A Data Driven Framework to Predict the Fatigue among Manufacturing Workers Using Wearable Sensors

Abstract

Worker fatigue has been known as a significant phenomenon in the manufacturing occupations. In these occupations, physical fatigue is a challenging ergonomic/safety issue as it lowers productivity and boosts the incidence of injuries. The objective of this dissertation is to prevent the fatigue occurrence in the manufacturing occupations by monitoring the individual’s body using the wearable sensors on the wrist, torso, ankle, and hip coupled with a heart rate sensor. Specifically, this research,1) examines whether the commercially wearable sensors, extracting appropriate ergonomic-related metrics, can be used to detect the occurrence of fatigue on an individualized level for different occupational tasks, 2) proposes a comprehensive framework consisting of four phases including detection, identification, diagnosis, and recovery to manage fatigue in manufacturing occupations using wearable sensors. Overall, the goal of this research is to develop analytical models that present important findings for accident and injury prevention by managing fatigue in the manufacturing occupations.