This conference paper was submitted for presentation at the NAFEMS World Congress 2025, held in Salzburg, Austria from May 19–22, 2025.

Abstract

Remanufacturing worn-out industrial parts, prolonging their lifecycle, can decrease greenhouse emissions from manufacturer processes by up to 45% of their totals and also has the benefit of reducing production costs. Stricter environmental policies and regulations, focusing on a net-zero economy, increase the attractiveness of these processes to manufacturers. However, the validation of remanufactured parts relies mostly on physical testing, which counteracts some of the benefits from prolonging their lifecycle. This makes the shift from physical experiments to digital models and simulations, including systems such as digital twins, an area of focus. The ReMake testbed, developed by the National Manufacturing Institute Scotland (NMIS), aims to replace physical testing for validating reconditioned industrial parts. The main blocker to the adoption of Digital Verification and Validation (DVV) frameworks remains the credibility of models and simulations. Low credibility in model and simulation outputs can hinder decision-making due to the lack of stakeholder confidence in these outputs. This is particularly prominent in high-risk applications that require the credibility of results to be more robust to support a decision. Under the auspices of the ReMake testbed, the National Physical Laboratory (NPL) Data Science group has developed a data model based on a NASA Standard for Models and Simulations capable of enhancing the credibility of the data used by the ReMake testbed. NPL'™s Data Science group has also carried work to link risk events related to the use of remanufactured parts'”such as the risk level and impact associated with critical failure'”to the trustworthiness that models and simulations need to achieve to aid decision-making. The baseline credibility that a model needs for use in DVV is set by a risk assessment, which is often a time-consuming and human-centric activity. Risk factors were mapped into a risk data model to streamline the risk assessment generation process. This reduces human bias and creates a traceable pathway to the risk data for validation and audit purposes. The ReMake model focuses on data traceability and provenance, and the risk model maps risk factors to credibility requirements. In particular, the risk model establishes a solid foundation for objectively capturing risk aspects by leveraging FAIR'”Findable, Accessible, Interoperable, Reusable'”data, including historic data from previous DVV activities. We propose a cohesive DVV framework that combines risk assessment and data traceability to enhance model credibility and automate decision-making for all engineering disciplines. This framework would have the ability to enhance simulation credibility while also providing the infrastructure to automate decision-making processes. This integration optimises good simulation data management practices for DVV by ensuring that data credibility remains the key factor for building trust in the process. This increase in demonstrably trustworthy model and simulation results yields a higher number of decisions made without recourse to physical testing, bringing benefits to stakeholders, such as manufacturers or policymakers, in terms of efficiency and sustainability. In the context of the ReMake testbed, this would equate to more remanufactured parts being safely deployed based on the DVV framework, aligning with the larger goals of supply chain efficiency, and promoting net zero. This framework could act as a baseline for industries seeking to enhance their sustainability efforts through digital transformation. By implementing DVV-focused best practices, with a focus on the role of risk for DVV, organisations can set minimum credibility standards for their models and simulations. This approach not only enhances decision-making but also promotes greater trust in digital methodologies.