Inventory management demand forecasting

In this section, it will be performed a literature review for each one of the four categories of the Demand management - Statistical Forecast, Sales and Operations Planning (S OP), Collaborative Plaiming and Forecasting Replenishment (CPFR) and Vendor managed Inventory (VMI). This review allowed identify the DDSC characteristics for each category which was used to develop the five level maturity model. 

Each of these types of inventory behaves differently and has to be managed differently. The major distinction between them is how their demand is managed. The demand for the independent inventory must be forecast as explained in Chapter 8. The demand for the dependent inventory can be calculated, because the amount of the type of inventory that is needed is directly related to the demand for the final product. For example, the demand for some components, which are in a finished item, is always dependent on the number of a finished item that we need. First, we will examine how to manage independent inventory. Then we will investigate the methods to manage dependent inventory. 

Throughout this book we have described the barriers to more effechve supply chains. We have also described many of the tools and techniques that address this barrier. For example. Section 3-7.2 describes a eollabo-ration effort called Collaborative Planning Forecasting and Replenishment, CPFR for short. CPFR requires a business relationship between partners and has taken root in the retail industry. Collaboration strives to better match demand and supply, improve inventory management practiees, and capitalize on new systems through sharing. CPFR is primarily a link between retailers and their manufacturer suppliers. However, it is expected that the CPFR concept will expand to other industries. 

Common Approaches for Modeling Demand Uncertainty and Forecast Evolution in the Inventory Management Literature... 

In this section we briefly describe a linear state space model that serves as a building block for the main models of collaborative forecasting processes we present in this chapter. We then present a well-known forecasting technique associated with this model namely, the Kalman filter. Let Xt be a finite, n-dimensional vector process called the state of the system. In the context of inventory management, this vector may consist of early indicators of future demand in the channel, actual demand realizations at various points of the channel, and so forth. Suppose that the state vector evolves according to ... 

In the previous section we presented a set of models of demand/forecast evolution that are fairly common in the inventory management literature. Clearly, this collection of models was not intended to be exhaustive. Nevertheless, these models share an important advantage They are sufficiently descriptive of demand processes in a large variety of settings, and at the same time, they are simple enough to be embedded into inventory decision models without making them virtually intractable. Indeed, the purpose of this section is to explain some of the complexities associated with the control of inventory for products that face the above types of demand processes, and to direct the reader to the relevant literature. 

Remark. A different type of approach to study forecasting issues in inventory management is that taken, e.g., by Chen et al. (1999). In their models they assume that the decision-makers are not aware of the exact characteristics of the demand process, and hence they resort to popular forecasting mechanisms such as the moving-average technique when making replenishment decisions. Chen et al. also propose heuristic policies that are very similar in their nature to those proposed in this paper. 

Aggregate planning manages supply (capacity and inventory) to handle demand variability. Hence it is a reactive process given a demand forecast. Typically, manufacturing departments are responsible for aggregate planning. 

Logistics and supply chain management have conventionally been forecast-driven rather than demand-driven. In other words, the focus has been to look ahead over a planning horizon and to predict demand at a point in time and then to build inventory against that forecast. As markets become more volatile and turbulent... 

Warehouse management was implemented in three DCs, two return centers, and several stores, to integrate inventory with other systems. It was initially focused on the DC-to-store channel and was then integrated into the catalog and internet channels. The replenishment procedure calculates daily orders in response to actual sales and updates inventory positions. Each inventory item is forecast weekly on a rolling horizon basis, and order projections are provided to the DCs and vendors. Additional capabilities include system-generated seasonal profiles, demand alerts, purchase order alerts, and order frequency optimization. 

Demand forecasts form the basis of all supply chain planning. Consider the push/pull view of the supply chain discussed in Chapter 1. All push processes in the supply chain are performed in anticipation of customer demand, whereas all pull processes are performed in response to customer demand. For push processes, a manager must plan the level of activity, be it production, transportation, or any other planned activity. For puU processes, a manager must plan the level of available capacity and inventory, but not the actual amount to be executed. In both instances, the first step a manager must take is to forecast what customer demand will be. 

A key point from the Red Tomato supply chain examples we have considered in this chapter is that when a firm is faced with seasonal demand, it should use a combination of pricing (to manage demand) and production and inventory (to manage supply) to improve profitability. The precise use of each lever varies with the situation. This makes it crucial that enterprises in a supply chain coordinate both their forecasting and planning efforts through an S OP process. Only then are profits maximized. 

The manager at Sporlmart, a sporting goods store, has to decide on the number of skis to purchase for the winter season. Based on past demand data and weather forecasts for the year, management has forecast demand to be normally distributed, with a mean of ix = 350 and a standard deviation of o- = 1(X). Each pair of skis costs c = 1(X) and retails for p = 250. Any unsold skis at the end of the season are disposed of for 85. Assume that it costs 5 to hold a pair of skis in inventory for the season. How many skis should the manager order to maximize expected profits ... 

---