Gonçalo Reis Figueira

Abstract
Supplier selection for strategic items requires a comprehensive framework dealing with qualitative and quantitative aspects of a company's competitive priorities and supply risk, decision scope, and uncertainty. In order to address these aspects, this study aims to tackle supplier selection for strategic items with a multi-sourcing, taking into account multi-criteria, incorporating uncertainty of decision-makers judgment and supplier–buyer parameters, and integrating with inventory management which the past studies have not addressed well. We develop a novel two-phase solution approach based on integrated multi-criteria decision-making (MCDM) and multi-objective simulation-optimization (S-O). First, MCDM methods, including fuzzy AHP and interval TOPSIS, are applied to calculate suppliers’ scores, incorporating uncertain decision makers’ judgment. S-O then combines the (quantitative) cost-related criteria and considers supply disruptions and uncertain supplier–buyer parameters. By running this approach on data generated based on previous studies, we evaluate the impact of the decision maker's and the objective's weight, which are considered important in supplier selection. © 2023 Elsevier Ltd

Abstract

Abstract
Supplier selection has received substantial consideration in the literature since it is considered one of the key levers contributing to a firm's success. Selecting the right suppliers for different product items requires an appropriate problem framing and a suitable approach. Despite the vast literature on this topic, there is not a comprehensive framework underlying the supplier selection process that addresses those concerns. This paper formalizes a framework that provides guidance on how supplier selection should be formulated and approached for different types of items segmented in Kraljic's portfolio matrix and production policies. The framework derives from a thorough literature review, which explores the main dimensions in supplier selection, including sourcing strategy, decision scope and environment, selection criteria, and solution approaches. 326 papers, published from 2000 to 2021, were reviewed for said purpose. The results indicate that supplier selection regarding items with a high purchasing importance should lead to holistic selection criteria. In addition, items comprising a high complexity of supply and production activities should require integrated selection and different sources of uncertainty associated with decision scope and environment, respectively, to solve it, as well as hybrid approaches. There are still many research opportunities in the supplier selection area, particularly in the integrated selection problems and hybrid solution methods, as well as in the risk mitigation, sustainability goals, and new technology adoption.

Abstract

Abstract
The emergence of Industry 4.0 is making production systems more flexible and also more dynamic. In these settings, schedules often need to be adapted in real-time by dispatching rules. Although substantial progress was made until the ???90s, the performance of these rules is still rather limited. The machine learning literature is developing a variety of methods to improve them. However, the resulting rules are difficult to interpret and do not generalise well for a wide range of settings. This paper is the first major attempt at combining machine learning with domain problem reasoning for scheduling. The idea consists of using the insights obtained with the latter to guide the empirical search of the former. We hypothesise that this guided empirical learning process should result in effective and interpretable dispatching rules that generalise well to different scenarios. We test our approach in the classical dynamic job shop scheduling problem minimising tardiness, one of the most well-studied scheduling problems. The simulation experiments include a wide spectrum of scenarios for the first time, from highly loose to tight due dates and from low utilisation conditions to severely congested shops. Nonetheless, results show that our approach can find new state-of-the-art rules, which significantly outperform the existing literature in the vast majority of settings. Overall, the average improvement over the best combination of benchmark rules is 19%. Moreover, the rules are compact, interpretable, and generalise well to extreme, unseen scenarios. Therefore, we believe that this methodology could be a new paradigm for applying machine learning to dynamic optimisation problems.