Toward a Resilient Automotive Supply Chain: New Framework for Sustainable Supplier Selection using MCDM and Machine Learning Tools

The supplier selection field is a crucial part of the automotive supply chain. Despite its importance, it faces several challenges, such as geopolitical instability, intense competition among organizations, and regulatory and environmental pressures. In this context, it becomes essential to integrate sustainability and resilience considerations into the supplier selection phase. This integration is primarily achieved through sourcing activities, which directly influence overall supply chain performance. In this article, we present a clear methodology that can be used by companies for sustainable supplier selection. This methodology is based on a comprehensive literature review of supplier selection criteria and methods. Building on this theoretical foundation, a survey was conducted with experts from the automotive industry to validate the identified criteria. Based on this theoretical and practical groundwork, we collected data and developed a framework for supplier selection within the automotive supply chain. Supplier selection is recognized as a Multi-Criteria Decision-Making problem. To address this, we apply Subjective Weighting Method SWM with Survey results to rank suppliers. To enhance our analysis, we compare the results obtained from SWM with those generated by Machine Learning techniques, specifically K-means clustering and RankNet. The primary objective of this research is to propose and validate an integrated decision-making framework that enhances sustainable and resilient supplier selection in the automotive sector. Secondary objectives include identifying the most critical supplier selection criteria through expert validation, assessing the consistency and complementarity of SWM and Machine Learning outputs, and providing practitioners with a replicable, data-driven methodology adaptable to various automotive commodities.