The Modular Maintenance & Spare Parts Decision Intelligence Engine

The management of maintenance and spare parts is an imperative function in equipment-intensive asset management industries, and its effectiveness is essential to ensuring continuous production, limiting environmental impacts and safety risks, and maximizing production output, asset value, and organizational competitiveness. Maintenance has evolved from a skilled trade to an industrialized, multidisciplinary domain reliant on plans, decision-making processes, and information. At the same time, resilient spare parts management (SPM) and supply chains have become indispensable, as maintenance cannot be executed without them. Achieving effective maintenance and SPM relies on efficient and effective decision-making supported by information. However, as assets and operations become increasingly interconnected and complex, decisionmaking has become increasingly complex and challenged. Despite the large volume of valuable empirical data existing in organizations’ computerized maintenance management systems (CMMS), this resource remains underutilized in decision-making due to data fragmentation, scarcity, and siloing. Decision-making practices are found to be siloed and reliant on distinct information bases, tacit knowledge, and simple yet suboptimal support. In contrast, research advances toward complex optimization models that rely on assumptive, complete, and high-frequency data, not reflecting the real dynamics of empirical maintenance and spare parts data. This PhD research thesis addresses this gap between research and industry practices in data utilization for decision-making through an engineering design perspective, applying the design research methodology (DRM) through a problem-based and theory-based approach. In the context of a major offshore oil and gas case company, the research explores how existing empirical CMMS data can be leveraged using modularization theory. This paper-based thesis presents seven papers that propose a modular maintenance and spare parts decision intelligence engine, as a collective representation of the artifacts developed through the research. This collective proposal structures, links, and contextualizes empirical data to transform it into actionable information modules that can be combined and linked to decision-making areas through interfaces, supporting, improving, and automating operational, tactical, and strategic decision-making practices, while bridging organizational and IT system silos. The research shows that using empirical data with modularization principles leads to noticeable improvements in decision-making, including reduced data-gathering time and enhanced data availability. Quantitative results show reductions between 10-15% in overall stock value and 34-45% in scope stock value. Efficiency improved by 76-91% in spare parts availability decisionmaking, and maintenance planning time decreased by 88%. Decision quality improved by 4-5%, spare parts supply capabilities increased by 15%, spare parts selection errors decreased from 6.1% to 0.6% in maintenance planning, and first-attempt maintenance execution increased from 51.1% to 86.1%. Despite the scarcity and fragmentation of empirical data, valuable support and improvements to decision-making can be achieved by leveraging this data and modularization principles. From the foundation presented in this thesis, alignment of maintenance needs and spare parts availability can be supported with the continuous improvement and variety control cycle among the maintainable asset, the spare parts portfolio, and the maintenance program. As a result, this advances maintenance a step further toward effective data-driven decision-making practices.