Executive Summary A number of organizations sell products in containers that can be reused. Illustrative cases include beverages (in kegs, plastic cartons, etc.) and liquid gases (in cylinders). The time from issue to return of an individual container is usually not known with certainty and there is a chance that the container is never returned (because of loss or irreparable damage). In such a situation, even if the demand pattern is known and level with time, it is still necessary to acquire new containers from time to time. Such an acquisition must be initiated when the inventory level of containers drops too low in relation to the anticipated net demand (demand minus returns) during the replenishment lead time. Thus, it is important to forecast the net demand as well as to obtain an estimate of the accuracy of the forecast. In this paper, four different forecasting procedures, based upon different amounts of information, are developed. The information possibilities include recent issues period by period and recent returns (either individually identified or on an aggregate basis). The methods are compared on a wide range of simulated data, including some cases based on empirical data obtained from industry. Not surprisingly, the use of additional information improves the performance. However, most of the benefit associated with using the costly approach of identifying and tracking the issues and returns of individual containers is achieved by the more practical method of recording only aggregate issues and aggregate returns period by period.

No abstract

No abstract

No abstract

No abstract