Implementation Issues

In this chapter we consider some of the issues relating to implementation of stockage models. We first discuss comparisons of VARI-METRIC with existing stockage policies, and point out errors favoring VARI-METRIC that are sometimes made.

Then we address the amount of data required by the models, and the difficulty in keeping the information current. We distinguish between two types of data: (1) primary data such as demand rates that relate to the “physics” of the problem; (2) derived data such as average repair times that are influenced by management decision-making. We argue that standard times may be more appropriate than measured values for derived data.

We discuss the effect of errors in the data on the model decisions. In many cases the estimation process is robust in the sense that large errors in the data result in small errors in the decisions. This enables us to concentrate on those data elements such as demand rates and variances whose impact on stockage decisions is most critical.