- June 19, 2013 - 13:00
- Zaragoza Logistics Center, Room 221
Order variability in supply chain operations is one of the major concerns of industry and researchers. This type of variability can have different manifestations: variations in volume and variations in time. The dimension and direction of the relations among these manifestations are not so clear. Nevertheless, these ordering variations and their relationships affect different operational decisions at upstream levels. This research aims to understand more clearly, through an empirical study, the relationship between these types of order variability components, and the measurement issues that influence them. Also, we attempt to unveil how the inventory policies and end demand distributions affect such relationships. We found that the traditional way variability is measured does not properly address the measurement of these two components; therefore we proposed a new simple measure. Evaluation of the variability component relationships by means of this new measure suggests that specific ordering policies and specific end demand distributions are largely relevant for the behavior of the components.The latter may help guide practical decisions on which inventory policy to use on dampening a specific variability component in order to reduce certain types of costs within an organization.Keywords: Order variability components, inventory policies, demand distribution, measure.
“Order Variability Decomposition: A New Variability Measure on Real Data”
It has been shown that, at least in simulated scenarios of size and frequency variability components, the way they are measured largely determines the shape of their relationships. This study aims to build on this specific finding and tests how these measures of variability components behave on real data. Moreover, taking advantage of the type of data available, several models are setup to assess amplification on such variability components, and to evaluate the relationships of the type of product for both: amplification and component variability behaviors. We do this by performing model assessment with the traditional un-weighted C.V. measure, and then replicating the same evaluation with the recently proposed ADV measure.Keywords: Variability Components, Measure, Real Data, Amplification.