Integrating human factors aspects into performance optimization models of a serial assembly system

Abstract

This dissertation investigates the effects of human factors (HF) of the working environment on the performance of an operation system. Poor HF design of the workplace interrupts the balance of the working environment and reduces employees' overall work performance creating a substantial economic burden on organizations. This thesis focuses on integrating HF aspects into performance optimization models of the serial system. For this reason, a modeling framework has been developed for hierarchical consideration of HF consequences at the individual, workstation and system levels. The developed framework provides a road map for the three analytical phases of this PhD research. In the first analytical phase, a two-state Markov chain is developed to quantify the connection between Work-related Ill Health (WIH) risk factors (ergonomic conditions in the workplace) and employee health-state in a probabilistic way. Subsequently, an optimization model is developed to minimize the total cost of the assembly system with regard to employee health-related productivity loss. Numerical results indicate that there is between 0.5% and 8% difference in the optimal cost of the system with and without including HF effects. In the second analytical phase, a three health-state Markov chain models the connection between HF aspects of the workplace and the employees' work-related productivity and quality variations. Results show between 0.02% and 32% increase for the optimal total cost when both employee productivity and quality losses due to poor HF design of the workplace are integrated into the optimization model. In the third analytical phase, the uncertainty involved in customer demand is considered by developing a two-regime switching model, using a pentanomial lattice. The developed modeling approach investigates the effects of both work-related employee performance variation and demand behavior on the optimal cost of the serial assembly system. Results show that a prediction of the demand distribution throughout the product life cycle is necessary to reduce the over/under cost estimation of the system, due to the stochastic behavior of the demand. This research opens a new window for considering HF intervention not only as occupational health and safety but also as operation improvement method leading to design safer and more efficient systems.