期刊名称：CORE Discussion Papers / Center for Operations Research and Econometrics (UCL), Louvain
出版年度：2010
卷号：2010
期号：1
出版社：Center for Operations Research and Econometrics (UCL), Louvain
摘要：This paper introduces a novel way to measure the variability of order flows in supply chains, the
peakedness. The peakedness can be used to measure the variability assuming the order flow is a
general point pro- cess. We show basic properties of the peakedness, and demonstrate its computation
from real-time continuous demand processes, and cumulative demand collected at fixed time intervals
as well. We also show that the peakedness can be used to characterize demand, forecast, and
inventory variables, to effectively manage the variability. Our results hold for both single stage and
multistage inventory systems, and can further be extended to a tree-structured supply chain with a
single supplier and multiple retailers. Furthermore, the peakedness can be applied to study traditional
inventory problems such as quantifying bullwhip effects and determining safety stock levels. Finally,
a numerical study based on real life Belgian supermarket data verifies the effectiveness of the
peakedness for measuring the order flow variability, as well as estimating the bullwhip effects.
