Dependency-aware maintenance optimization of multi-component multi-asset systems: A Simulation-based approach

Industry 4.0 has made continuous monitoring of assets a viable option, enabling decision-makers to make more informed and cost-effective maintenance planning decisions. This paper presents a comprehensive framework that integrates a simulation-based optimization technique to address condition-based maintenance and logistics planning policies for systems comprising multiple assets and components. The objective is to minimize the total expected cost per time unit by determining the optimal replacement and ordering limits for each asset's components. A key novelty of this study is the simultaneous consideration of four primary types of dependencies: structural, stochastic, economic, and resource, all within individual assets and among components across different assets. Unlike prior research, which primarily focuses on dependencies within single systems, our framework explicitly accounts for inter-asset dependencies, reflecting real-world operational challenges in multi-asset environments. Besides, we introduce the novel concept of an opportunistic ordering limit, which enables dynamic order consolidation, reducing procurement costs while maintaining spare part availability. Given the problem's complexity and stochastic nature, simulation is employed to model and effectively solve the problem. The findings demonstrate that integrating dependency-aware maintenance strategies and flexible ordering policies can significantly reduce system costs, offering at least 10% savings on the total expected cost per time unit. Additionally, this study provides a systematic approach to identifying and analyzing primary cost drivers, which can directly inform and improve supplier evaluation and selection processes in real-world asset management and service logistics planning under uncertainty.