The Impact of Inspector Fallibility on the Inspection Policy in Serial Production Systems

TLDR

Inspection plays a critical role in ensuring quality within production systems. The study develops a model to determine the optimal placement of inspection stations in n‑stage linear production systems. The model incorporates predictable and erratic inspector fallibility, optimizes cost per good unit, and examines the cost‑quality surface through sequential sensitivity analysis. Results show that predictable inspector fallibility significantly affects inspection station placement and cost, while variability has little effect, offering managers guidance on optimal station numbers and policy trade‑offs.

Abstract

Appropriate inspection is a significant component of production systems. In this paper a model is developed to determine the optimal placement of inspection stations within n-stage linear production systems. This model accommodates two types of inspector fallibility: “predictable,” which implies that the error rates are known and constant, and “erratic,” which requires a random variable to describe inspector performance. Cost per good unit accepted by the customer is used as the optimizing criterion. The cost-quality response surface is explored through a sequential sensitivity analysis. Our results indicate that under certain conditions the level of predictable inspector fallibility significantly impacts the number and placement of inspection stations as well as cost per good unit produced. The modeled systems, however, were quite insensitive to the variability of inspector performance. This production-inspection model provides management with information on the optimal number and placement of inspection stations for specific planned or existing serial production systems. It can also be used by management to explore various policy options, such as the cost implications of increasing the quality vs. the quantity of inspection stations. The data required by the model can be obtained at reasonable cost provided management is willing to estimate or determine judgementally certain of the variables.