Abstract
Production planning and scheduling in the process industry in general and in the pulp and paper (P&P) sector in particular can be very challenging. Most practitioners, however, address those activities relying only on spreadsheets, which is time-consuming and sub-optimal. The literature has reported some decision support systems (DSSs) that are far from the state-of-the-art with regard to optimization models and methods, and several research works that do not address industrial issues. We contribute to reduce that gap by developing and describing a DSS that resulted from several iterations with a P&P company and from a thorough review of the literature on process systems engineering. The DSS incorporates relevant industrial features (which motivated the development of a specific model), exhibits important technical details (such as the connection to existing systems and user-friendly interfaces) and shows how optimization can be integrated in real world applications, enhanced by key pre- and post-optimization procedures.

Abstract
Path relinking has been used for solving deterministic problems by exploring the neighborhood of elite solutions in an intelligent way. We present an algorithm that combines a mixed-integer linear solver with a truncated path-relinking method in order to solve two-stage stochastic integer problems with complete recourse and first-stage integer variables. This method takes advantage of a possible scenario-based decomposition in an innovative way. Therefore, path relinking is used to combine optimized solutions from different scenarios in order to pursue good stochastic solutions. To assess the computational performance of this method, we use the stochastic lot sizing and scheduling problem dealing with perishable products. In this problem, first-stage decision variables are linked to production sequences and production quantities. After the uncertain demand is unveiled, the second-stage variables decide on the inventory usage. Computational results show a clear advantage of the proposed method when compared to a state-of-the-art mixed-integer linear solver.

Abstract
This research aims at tackling a real-world long-haul freight transportation problem where tractors are allowed to exchange semi-trailers through several transshipment points until a request reaches its destiny. The unique characteristics of the considered logistics network allow for providing long-haul services by means of short-haul jobs, drastically reducing empty truck journeys. A greater flexibility is achieved with faster responses. Furthermore, the planning goals as well as the nature of the considered trips led to the definition of a new problem, the long-haul freight transportation problem with multiple transshipment locations. A novel mathematical formulation is developed to ensure resource synchronization while including realistic features, which are commonly found separately in the literature. Considering the complexity and dimension of this routing and scheduling problem, a mathematical programming heuristic (matheuristic) is developed with the objective of obtaining good quality solutions in a reasonable amount of time, considering the logistics business context. We provide a comparison between the results obtained for 79 real-world instances. The developed solution method is now the basis of a decision support system of a Portuguese logistics operator (LO).

Abstract
Production and distribution problems with perishable goods are common in many industries. For the sake of the competitiveness of the companies, the supply chain planning of products with restricted lifespan should be addressed with an integrated approach. Particularly, at the operational level, the sizing and scheduling of production lots have to be decided together with vehicle routing decisions to satisfy the customers. However, such joint decisions make the problems hard to solve for industries with a large product portfolio. This paper proposes an adaptive large neighbourhood search (ALNS) framework to tackle the problem. This metaheuristic is well known to be effective for vehicle routing problems. The proposed approach relies on mixed-integer linear programming models and tools. The ALNS outperforms traditional procedures of the literature, namely, exact methods and fix-and-optimize, in terms of quality of the solution and computational time of the algorithms. Nine in ten runs of ALNS yielded better solutions than traditional procedures, outperforming on average 12.7% over the best solutions provided by the latter methods.

Abstract
Large food retailers have to deal with a complex distribution network with multiple distribution centres, different temperature requirements, and a vast range of store formats. This project used an optimization-simulation approach to help food retailer Sonae MC make the best decisions regarding product-warehouse-outlet assignment, product delivery modes planning and fleet sizing.

Abstract
The design of retail backroom storage has a great impact on in-store operations, customer service levels and store life-cycle costs. Moreover, backroom storage in modern retail stores is crucial to several functions, such as acting as a buffer against strong demand lifts yielded by an increasing promotional activity, seasonal peak demand and e-commerce activities. Despite having similar functions to a distribution center, backroom storage facilities have particularities that deserve a distinct analysis. In this paper we aim to draw attention to the lack of research about this topic.