Demand Distribution Classification for Safety Stock Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current safety stock planning in supply chain management systems is overly simplistic, failing to accurately characterize demand data, particularly distinguishing between regular and sporadic demand, and not accounting for differences in historical and planning periods.
Innovation Solution
A system and method that evaluates historical demand characteristics such as mean value, standard deviation, and fraction of periods without demand to categorize demand as regular or sporadic, selecting appropriate mathematical models (Gaussian or Gamma distributions) for safety stock calculations, and adjusting parameters to match planning periods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a simple variation coefficient threshold (0.5) is used to distinguish regular and sporadic demand, then the classification process is simple and quick, but the accuracy of demand characterization deteriorates
Solution Approach 1:
The patent transforms the single-parameter classification (variation coefficient only) into a multi-parameter evaluation system incorporating mean demand, standard deviation, and multiple time period analyses. This changes the classification parameters from simple to comprehensive, resolving the contradiction between simplicity and accuracy by making the classification process more informative without being overly complex.
Solution Approach 2:
The patent segments the demand classification process into distinct evaluation stages: calculating variation coefficient, calculating mean demand, calculating standard deviation, and analyzing multiple time periods. This segmentation allows each parameter to be evaluated independently and systematically, improving accuracy while maintaining operational clarity through structured analysis.
2Device complexity
If historical demand data is used directly without period adjustment, then the calculation process is simple, but the accuracy of safety stock planning deteriorates due to period length mismatches
Solution Approach 1:
The patent applies preliminary period adjustment to historical demand data before performing safety stock calculations. By pre-adjusting the variation coefficient and other parameters to match the planning period length, the system eliminates period mismatch errors in advance, resolving the contradiction between calculation simplicity and planning accuracy.
3Device complexity
If only variation coefficient is considered for demand classification, then the classification method is simple, but the reliability of safety stock planning deteriorates due to insufficient demand characteristics analysis
Solution Approach 1:
The patent creates a composite classification approach that combines multiple demand characteristics (variation coefficient, mean demand, standard deviation, and period-specific metrics) into a unified demand classification system. This composite methodology reliably distinguishes between regular and sporadic demand by synthesizing multiple indicators, resolving the contradiction between method simplicity and planning reliability.
Data Source
AI summary
A system and method are described for identifying an appropriate demand distribution to use for safety stock planning within a supply chain management (“SCM”) system. For example, one embodiment of the invention comprises a computer-implemented method comprising: extracting historical demand characteristics for a product from a specified data source, the demand characteristics including the mean value of the demand for specified periods of time and/or the fraction of periods without any demand; evaluating the historical demand characteristics including the mean value of the demand for the specified periods and/or the fraction of periods without any demand; based on the evaluation, categorizing the demand characteristics into one of two or more different predefined categories; and using the classification to perform safety stock calculations for the product over the specified periods of time.