Supplier management: multicriteria group decision model supported by computing with words and machine learning techniques

• Grant number: 24/19620-3
• Support Opportunities: Regular Research Grants
• Start date: April 01, 2025
• End date: March 31, 2028
• Field of knowledge: Engineering - Production Engineering
• Principal Investigator: Luiz Cesar Ribeiro Carpinetti
• Grantee: Luiz Cesar Ribeiro Carpinetti
• Host Institution: Escola de Engenharia de São Carlos (EESC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated researchers: Francisco Rodrigues Lima Junior ; Lucas Daniel Del Rosso Calache ; Lucas Gabriel Zanon ; Rafael Ferro Munhoz Arantes

Abstract

The process of supplier evaluation and management can contribute not only to guiding purchasing decisions but also to supplier development. This approach also enhances partnership relations, strengthening the links in the supply chain. The literature on multicriteria decision techniques and computing with words applied to supplier selection and development is quite extensive. Additionally, studies proposing the application of machine learning techniques have been growing. However, supplier evaluation is not limited to the objective data analysis typical of a supervised learning. It also involves considering subjective criteria. In this case, human observation and judgment become essential for supplier evaluation. Often, judgments must be endorsed by multiple decision-makers, which adds another layer of complexity to the decision-making process. In this context, the present project aims to propose a decision support model for supplier evaluation and development, grounded in the concept of hybrid intelligence, combining human decisions with machine intelligence. The research questions guiding the construction of this hybrid decision support model are as follows: Q1 - How can machine learning techniques and SHAP be used for supplier evaluation and recommendation? And Q2 - Can machine learning techniques, in combination with CRP and SHAP techniques, automate and support the interpretation of consensus building? To answer these questions and achieve the proposed objective, this project sets forth several goals: literature review; development of a conceptual model for supplier classification and recommendation; development of a conceptual model for CRP automation; pilot application; and dissemination of results in international conferences and journals. The project is expected to yield academic results, such as the development of machine learning applications in production engineering, training of individuals, and scientific dissemination. Through pilot application and interaction with professionals, this project is also expected to contribute to the evolution of supplier management practices within the industrial community. (AU)