistency in the agreement with the motion-capture cameras ranging from “fair” (κ < 0.4) to “moderate” (κ > 0.4) agreements, caused by both self-occlusion and object occlusion. Therefore, wearable technology showed to be more suitable than marker-less optical technology for in-field ergonomic risk assessment.