Digital Twin (DT) technologies are increasingly adopted as valuable methods to support the analysis and optimization of Business Processes (BPs). The use of simulation-based techniques in the BPM context is not new. Nonetheless, existing approaches for BP simulation primarily focus on performance-related aspects. In this respect, this paper introduces a data-driven framework for the automated generation of reliability-aware DTs derived from event and state logs. The proposed method integrates process mining techniques, resource reliability modeling, and model-driven transformations to produce executable DT simulations capable of predicting the failure behavior of process resources. The proposed framework has been applied to the predictive maintenance domain, where DT-based simulations are used to estimate resource availability, identify possible failure scenarios, and support the timely scheduling of preventive maintenance interventions. A manufacturing case study shows that the proposed approach can enhance process operability and reduce downtime compared to conventional BP execution without DT-based reliability insights.
Bocciarelli et al. (Fri,) studied this question.