Quantifying Cascading Delays: An Epidemic Modeling Framework for Risk Propagation Analysis in Smart Manufacturing Systems

Current smart manufacturing systems feature complex resource interactions and tight coupling, posing significant challenges for analysis and optimization. Minor disruptions, like material delays, can trigger cascading failures where risks propagate through the manufacturing network, amplifying impacts and altering equipment performance dynamically. This nonlinearity complicates predicting the effect of local adjustments in large systems. While prior research often focused on supply chains using probabilistic models or complex network theory, this research targeted plant-level production systems, aiming to quantitatively analyze dynamic risk propagation in smart manufacturing systems, and identify key factors enhancing network stability. Leveraging the Susceptible-Infected-Susceptible epidemic model principles, this research formulated the trend of delay propagation and represented network dynamics via differential equations. The critical conditions determining whether a delay will spread exponentially, stabilize, or fade were identified, which can be used as prediction tools in digital twins. Discrete event simulations were employed to demonstrate the sensitivity of maintenance parameters for cascading failure containment under disruptions. The findings provide insights into risk propagation mechanisms and offer manufacturers targeted recommendations to improve system resilience.