A common obstacle in manufacturing projects involving digitalization and modelling is the lack of complete and usable data. This issue is especially evident in discrete event simulation (DES), where missing or inconsistent inputs can affect the ability to effectively build and use models. This paper presents a heuristic method based on iterative simulation and calibration to construct DES models in the presence of limited data quality, in the context of a manufacturing job shop with unclear processing and waiting times. The approach incrementally calibrates processing times until the simulated system meets observed performance, especially regarding punctuality and machine utilization. The system is therefore mathematically described and simulated, even in the absence of the complete initial data. This allows for the anticipation of model development phases and generation of tangible early outputs that help maintain strong stakeholder engagement, particularly with respect to future investments in data quality. This paper presents a case study to illustrate the method, its strengths, its limitations, and possible directions for future work.
Lanzini et al. (Thu,) studied this question.