Transparency in Machine Learning (ML) and Artificial Intelligence (AI) models is increasingly important for reliable decision-making in safety-critical industries. This
thesis examines the role of Explainable Artificial Intelligence (XAI) methods in revealing the reasoning behind model outputs, in contrast to traditional black-box ML
approaches.
Through a structured review and assessment of XAI tools, this research develops a
methodology to support practitioners in applying explainability techniques in a practical context. The methodology is designed to bridge the gap between ML experts and
non-specialists by adapting complex XAI outputs into formats that are more intuitive
and accessible for domain experts, thereby enhancing interpretability and supporting
informed decision-making.
A framework is introduced and demonstrated through a real-world case study in
condition monitoring. The framework combines textual and visual explanations, presenting ML-generated insights in ways that can be more readily understood without
reducing analytical value.
Validation of the framework is carried out using established principles and engineering knowledge to assess the consistency and usefulness of the transformed explanations.
This focus on both accessibility and reliability aims to encourage wider adoption of XAI
in engineering practice, addressing the need for clearer and more dependable ML insights in high-stakes environments.