Content area
Abstract
In our research, we study the case of a pulp and paper producer who decides to establish a partnership with one of its clients. Since production and distribution capacities are limited, the producer has to plan operations in order to satisfy the demand of the partner and the demand of other clients. The partner is a merchant, thus he buys products and sells them to the final consumer, without transforming the merchandise. The merchant can choose to order from the producer or from a second supply source, depending on the price and lead times offered. Even if each partner wants to create a real partnership with mutual benefits, they usually make decisions based on their local costs rather than the global costs of the system. The producer plans operations in order to minimize production, distribution and inventory costs, while the merchant orders products so as to minimize buying, ordering and inventory costs. For this context, we try to identify which collaboration model should be used in order to exchange products and information efficiently, and generate maximum benefits for the network and for each partner.
Many authors study this problem, using game theory and the Nash equilibrium. However, this kind of approach limits the number of parameters studied. Consequently, the models developed do not keep industrial realities in view. So in order to take into account characteristics of production and distribution systems, we study the problem using the following methodology. First, we identify four potential approaches for the case study, namely a traditional system without any collaboration scheme, the regular replenishment method, VMI (Vendor Managed Inventory) and CPFR (Collaborative Planning, Forecasting and Replenishment). For each approach, we develop a decision model from the point of view of the producer and a decision model from the point of view of the merchant. More precisely, we use mixed-integer linear programs to identify the costs, revenues and constraints involved in using each approach. Afterwards, we test and compare the models, so as to find the type of relationship that is the most profitable for the system and for each partner. We also develop different incentives in order to change partners' behavior, increase the profit of the system and share collaboration benefits. Our objectives are to better understand all the dynamics of enterprise collaborations and the effect of various strategies on the planning process of each partner. We also want to verify if a particular collaboration mode can be the most profitable approach for both the producer and the merchant.
To realize this, we proceed with numerical experiments, using AMPL Studio and the Cplex solver. The experiment leads us to some observations. To begin with, the CPFR model generates the greatest total system profit, because of an efficient optimization of both shipping and inventory costs. The VMI mode is second best, since the transportation cost is optimized. The regular replenishment mode and the traditional system obtain the lowest total system profit. After comparing each model using the profit of the system, we base our investigation on the profit of each partner. More precisely, we compare two types of relationship with different levels of interaction, namely a traditional system and the CPFR method, in order to verify if the same approach can generate the highest profit for both the producer and the merchant. We observe that the CPFR mode generates the greatest profit for the producer, while the traditional system is the most beneficial for the merchant. For this reason, we define a method to share benefits so as to obtain a CPFR collaboration that is profitable for each partner. We prove that if the producer shares a part of the transportation savings with the merchant, the profit of the merchant is higher than the profit obtained with a traditional system, and the producer obtains a higher profit than the one generated by the other approach. Since CPFR is a complex method that necessitates an important implementation cost, we also investigate the use of three different incentives in order to change partners' behavior and increase the profit of a traditional system. We demonstrate that if incentives are adequately defined, they can considerably improve the profit of both the producer and the merchant, without requiring an important information exchange. However, the creation of efficient enterprise collaborations is not an easy thing. Partners must take into consideration costs, risks, information needs and all the difficulties involved in using each collaboration approach. They also have to adjust their relationship depending on the context and share benefits equitably. (Abstract shortened by UMI.)
