Modell zur Analyse und Gestaltung des Bestellverhaltens für die variantenreiche Serienproduktion

Abstract

Ein immer turbulenter werdendes Umfeld und die damit verbundenen höheren Anforderungen an die Stabilität und Flexibilität von ganzen Netzwerken stellt Industrieunternehmen vor eine neue Ausgangssituation. Die veränderten Umfeldbedingungen sind mit Risiken verbundenen, bieten aber allen Beteiligten auch die große Chance, sich zukünftige Wettbewerbsvorteile zu sichern. Das Thema gewinnt insofern an Bedeutung, als eigentlich jedes Industrieunternehmen ungeachtet seiner Größe tagtäglich in Netzwerken über seine zahlreichen Schnittstellen agiert. Genau hier unterstützt die vorliegende Arbeit die Schaffung einer neuen Qualität der Prozess-Synchronisation an der Schnittstelle zu anderen Unternehmen. Die hier formulierten Anforderungen und die Darstellung des Standes der Technik zeigen, dass verbesserte und neue Prinzipien, Verfahren und Modelle zur Harmonisierung von Kunden-Lieferanten-Beziehungen notwendig sind. Das in dieser Arbeit entwickelte Modell basiert auf den zwei Kernbausteinen Analyse und Gestaltung von Bestellverhalten. Im ersten Teil wird eine Methode zur Quantifizierung von Bestellverhalten beschrieben, mit deren Hilfe das Bestellverhaltens einer Güteklasse zugeordnet wird. Ein Unternehmen kann damit sowohl sein eigenes Bestellverhalten oder das seines Kunden bewerten. Auf Basis dieser Ergebnisse wird dann das Bestellverhalten interpretiert. Diese Interpretation belegt, inwieweit das gemessene Bestellverhalten ein Problem im Netzwerk darstellt, und inwieweit dieses Bestellverhalten „logisch nachvollziehbar“ ist. Das Analysemodell berücksichtigt dabei die wesentlichen vier Handlungsbedarfe, die bei der Bewertung des Standes der Technik identifiziert wurden. Diese vier Handlungsbedarfe beziehen sich auf die ganzheitliche Berücksichtigung der drei Dimensionen Bestellmenge, Bestelltermin und Zeitpunkt der Bestellung, die Identifikation des tatsächlichen Kundenwunsches, die mit einer Schwankung verbundenen Risiken sowie den Zusammenhang zwischen Schwankungscharakteristik und möglichen Schwankungsursachen. Diese Interpretation liefert damit die Basis für den Gestaltungsteil des hier entwickelten Modells. Darin werden konkrete Handlungsempfehlungen für eigen- und fremdinduzierte Turbulenzkeime vorgeschlagen. Bei der Reduzierung eigeninduzierter Bestellschwankungen helfen konkrete Ursachen-Wirkungs-Zusammenhänge zwischen beispielhaften Schwankungsmustern und dazugehörenden typischen „Stolpersteinen“ aus der Praxis. Da sich eine Vielzahl an Turbulenzen aufgrund ihres fremdinduzierten Charakters nicht vermeiden lassen, wird dargestellt, wie Kunde und Lieferant die zu erwartenden Turbulenzen mittels Schwankungskorridoren vereinbaren können. Bei der Analyse des „Standes der Technik“ wurden zwei besonders wichtige Kriterien identifiziert: die Berücksichtigung spezifischer Abhängigkeiten zwischen Produkten oder Varianten bezüglich der Lieferflexibilität sowie eine praxisnahe Vorgehensweise für Kunde und Lieferant zur Erlangung einer derartigen Vereinbarung. Das hier vorgestellte Modell leistet einen wichtigen Beitrag zur besseren Zusammenarbeit an der Schnittstelle Kunde-Lieferant und damit zur Stabilität ganzer Netzwerke. Die Verifizierung des Modells mit einem abstrahierten Datenmodell belegt, dass die Anforderungen an ein derartiges Modell erfüllt werden können. Darüber hinaus können auch Teilelemente des entwickelten Modells in der Praxis eingesetzt werden, etwa bei der Vereinbarung einer Kunden-Lieferanten-Vereinbarung, die auch ohne vorangehende retrograde Analyse des Bestellverhaltens möglich ist.

Recent developments have called into question the established principles of successful business management. These developments find expression in the increasing saturation of “first world“ markets, where demand tends towards individualization. Reasons are a demographic shift resulting in an aging population and a fragmentation of target groups. In addition, an ongoing and deepening process of internationalization and deregulation has taken place over the past years. It has opened up new sales and supply markets, but also initiated a new type of competition at international level. From a technological point of view, the innovation cycles have steadily shortened due to new technological solutions at product and process level and in information and communication technology. Numerous expert papers on manufacturing engineering describe the trend towards a more dynamic environment, along with more customized products, using the terms „turbulence“ or „environmental turbulence“. A highly complex and dynamic environment makes it difficult to generate forecasts and confronts manufacturing companies with new challenges in global competition. Customers call for highly customized products with state-of-the art technology to be delivered as soon as possible, while being less and less willing to pay for it. To remain successful, companies must counteract the increasing requirements at product, price and service level. The future will only see those companies survive in the market that continue to meet the exact needs of their customers. These companies must be able, at any time, to identify new trends or spark new ones. With a lower degree of vertical integration, they need to focus on their core competencies and produce products in value chains or networks of competent, creative and innovative companies. That is why learning how to manage a network, and to think and to act in terms of a network, are essential capabilities. Many of the activities carried out by companies - often to cut down on costs - increase the significance of high-quality network management. Examples are known as „global sourcing“, „outsourcing“ (focus on core competencies by contracting out value-adding activities, e.g. modularization in vehicle construction), but also the ongoing reduction of stocks and the amount of capital tied-up. A greater inclusion of partners distributed across the world, as well as the reduction of safety stocks to a minimum, require a well-functioning network. To this must be added the growing market demand for short delivery times, for a more flexible handling of customer change requests and for high schedule reliability. As a result, the call for greater stability and flexibility does not only affect the network as a whole, but each individual customer-supplier interface in the network. Moreover, the number of interfaces grows with ongoing intertwining of the network. Above all, these changes require raising the quality of process synchronization in order to optimize the customer-supplier-interfaces. Several solutions have been pursued as part of supply chain management, but they do not go far enough. The following three key problems arise in industrial practice: - There are no universal indicators - resulting in a lack of visibility - which allow to measure the quality of a customer’s order and call-off patterns. That’s why this advance information is not considered appropriately, even though it has a considerable impact on the delivery performance of the supplier. - There are numerous turbulence germs in the companies that trigger fluctuations in the order and call-off patterns. There is a lack of methods to identify and eliminate self-induced turbulence. - Growing environmental turbulence results more often in externally driven turbulence, i.e. market- or consumer-driven turbulence. There are no appropriate approaches in science and business that enable arrangements which minimize turbulence between customer and supplier. The aim of this thesis is therefore to find concrete solutions to the three problems described above in order to contribute to the process synchronization of customer-supplier interfaces. The quality of the customer-supplier interface is a key determinant of competitiveness in a manufacturing environment where series products are manufactured at regular intervals over an extended period. Both the continuous order flow and the deliveries at short intervals provide the necessary basis for a model that allows analyzing and designing the order pattern. The model developed in this thesis is based on two key components for the analysis and design of order patterns. The first part describes a method to quantify order patterns so that they can be sorted into quality categories. A company can thus assess both its own order pattern as well as that of its customers. The outcome allows interpreting the order pattern. The interpretation determines if and how problematic the measured order pattern is for the network and if this order pattern is „logically comprehensible“. The analysis model considers the four key trouble spots identified when assessing the state of the art. These four areas of action require taking into full consideration the three dimensions of order quantity, order date and delivery date, the identification of actual customer needs, the risks associated with fluctuation, and the relationship between fluctuation characteristics and possible causes of fluctuations. This interpretation provides the basis for the design part of the developed model. It recommends specific actions to eliminate self- and externally induced turbulence germs. Specific cause-and-effect relationships referring to exemplary fluctuation patterns and associated typical ‘obstacles’ in real life help to reduce the self-induced order fluctuations. Due to the fact that many turbulence germs are externally induced and therefore cannot be avoided, this thesis suggests how customer and supplier can handle the expected fluctuations by agreeing on fluctuation margins. The analysis of the „state of the art“ revealed two major criteria: specific dependencies between products or variants in terms of flexible delivery need to be considered, and a practical approach developed for customers and suppliers to come to such an agreement. The model presented here makes an important contribution to improving collaboration at the customer-supplier interface and thus to making networks more stable. The model has been verified by an abstract data model proving that the requirements set for such a model can be satisfied. Moreover, it is also possible to apply subcomponents of the developed model in practice, for example, when working out customer-supplier agreements, which is also possible without an advance retrograde analysis of the order pattern.