Name
Abstract | ||
|---|---|---|
Demand planning in the semiconductor industry is typically divided into different planning horizons, mid-term and short-term. Accurate demand forecasting is crucial because of long capacity installation times, long lead-times, short product life cycles, and constantly new technological advances. As demand forecasting for short and mid-term horizons are often made on different product and time granularities using different planning tools, we may see demand fluctuations (on the same granularity) within individual horizons and at the intersections of different granularities. This paper discusses stability of demand forecasts depending on time and product granularity and introduces definitions of good and bad stability, using Symmetric Mean Absolute Percentage Error (SMAPE) as a measure for stability. We show that time and product granularities have a significant effect on the intra-horizon stability of a demand plan and that planning on different granularities can lead to artificial demand fluctuations at the intersections of planning horizons.
|
Year | DOI | Venue |
|---|---|---|
2017 | 10.5555/3242181.3242499 | WSC '17: Winter Simulation Conference
Las Vegas
Nevada
December, 2017 |
Field | DocType | ISBN |
Symmetric mean absolute percentage error,Artificial demand,Industrial engineering,Demand forecasting,Systems engineering,Computer science,Bridging (networking),Capacity planning,Granularity,Semiconductor industry,Seven Management and Planning Tools | Conference | 978-1-5386-3427-1 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
4 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Nicola Schuster | 1 | 0 | 0.34 |
| Hans Ehm | 2 | 40 | 12.67 |
| Andreas Hottenrott | 3 | 0 | 0.68 |
| Tim Lauer | 4 | 0 | 0.34 |