Impact of Machine Reliability Data Uncertainty on the Design and Operation of Manufacturing Systems

Abstract

Decision making in the design and operation of advanced multi-stage manufacturing systems is more and more supported by digital manufacturing tools. In order to be effective in their scope, such tools have to be based on high-fidelity virtual representations of the real system. To achieve this goal, they are continuously fed with process and system data directly collected from the field. Once validated, these digital tools can be used to evaluate and generate alternative system improvement actions and optimized re-designs of the system, based on scenario analysis. Traditionally, manufacturing systems engineering methods suitable to this scope include analytical methods and simulation. While evaluating the performance of the system under a given configuration, they typically assume that machine reliability parameters (Mean Time to Failure and Mean Time to Repair) are precisely known. However, in practical situations, these parameters are either estimated from real life data or based on experts’ knowledge. In both cases, they are subject to estimate uncertainty. This paper investigates the risks and the potential performance losses due to design and operation decisions derived by neglecting machine reliability uncertainty in the digital manufacturing tools. The proposed method paves the way to the on-line adoption of digital models for manufacturing system continuous improvements.