Generative AI–Enabled Resilience Assessment of Manufacturing Supply Chains

Abstract

Manufacturing supply chains are vulnerable to disruptions such as supply delays, demand fluctuations, equipment failures, and logistics interruptions, which may cause inventory imbalance, capacity reduction, and delivery delays.Existing studies often rely on predefined scenarios and static indicators, limiting the characterization of dynamic degradation and recovery processes.This study proposes a generative artificial intelligence-enabled simulation-based method for manufacturing supply chain resilience assessment.A discrete-event simulation model integrating suppliers, inventories, production workshops, finished-goods warehouses, and customer orders was developed.Generative artificial intelligence was used to generate diverse disruption scenarios, which were converted into structured simulation inputs.Through multiple simulation experiments, resilience was evaluated based on performance degradation, recovery time, order fulfilment, inventory stability, and operational cost.The proposed framework supports supply chain stress testing, vulnerability identification, and recovery strategy evaluation under complex disruptions.