A metaheuristic-driven physical asset risk management framework for manufacturing system considering continuity measures

Abstract

To ensure the uninterrupted operation of physical assets, the effective implementation of risk mitigation plans (RMPs) and business continuity plans (BCPs) for critical assets is required. However, manufacturing decision-makers face the challenge of allocating limited resources between RMPs and BCPs while prioritizing sustained success. To address this challenge, a novel approach proposes integrating a metaheuristic algorithm into a risk management framework that considers interconnected risks. The framework utilizes a Bayesian Network (BN) to model interdependencies among critical physical asset risks. It determines the optimal combination of BCPs and RMPs, considering continuity measures and resource limitations. A physical asset risk assessment process evaluates operational and disruption risks based on expected loss and probabilities within the risk network. Given the complexity, the proposed framework incorporates hybrid multi-objective metaheuristic algorithms (e.g., Nondominated Sorting Genetic Algorithm II (NSGA-II), Strength Pareto Evolutionary Algorithm 2 (SPEA2), Multi-Objective Particle Swarm Optimization (MOPSO), and Pareto Envelope-based Selection Algorithm II (PESA-II)) combined with a reinforcement learning approach. A case study of a manufacturing company demonstrates the framework’s applicability and discusses the results. The findings show a significant reduction in the expected loss within the risk network and migration of most physical asset risks from high-risk to acceptable areas. Additionally, a sensitivity analysis examines the available budget, aiding decision-makers in determining the necessary compromise between asset availability and network risk level. In conclusion, the proposed framework empowers decision-makers to allocate resources effectively, prioritize RMPs and BCPs, and ensure continuity for critical physical assets, thereby fostering sustained success in manufacturing systems.