Optimizing the tactical planning in the Fast Moving Consumer Goods industry considering shelf-life restrictions

https://doi.org/10.1016/j.compchemeng.2014.01.020Get rights and content

Highlights

  • Developed two computationally efficient methods to consider shelf-life limitations.

  • Hybrid method: more efficient than common methods and obtains near optimal solution.

  • Indirect method: very efficient and yields solution within few percent of optimum.

  • Indirect method with previously developed algorithm can solve cases up to 1000 SKUs.

Abstract

This paper addresses the optimization of the tactical planning for the Fast Moving Consumer Goods industry using an MILP model. To prevent unnecessary waste and missed sales, shelf-life restrictions are introduced using three methods. The direct method tracks the age of products directly. While it provides optimal solutions, it is computationally inefficient. The indirect method forces products to leave inventory before the end of their shelf-life. It obtains solutions within a few percent of optimality. Moreover, compared to the direct method, the computational time was on average reduced by a factor 32. The hybrid method models the shelf-life directly in the first and indirectly in the second storage stage. It obtains near optimal solutions and, on average, reduces the required computational time by a factor 5 compared to the direct method. Cases containing up to 25, 100, and 1000 SKUs were optimized using the direct, hybrid and indirect method respectively.

Introduction

Due to the increasingly competitive global market, companies with a global supply chain have to continuously optimize their supply chain operations. Optimizing these operations could, for example, allow a company to reduce the inventory while maintaining high customer satisfaction levels (Papageorgiou, 2009). Grossmann (2005) and Varma, Reklaitis, Blau, & Pekny (2007) review the research on Enterprise-Wide Optimization (EWO), which focuses on optimizing the procurement, production and distribution operations.

In this paper, we consider these procurement, production and distribution operations over a one year horizon. Specifically, we optimize these decisions on the tactical planning level for a Fast Moving Consumer Goods (FMCG) company. Examples of FMCG are yoghurt, ice cream and shampoo.

FMCG are products that are replaced/used up within a relatively short period, which depending on the product ranges from days to a year. They are usually quickly substituted when not available, and they are generally produced in large quantities. Because of these large quantities, they are profitable despite typically low profit margins. Therefore, optimizing the tactical planning of a FMCG company is important to ensure that the products remain profitable, while ensuring that they are available in the right place at the right time.

For example, Kellogg greatly reduced its production, distribution, and inventory cost through the use of Linear Programming (LP) planning models (Brown, Keegan, Vigus, & Wood, 2001). For an extensive review on quantitative optimization methods for the food supply chain, we refer to Akkerman, Farahani, & Grunow (2010). These authors mention that the perishability of the products is an important challenge in the optimization of the operations in a food supply chain.

Considering perishability is important because product freshness is one of the primary concerns for consumers when buying food products. Consumers can judge the freshness of a product either by evaluating the sensory qualities of the product or by the Best-Before-Date (BBD) listed on the packaging. Since many products are fully packed, the consumer must often rely on calculating the remaining shelf-life based on this BBD (Entrup, 2005).

Shelf-life is defined by the Institute of Food Science and Technology (1993) as “the time during which the food product will remain safe, be certain to retain the sensory, chemical, physical and microbiological characteristics, and comply with any label declaration of nutritional data.”

Because product freshness is important for consumers, the retailers require that the products they receive have a certain minimum remaining shelf-life. Therefore, only part of the shelf-life can be used in the supply chain up to the retailers. For the remainder of this paper, shelf-life refers to the part of the shelf-life that may be used in the supply chain before the retailers.

If the shelf-life is not considered in the tactical planning problem, part of the inventory could exceed its shelf-life. This would not only result in disposal costs, but the reduced inventory might not be sufficient to meet the demand, which would lead to missed sales. Therefore, considering shelf-life limitations in the tactical planning problem is crucial. Nevertheless, the implementation of shelf-life limitations in the tactical planning has only received limited attention in literature.

Much of the research regarding implementing shelf-life limitations focuses on adding shelf-life constraints to the Economic Lot Scheduling Problem (ELSP). An overview of the major contributions in this area is given in Soman, van Donk, & Gaalman (2004) and Entrup, Gunther, Van Beek, Grunow, & Seiler (2005). However, these models typically assume a constant demand rate. This is unrealistic for the food industry, which has many seasonal products and intense promotional activities (Entrup et al., 2005).

Another part of the research in this area focusses on the quality degradation over time. Entrup (2005) integrates shelf-life in the advanced planning for fresh food industries. He relates the revenue of a product to its remaining shelf-life. The longer the remaining shelf-life, the more valuable the product. The shelf-life is modeled by tracking the production day and selling day of each product.

Farahani, Grunow, and Guenther (2012) propose an iterative scheme that integrates the production and distribution decisions for a perishable food company. They compare their integrated approach to a sequential planning approach. A penalty is added to the objective function for the quality decay of the products. They assume a linear decay for each day that a product remains in storage. Ahumada and Villalobos (2011) consider a similar linear decay penalty for the production and distribution of fresh produce. In addition, they limit the maximum shelf-life based on the harvest period and the sales period.

Rong, Akkerman, and Grunow (2011) optimize a food supply chain, while managing the food quality. The quality degradation per period is linearly dependent on the temperature, which can be varied for each location. The shelf-life is then considered by imposing a minimum quality requirement.

Amorim, Gunther, and Almada-Lobo (2012) consider the shelf-life using two methods. In the first method, the maximum shelf-life is enforced directly through the dates of production and sales of the products. In the second method, similar to Rong et al. (2011), they adjust the remaining shelf-life in each period according to the storage conditions. They use two objective functions. In the first one, the overall costs are minimized. In the second objective, the remaining shelf-life of the products sent to the distribution centers is maximized. Using these two objectives, they consider the trade-off between costs and the value of freshness.

Eksioglu and Jin (2006) optimize the tactical planning for perishable products in a two-stage supply chain, consisting of production facilities and retailers. They add a constraint to ensure that the inventory at a production facility in any period cannot exceed the amount that is sent to the retailers in the next X periods, where X is the shelf-life. However, their model formulation limits the retailers to receiving product from a single factory.

Gupta and Karimi (2003) consider the shelf-life of intermediate products in the short-term scheduling of a two-stage multiproduct process. They introduce a constraint that forces the second stage processing of a batch of product to start before the end of the first stage processing of a product lot plus the shelf-life of the product. Using a big-M formulation, they relax this constraint for second stage batches that are not produced from this first stage lot.

Finally, Susarla and Karimi (2012) optimize the tactical planning for pharmaceutical companies while considering the shelf-life. They directly model the age of each product, and set the maximum age equal to the shelf-life.

In summary, when shelf-life is considered in literature, it is typically considered directly: either by tracking the age of products, by tracking the production and sales dates, or through the product quality. While directly tracking the shelf-life is accurate, it is relatively inefficient, as we will show in this paper. Therefore, it might not be a tractable method for larger, more realistically sized problems. In this paper, we propose two other, computationally more efficient, methods that also accurately consider the shelf-life limitations.

The remainder of this paper is organized as follows. The problem is defined in Section 2. In Section 3, the base tactical planning MILP model without shelf-life constraints is given. Section 4 introduces the three methods to include shelf-life restrictions into this tactical planning model. The results are discussed in Section 5, and the conclusions are drawn in Section 6.

Section snippets

Problem definition

Given is a set of Stock-Keeping Units (SKUs). These SKUs are products that may differ in composition and/or packaging. Given is also a supply chain consisting of suppliers, factories, warehouses, distribution centers and retailers. The operation of the supply chain is considered over a one year horizon that is divided into 52 weekly time periods to account for the seasonality of the demand.

The objective is to minimize the total costs of operating the supply chain. The costs include the

Tactical planning model

The problem described in the previous section can be represented by an MILP model. We start with the MILP model we previously proposed in van Elzakker, Zondervan, Raikar, Hoogland, and Grossmann (2014), which describes this problem without considering the shelf-life restrictions. An overview of this model will be given below and afterwards the possible methods for introducing shelf-life constraints into this model will be discussed.fTransIngh,f,s,tMaxSupplyh,s,th,s,thINVIngh,f,tINVIngCapf

Shelf-life

We consider three methods of implementing the shelf-life restrictions into the tactical planning model.

Results

First, these three shelf-life implementation methods have been applied to several relatively small case studies. The time horizon in these case studies consists of 52 weekly periods, and the supply chain consists of 5 suppliers, 2 factories, 2 warehouses, 4 distribution centers and 8 retailers. Each of these case studies contained 10 ingredients and 5 SKUs. The SKUs belonged to 2 different mixing families, 4 packing families, and 5 SKU families. Later in this paper, case studies with a larger

Conclusions

Three different methods for introducing shelf-life restrictions into a tactical planning MILP for a FMCG company were proposed and compared. The direct method, which keeps track of the age of all SKUs, provides optimal solutions but is computationally inefficient. Therefore, it is only suited for small problems. For larger problems, the hybrid method is more suitable. It tracks the age of SKUs in the first storage stage directly, while indirectly enforcing the maximum shelf-life in the second

Acknowledgements

This research is supported by Unilever, which is gratefully acknowledged by the authors.

References (17)

There are more references available in the full text version of this article.

Cited by (0)

View full text