Abstract
This paper investigates how worker productivity differs in each consecutive work period based on tests conducted on a manufacturing company. We first employed a two-dimension fixed effect model to unbalanced panel data with 6,117,317 records which contain 102 employees and 1680 workdays. We also found that worker productivity between two adjacent rest breaks shows a U-shaped trend. This finding supplements the previous researches about interruption events, which only explored the effect of interruption after its occurrence. Based on our empirical findings, to obtain the lowest production system variability, a mathematical model which aims to optimally configure work and break scheduling of a production line is then presented. The optimal schedule is that one worker either begins their work when another worker from the same workstation starts their break immediately or has worked for a half of a completed work and break period. Our analysis combines empirical studies and model analysis to improve the operations management researches and practices.