Abstract
Manufacturing environments commonly present uncertainties and unexpected schedule disruptions. The literature has shown that in these environments simple and fast dynamic dispatching rules are efficient sequencing methods. However, most of the works in the automated designing of these rules have considered deterministic processing times. This work aims to design dispatching rules for problem settings similar to the ones found in real environments such as uncertain processing times and sequence-dependent setup times. We use Genetic Programming to generate efficient rules for stochastic job shops with setup times. We show that the generated rules outperform benchmark dispatching rules, specially in settings with high setup time levels. © 2020, Springer Nature Switzerland AG.

Abstract
Soon, a new generation of Collaborative Robots embodying Human-Robot Teams (HRTs) is expected to be more widely adopted in manufacturing. The adoption of this technology requires evaluating the overall performance achieved by an HRT for a given production workflow. We study this performance by solving the underlying scheduling problem under different production settings. We formulate the problem as a Multimode Multiprocessor Task Scheduling Problem, where tasks may be executed by two different types of resources (humans and robots), or by both simultaneously. Two algorithms are proposed to solve the problem - a Constraint Programming model and a Genetic Algorithm. We also devise a new lower bound for benchmarking the methods. Computational experiments are conducted on a large set of instances generated to represent a variety of HRT production settings. General instances for the problem are also considered. The proposed methods outperform algorithms found in the literature for similar problems. For the HRT instances, we find optimal solutions for a considerable number of instances, and tight gaps to lower bounds when optimal solutions are unknown. Moreover, we derive some insights on the improvement obtained if tasks can be executed simultaneously by the HRT. The experiments suggest that collaborative tasks reduce the total work time, especially in settings with numerous precedence constraints and low robot eligibility. These results indicate that the possibility of collaborative work can shorten cycle time, which may motivate future investment in this new technology.

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
Food waste reduction represents a potential opportunity to enhance environmental sustainability. This is especially important in fresh products such as fresh seafood, where waste levels are substantially higher than those of other food products. In this particular case, reducing waste is also vital to meet demand while conserving fisheries. This paper aims to promote more sustainable supply chains by proposing daily fresh fish demand forecasting models that can be used by grocery retailers to align supply and demand, and hence prevent the production of food waste. To accomplish this goal, we explored the potential of different machine learning models, namely Long Short-Term Memory networks, Feedforward neural networks, Support Vector Regression, and Random Forests, as well as a Holt-Winters statistical model. Demand censorship was considered to capture real demand. To validate the proposed methodology, we estimated the demand for fresh fish in a representative store of a large European retailing company used as a case study. The results revealed that the machine learning models provided accurate forecasts in comparison to the baseline models and the statistical model, with the Long Short-Term Memory networks model yielding, in general, the best results in terms of root mean squared error (27.82), mean absolute error (20.63) and mean positive error (17.86). Thus, the implementation of these types of models can thus have a positive impact on the sustainability of fresh fish species and customer satisfaction.

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.

Abstract
In the current global market, managing supply is not a straightforward process and it becomes even more complex as uncertainty and disruptions occur. In order to mitigate their impact, the selection of suppliers of strategic items should have a more holistic view of the operations in the supply chain. We propose an integrated model for supplier selection, considering inventory management and inbound transportation. We approach this problem, incorporating stochastic demand and suppliers' imperfect quality. Imperfect quality triggers additional costs, including external failure and holding costs. Supply disruptions also affect the suppliers' lead time, resulting in delivery delays. We develop a methodology to address this challenge with simulation-optimisation. A genetic algorithm determines supplier selection decisions, while inventory decisions are computed analytically. Discrete-event simulation is used to evaluate the overall performance, as well as to update the lead time dynamically, according to the disruptions. Finally, sensitivity analysis providing managerial insights reveals that criteria in supplier selection should be given a different priority depending on the characteristics of the items, and the effectiveness of disruption mitigation strategies depends on the disruption characteristics.