Increasing the number of pieces per pallet for consumables stacking: A data-driven solution

Abstract

This in-company project focuses on data-driven solutions to increase the number of pieces per pallet. It is part of the 'Transform 2 Grow' program, which strives for step-by-step operational efficiency and commercial excellence by accelerating the execution of strategic priorities and value creation. A first stage of the project includes de defining possible opportunities to increase the number of pieces per pallet. In order to signal where improvements are possible, multiple interviews with the R&D, packaging, and supply chain department are conducted. In addition, the Dourges plant is visited to get a grasp of the production constraints. Simultaneously, a literature study is performed where insight is gained on the main drivers of all levels of packaging and where models for optimal pallet stacking are studied. Two main areas of improvement are defined and further elaborated on: the packaging matrices, more precisely the rationale behind them, and the alignment of sales and supply chain objectives considering bag efficiency. During the second stage of the project, a solution is proposed for both areas in the form of an excel tool that not only serves as a means to make an analysis of the current situation but also to continuously measure performance in an efficient and effective manner. The first opportunity explored considers the packaging matrices, which give an overview of the bags that are in use for a certain product and how many pieces can maximally fit into that bag (pack count). These bag sizes are determined in order to achieve as many pieces as possible on a pallet. However, some doubt exists on whether all configurations of bags into cartons and cartons onto pallets are taken into account before a set of bags is selected for a packaging matrix. In order to analyze whether improvements in pieces per pallet are possible by adjusting the packaging matrices, the current method of matrix construction is challenged. A VBA-programmed excel tool 'Matrix optimization' is developed in order to calculate a set of bags that could outperform the current matrix by calculating all configurations for bag widths 75 mm to 500 mm and number of bags per carton from 1 to 9. This tool is based on a heuristic inspired by literature [1, 2]. Two matrices from the incontinence group, Grace and Caroline, are analyzed. A set of bags is selected that results in the highest number of pieces per pallet on average. It is concluded that for both of these matrices, improvements are possible, as the new sets of bags outperform the current set of bags. For the matrix Grace, the new set of bags results in an extra 174 pieces on EU pallets, equal to an increase of 5.0 %. On UK pallets, which are used less often, 156 pieces extra can be placed, equal to an improvement of 3.5 %. For the matrix Caroline, the new set of bags results in 51 extra pieces on EU pallets (6 for UK), equal to an increase of 2.4 % (0.2 % for UK). The tool thus performs well as a stacking generator. However, some improvements in automation and bag selection, primarily by taking into account sales objectives and thus pack counts, need to take place to translate the tool's output into a matrix selection more efficiently and effectively. The performed analysis thus provides a first step towards a quick and effective method for packaging matrix revision and construction.