Abstract

Abstract
Supplier development contributes to supply chain integration and performance, thus playing an essential role in any industry. In the retail market, no formal procedures exist for the selection-assignment of supplier development programs, meaning that respective management is complex and unclear given actor and requirement heterogeneities. Multi-criteria techniques are currently used for selection and evaluation processes. However, no models exist that integrate clustering and multi-criteria techniques together to efficiently select-assign supplier development programs. This study proposes a four-phase methodology - diagnosis, program design, assignment, and proposal - that considers supplier clustering through K-means and multi-criteria fuzzification. Additionally, case-study analysis of 149 retail suppliers in Colombia clustered businesses into high (8), medium (32), and low (109) risk sets, each of which was assigned tailored development programs. The obtained results support use of the proposed methodology to improve supply chain performance for organisations with many suppliers requiring development programs.

Abstract
In a more and more competitive and industrialized market, it is essential that companies realize that the way forward must go through the optimization of their production processes, reducing the costs and increasing product quality. Nowadays it's necessary to adopt innovative management models that can provide increased productivity at minimal costs, such as the Lean thinking. The metalworking industry is integrated into one of the most competitive existing markets in Portugal. Given this, it's fundamental to reduce the waste in all sectors of the production process, using the good Lean principles and practices, such as the Single Minute Exchange of Die, also known as SMED methodology. This paper presents a project of implementing the SMED methodology in the cold profiling process, in a population of five different profiling machines. The results of the SMED implementation show an average OEE improvement of 10,8%.

Abstract
Brazil is the largest sugarcane producer in the world and the leader in the production of sugar and ethanol. Although sugarcane has become an important factor in the Brazilian economy, cultivation has presented many issues, for example, the problems due to burning before the manual harvest. The Brazilian authorities have approved a law that prohibits this practice and mechanized harvesting has thus become the most fitting approach. Given this development, areas for sugarcane plantation must be properly rebuilt to accommodate the new way of harvesting. The main characteristic demanded of sugarcane plots to use harvesting machines is that they must be rectangular. In the present paper, we propose a methodology for dividing the plantation area into plots and planning their allocation so as to accommodate mechanized harvesting. In view of the requirement for plots to be rectangular, we represented this problem as a two-dimensional cutting problem, and to find a solution we adopted the AND/OR graph approach. The computational experiments were conducted using real cases, and the proposed strategy was shown to perform well.

Abstract
The load balance aspect of the Container Loading Problem (CLP) has been handled in an simplified way in the literature. Either load balance has been treated as a soft constraint or the geometrical centre of the container has been assumed to be the ideal location for the centre of gravity of the cargo, or both, which does not meet regulatory directives and transportation legislation. In this paper, we treat load balance as a 'hard constraint and adopt vehicle specific diagrams that define the feasibility domain for the location of the centre of gravity of the cargo, according to the vehicle specific technical characteristics, thus fulfilling and complying with real-world regulations and legislation. We propose a multi-population biased random-key genetic algorithm (BRKGA), with a new fitness function that takes static stability and load balance into account. Extensive computational experiments were performed with different variants of the proposed approach. Also solutions taken from the literature were evaluated in terms of load balance. The computational results show that it is possible to obtain stable and load balanced solutions without compromising the performance in terms of container volume utilization, and demonstrate also the advantage in incorporating load balance in the packing generation algorithm.

Abstract
Retailers' individual products are categorized as part of product families. Merchandising rules specify how the products should be arranged on the shelves using product families, creating more structured displays capable of increasing the viewers' attention. This paper presents a novel mixed integer programming formulation for the Shelf Space Allocation Problem considering two innovative features emerging from merchandising rules: hierarchical product families and display directions. The formulation uses single commodity flow constraints to model product sequencing and explores the product families' hierarchy to reduce the combinatorial nature of the problem. Based on the formulation, a mathematical programming-based heuristic was also developed that uses product families to decompose the problem into a sequence of sub-problems. To improve performance, its original design was adapted following two directions: recovery from infeasible solutions and reduction of solution times. A new set of real case benchmark instances is also provided, which was used to assess the formulation and the matheuristic. This approach will allow retailers to efficiently create planograms capable of following merchandising rules and optimizing shelf space revenue.