US 8,712,822 B2Granted

Real-time demand prediction in a fast service restaurant environment

AssigneeHyperActive Technologies, Inc.

InventorsAbhishek Sharma, Robert Craig Coulter, Kerien W. Fitzpatrick

Life of this patent

Estimating the term…

Loading patent…

Abstract

Predicting future sales of a product includes determining a first, historical sales rate curve of historical sales of the product and determining a second sales rate curve for a time t from sales of the product occurring up to a time t−1. The second sales rate curve is determined for a window of time W that includes time t, the window of time W has a shorter duration than the first sales rate curve, and the first and second sales rate curves are for different periods of time. The second sales rate curve is compared to the first sales rate curve and a section of the first sales rate curve is identified based on predetermined criteria for the comparison. Future sales of the product are predicted as a function of a portion of the first sales rate curve in time after the identified section of the first sales rate curve.

Claims18

Dependent claims

A method of predicting future sales rate of a product comprising:

(a) electronically storing a first, historical sales rate curve of historical sales of a product;
(b) electronically determining for a time t a second sales rate curve for sales of the product up to a time t−1, wherein the second sales rate curve is determined for a window of time W that includes time t, the window of time W has a duration less than the first sales rate curve, and the first and second sales rate curves are determined from actual sales data acquired during mutually exclusive periods of time;
(c) electronically comparing the second sales rate curve to one or more portions of the first sales rate curve that have the same duration as window of time W;
(d) electronically identifying a section of the first sales rate curve based on a best fit for the electronic comparison in step (c);
and
(e) electronically predicting future sales rate of the product based on integration of the following equation from time t to a time t+w;
SaleRateHist
⁡
(
t
)
⁢
(
1
+
(
SalesRateHist
⁡
(
t
)
-
avgSalesHist
⁡
(
t
+
w
)
)
avgSalesHist
⁡
(
t
+
w
)
)
where:
w is a window of time of the first sales rate curve after the section of the first sales rate curve identified in step (d),
SalesRateHist is an historical sales rate pattern, and
avqSalesHist is average sales, and
wherein each sale rate curve is a discrete time series approximation of a first order derivative of sales.

40 more elements

A system for predicting future sales rate of a product comprising:

means for storing a first, historical sales rate curve of historical sales of a product;
means for determining for a time t a second sales rate curve for sales of the product up to a time t−1, wherein the second sales rate curve is determined for a window of time W that includes time t, the window of time W has a duration less than the first sales rate curve, and the first and second sales rate curves are determined from actual sales data acquired during mutually exclusive periods of time;
means for comparing the second sales rate curve to one or more portions of the first sales rate curve that have the same duration as window of time W;
means for identifying a section of the first sales rate curve based on a best fit for the electronic comparison;
and
means for predicting future sales rate of the product based on integration of the following equation from time t to a time t+w;
SaleRateHist
⁡
(
t
)
⁢
(
1
+
(
SalesRateHist
⁡
(
t
)
-
avgSalesHist
⁡
(
t
+
w
)
)
avgSalesHist
⁡
(
t
+
w
)
)
where:
w is a window of time of the first sales rate curve after the section of the first sales rate curve identified by the means for identifying,
SalesRateHist is an historical sales rate pattern, and
avqSalesHist is average sales, and
wherein each sale rate curve is a discrete time series approximation of a first order derivative of sales.

40 more elements

A non-transitory computer readable medium having stored thereon non-transitory computer readable program code which, when executed by a processor, causes the processor to perform the steps of:

(a) electronically store a first, historical sales rate curve of historical sales of a product;
(b) electronically determine for a time t a second sales rate curve for sales of the product up to a time t−1, wherein the second sales rate curve is determined for a window of time W that includes time t, the window of time W has a duration less than the first sales rate curve, and the first and second sales rate curves are determined from actual sales data acquired during mutually exclusive periods of time;
(c) electronically compare the second sales rate curve to one or more portions of the first sales rate curve that have the same duration as window of time W;
(d) electronically identify a section of the first sales rate curve based on a best fit for the electronic comparison in step (c);
and
(e) electronically predict future sales of the product based on integration of the following equation from time t to a time t+w;
SaleRateHist
⁡
(
t
)
⁢
(
1
+
(
SalesRateHist
⁡
(
t
)
-
avgSalesHist
⁡
(
t
+
w
)
)
avgSalesHist
⁡
(
t
+
w
)
)
where:
w is a window of time of the first sales rate curve after the section of the first sales rate curve identified in step (d),
SalesRateHist is an historical sales rate pattern, and
avgSalesHist is average sales, and
wherein each sale rate curve is a discrete time series approximation of a first order derivative of sales.

40 more elements

Dependencies are inferred from the claim text.

Facts

Application: 11/951,538Filed 6 Dec 2007 · us provisional application
Priority: 7 Dec 2006
Published: US 2008/0140511 A112 Jun 2008
Granted: 29 Apr 2014
Term ends: Not estimatedThe source records don’t support an estimate.

Classification3 codes

G06Q 30/02
Commerce: marketing; Price estimation or determination; Fundraising
G06Q 30/0202
Market modelling; Market analysis; Collecting market data: market predictions or forecasting for commercial activities
G06Q
Information and communication technology [ict] specially adapted for administrative, commercial, financial, managerial