Product Clustering Provides Transparency

Supply Chain Management (SCM) is about much more than just optimizing logistics processes and securing operational logistics processes. In addition to providing a solution for day-to-day business, it is used to validate existing processes in order to support strategic decision-making. A prerequisite for this is a reliable base of information. The Fraunhofer Institute for Technology and Business Mathematics (ITWM) in Kaiserslautern, Germany, has therefore joined forces with the Saarbrücken-based business consulting firm ORBIS to develop a solution that uses product clustering to provide a realistic base for optimization calculations.

Operational SCM as basis for controlling

Modern SCM applications provide businesses with planning algorithms that can be used to optimize the complete supply chain. SCM starts within the company. In production companies, in particular, the internal supply chains are a highly complex combination of processes and systems. This provides sufficient potential for optimizing the efficiency, transparency and controlling of logistics-related material flows, information flows and value flows across all company departments – from purchasing, through order fulfillment, production control and warehousing to transportation and shipping.
To find out where SCM can be used for the maximum benefit, a company must first develop an understanding of its own products and its own supply chain. It then needs to define the goals of its corporate strategy, and lay down qualitative and quantitative indicators so that success can be measured. Relevant questions to be addressed here include how to reduce stocks, reduce processing times, or improve the availability of the materials. Based on this information, realization scenarios can be developed that the company can use to estimate the potential for improvement.

Reducing costs – increasing performance

When optimizing a supply chain, a company must first investigate the extent to which it is possible to integrate existing solutions, and find out where new business processes and tools are needed. If a supply chain is to function effectively, a company will usually need to change its operational processes, or even introduce new organizational structures. Here, a company should not plan for the individual units in isolation, and this leads to inaccuracies and distortions. As a result, even small deviations in consumer demand become more and more distorted along the supply chain. The greater these effects, the more difficult it is for the production facilities to meet the exact customer requirements, and they are forced to increase warehouse stocks.
To reduced costs and planning effort, and to increase performance in logistics, a company also needs to define how products are to be manufactured and moved along the supply chain to the customer. The company defines the internal logistics network, that is the supply relationships between the plants, central warehouses, distribution centers and end customers on the basis of the physical material flow. On the basis of this, procedures for demand planning, network planning and detailed scheduling can be defined. This preliminary work determines subsequent business success.

SCM is usually motivated by strategy

Companies usually decide to implement SCM not just to achieve operational improvements but also to reform supply chains and thus increase their long-term competitiveness. For this strategic Supply Chain Management, SAP offers the SAP Advanced Planner and Optimizer (SAP APO). An important question that needs to be clarified when using SAP APO is what input should be used. The operational model is usually not suitable for answering strategic questions, because not every detail entered here is relevant for higher-level planning. On the contrary, the high volume of data leads to a loss of transparency. Strategic issues, in contrast, are not modeled.

Which questions should be modeled?

How does a company realize a solution that helps to address strategic issues and provides the necessary level of abstraction? First, it needs to define which questions are relevant to the task, and are therefore central to the modeling of the information base. In particular, it needs to define the basic framework for procurement, production and distribution. The results of subsequent calculations can only help a company to draw useful conclusions and apply these to the business strategy if the most important basic conditions have been modeled to correspond closely to reality.
Conventional approaches here are often susceptible to distortion. For example, a company may extract data by only taking account of the best-selling products in its range during modeling. With this approach, however, important information is lost – such as the fact that some low-selling products are very difficult to handle in distribution, and therefore generate very high costs. On the other hand, the alternative of mapping every product individually can only be achieved in companies with a very small range of products.

Limiting the loss of representative information


To solve this problem, the Fraunhofer Institute for Technology and Business Mathematics (ITWM) in Kaiserslautern, Germany, and the Saarbrücken business consulting firm ORBIS have together developed a procedure based on product clustering that can be used to create a base of information for strategic decision-making. This procedure uses a mathematical method to prove whether and in what form variables mount up in complex datasets. With product clustering, products can be described in any number of dimensions – from sales figures through process times to geometry. In this way, a company can identify a number of representative products while limiting the loss of representative information for the supply chain. The supply chain can then be modeled for just this considerably smaller range of products. Transparency is increased – alternative scenarios can be planned and evaluated. In addition, conclusions can be drawn about the complete process, and all products. Once the supply chain has been optimized for the representative products, it can be implemented in the operational sphere for all products. The company then models the supply chain on the basis of the information gained, starting with the relevant capacities, right through to bottleneck suppliers. Finally, different scenarios can be simulated to evaluate the strategic options.

The right strategy helps a company get ahead

The operational steps are important and unavoidable for efficient SCM. However, the use of further methods such as product clustering provides a company with additional options for the targeted use of the available SCM software products. Here, SAP APO, in particular, offers a number of benefits, from both an operational and strategic perspective. Ways of evaluating the success of strategic business decisions provide valuable information on whether and how the supply chain should be restructured and day-to-day business should be changed.

Frank Wilhelm