Monitoring a demand forecasting process

• US 20040088211A1
• Filed: 11/04/2002
• Published: 05/06/2004
• Est. Priority Date: 11/04/2002
• Status: Active Grant

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Smith, Nicole, Sasser, Marguerita, Kakouros, Steve, Kuettner, Dorothea
• Granted Patent: US 7,584,116 B2
• Time in Patent Office: Days
• US Class Current: 705/10
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
• CPC Class Codes: G06Q 10/06395 Quality analysis or management
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities

Abstract

Systems and methods of monitoring a demand forecasting process are described. In accordance with a demand forecasting monitoring method, a measure of forecast error variability is computed at each period of a selected time frame, and an indicator of forecast bias is computed at a given period within the selected time frame based on forecast error consistency over periods of the selected time frame prior to the given period. A computer program for implementing the demand forecasting monitoring method is described. A system for monitoring a demand forecasting process that includes a graphical user interface configured to display a measure of standard deviation of percent forecast error at each period of a selected time frame also is described.

Claims

1. A method of monitoring a demand forecasting process, comprising:

computing a measure of forecast error variability at each period of a selected time frame; and

computing an indicator of forecast bias at a given period within the selected time frame based on forecast error consistency over periods of the selected time frame prior to the given period.

2. The method of claim 1, wherein computing the forecast error variability measures comprises computing a measure of standard error at each period of the selected time frame.

3. The method of claim 2, wherein a standard error measure is computed by subtracting actual demand from forecasted demand at each period of the selected time frame.

4. The method of claim 2, further comprising dividing standard error measures by corresponding actual demand values to obtain measures of percent error at periods of the selected time frame.

5. The method of claim 4, further comprising computing a measure of standard deviation of percent error measures one or more periods of the selected time frame.

6. The method of claim 5, further comprising computing measures of inter-quartile spread based on the computed measures of standard deviation of percent error measures.

7. The method of claim 2, further comprising computing upper and lower control limits based on the computed measures of standard deviation of standard error measures.

8. The method of claim 7, wherein the upper and lower control limits are computed so that each computed standard error measure falls between the upper and lower control limits with a prescribed probability.

9. The method of claim 8, further comprising labeling the demand forecasting process at a period as being out of control based at least in part on whether or not the computed standard error measure at that period falls between the upper and lower control limits at that period.

10. The method of claim 1, further comprising contemporaneously displaying a computed forecast error variability measure and a selected acceptance limit value.

11. The method of claim 10, further comprising labeling the demand forecasting process with a state indicator based at least in part on whether or not the forecast variability measure is above or below the acceptance limit value.

12. The method of claim 1, wherein computing the forecast bias indicator at the given period comprises computing counts of positive and negative forecast errors at periods of the selected time frame prior to the given period.

13. The method of claim 12, wherein computing the forecast bias indicator at the given period comprises testing the counts of positive and negative forecast errors with respect to a binomial distribution with parameters of 0.5 and 0.5 to obtain a bias likelihood factor.

14. The method of claim 13, further comprising labeling the demand forecasting process with a bias indicator based on comparison of the bias likelihood factor with one or more confidence coefficients.

15. The method of claim 1, wherein computing the forecast bias indicator at the given period comprises computing a count of consecutive positive forecast errors after a most recent negative forecast error and computing a count of consecutive negative forecast errors after a most recent negative forecast error over periods of the selected time frame prior to the given period.

16. The method of claim 15, further comprising labeling the demand forecasting process with a bias indicator based on comparisons of the counts of consecutive positive and negative errors with a run control limit value.

17. The method of claim 16, further comprising labeling the demand forecasting process with a state indicator based at least in part on the bias indicator.

18. A computer program for monitoring a demand forecasting process, the computer program residing on a computer-readable medium and comprising computer-readable instructions for causing a computer to:

compute a measure of forecast error variability at each period of a selected time frame; and

compute an indicator of forecast bias at a given period within the selected time frame based on forecast error consistency over periods of the selected time frame prior to the given period.

19. A system for monitoring a demand forecasting process, comprising a graphical user interface configured to display a measure of standard deviation of percent forecast error at each period of a selected time frame.

20. The system of claim 19, wherein the graphical user interface is further configured to display an acceptance limit for the standard deviation of percent forecast error one or more periods of the selected time frame.

21. The system of claim 19, wherein a standard deviation of percent forecast error measure is based on percent forecast error measures equal to (F-A)/A, where F is forecasted demand and A is actual demand for each period of the selected time frame.

22. The system of claim 19, wherein the graphical user interface is further configured to display in a graph standard forecast error and upper and lower control limits for standard forecast error at each period of the selected time frame.

23. The system of claim 22, wherein the upper and lower control limits are computed so that each computed standard error measure falls between the upper and lower control limits with a prescribed probability.

24. The system of claim 19, wherein the graphical user interface is further configured to display at a given period within the selected time frame an indication of forecast bias computed based on forecast error consistency over periods of the selected time frame prior to the given period.

25. The system of claim 24, wherein the graphical user interface is further configured to display an indication of whether forecast error is positive or negative at each period of the selected time frame.

26. The system of claim 19, wherein the graphical user interface is further configured to display indication of forecast state along dimensions of forecast quality and forecast error variability.

27. The system of claim 19, wherein the graphical user interface is configured to display graphs of one or more demand forecasting evaluation metrics over time at different levels of aggregation.

28. The system of claim 27, wherein the different levels of aggregation are selectable by a user from:

individual product level, individual product line level, and multiple product lines level.

29. The system of claim 28, wherein the multiple product lines level is filterable in accordance with a user selected product category.

30. The system of claim 19, wherein the graphical user interface is configured to display an overview table containing summary metrics for one or more demand forecasting evaluation metrics for a given period.

31. The system of claim 30, wherein the graphical user interface is configured to display a graph over time of a forecast evaluation metric corresponding to a displayed summary metric selected by a user with a pointing device.