Based on the forecast volume of shipments, timely decisions must be made as to when and where new logistics centers need to be built.
- A calculation of the forecast took up to 3 weeks
- Potential of data volumes could not be met
- Forecast reliability was only possible for short periods of time
Result & customer benefit
- Processing of 7 times the amount of data results in a forecast with significantly higher quality
- Calculation time reduced from 3 weeks to 15 minutes with the new forecast model
- Machine learning was used to create a self-optimizing forecast model that enables the integration of additional data sources
