Ex Parte LI et al

Patent Trial and Appeal Board

Nov 25, 2015

12637593 (P.T.A.B. Nov. 25, 2015)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
12/637,593 12/14/2009 Sheng Li 247-098USCON 3585
131475 7590 11/25/2015
Dilworth IP - SAP
2 Corporate Drive, Suite 206
Trumbull, CT 06611
EXAMINER
STERRETT, JONATHAN G
ART UNIT PAPER NUMBER
3623
MAIL DATE DELIVERY MODE
11/25/2015 PAPER
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
___________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
___________

Ex parte SHENG LI,
KARSTEN SCHIERHOLT,
and RALF HEIMBURGER
___________

Appeal 2013–003374
Application 12/637,593
Technology Center 3600
___________

Before ANTON W. FETTING, MICHAEL C. ASTORINO, and
NINA L. MEDLOCK, Administrative Patent Judges.
FETTING, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE1
Sheng Li, Karsten Schierholt, and Ralf Heimburger (Appellants) seek
review under 35 U.S.C. § 134 of a final rejection of claims 1–18, the only
claims pending in the application on appeal. We have jurisdiction over the
appeal pursuant to 35 U.S.C. § 6(b).

1 Our decision will make reference to the Appellants’ Appeal Brief (“App.
Br.,” filed August 6, 2012) and Reply Brief (“Reply Br.,” filed January 3,
2013), and the Examiner’s Answer (“Ans.,” mailed November 7, 2012), and
Final Rejection (“Final Rej.,” mailed March 5, 2012).
Appeal 2013-003374
Application 12/637,593

2

The Appellants invented a way of forecasting demand for replacement
parts of a product that has been discontinued or phased-out. Specification
para. 2.
An understanding of the invention can be derived from a reading of
exemplary claim 1, which is reproduced below (bracketed matter and some
paragraphing added).
1. A non–transitory machine readable medium having
instructions stored therein which when executed cause a
machine to perform a set of operations comprising:
[1] retrieving data about a phased-out product
from a storage system;
[2] retrieving a group phase-out profile
associated with the phased-out product
the group phase-out profile generated by
sorting a plurality of phased-out products into
groups
and
aggregating a phase-out profile for each group
member;
and
[3] forecasting demand for the phased-out product
based on the group phase-out profile and retrieved data .
The Examiner relies upon the following prior art:
Li US 7,636,607 B2 Dec. 22, 2009
John R. Moore, Jr., “Forecasting and Scheduling for Past–Model
Replacement Parts,” Management Science, Vol. 18, No. 4, Application
Series, Part 1 (Dec. 1971), pp. B-200–B-213
Appeal 2013-003374
Application 12/637,593

3

Claims 1–18 stand rejected under obviousness type double patenting as
unpatentable over claims 1–19 of Li.
Claims 1–18 stand rejected under 35 U.S.C. § 103(a) as unpatentable
over Moore and Official Notice.
ISSUES
The issues of obviousness turn primarily on whether Moore discloses or
suggests the steps recited in claim 1.
FACTS PERTINENT TO THE ISSUES
The following enumerated Findings of Fact (FF) are believed to be
supported by a preponderance of the evidence.
Facts Related to the Prior Art
Moore
01. Moore is directed to describing a forecasting technique for past
model replacement parts. Moore B-200:Abstract.
02. Moore describes curve extrapolation and smoothing models of
demand forecasting having been thwarted in many product areas
by extreme volatility of demand and a lack of discernible sales
patterns. The desirable features of “steady state” demand
processes which permit a forecast to “catch up” and even
compensate for erroneous forecasts made in earlier periods are not
present in service part demand process. As a result, conventional
forecasting techniques produce sales estimates so far in excess of
Appeal 2013-003374
Application 12/637,593

4

actual demand that inventories larger than all-time requirements
are built up and maintained throughout a substantial portion of the
sales decay period. Only when demand actually vanishes is the
extent of an obsolescence loss recognized. Moore B-202.
03. Moore describes using standard deviation for all products
including phased out products in computing economic order
quantities (EOQ) for stock replenishment. Moore B-205.
04. Moore describes having retrieved data for and analyzed 200
completed parts histories representing nearly every parts category,
and the analysis resulted in a variety of sales decay patterns.
Moore describes a transformation of sales data from an arithmetic
scale to a logarithmic scale as producing interesting results.
Moore B-205.
05. Moore describes the analysis of data for major engine and
transmission parts, body panels, axles, springs and windshields.
The nature of the parts suggests very expensive repairs which
might be performed, if at all, with salvaged parts when the market
value of the vehicle is low. Moore found an elliptical sales decay
function with these parts. Moore B-205.
06. Moore describes the analysis of data for moderately priced
essentials and several cosmetic items for higher priced vehicles,
e.g., lamp assemblies, valves, connecting rods, speedometers, fuel
and water pumps and chrome body trim. Moore found a parabolic
sales decay function with these parts. Moore B-205.
Appeal 2013-003374
Application 12/637,593

5

07. Moore describes the analysis of data for a wide range of general
maintenance parts such as shock absorbers, brake cylinders, lines,
and pads, valve lifters and a host of seals, gaskets, brackets and
small fittings. Moore found a linear (exponential when the
logarithmic transformation is accounted for) sales decay function
with these parts. Moore B-205.
08. Moore describes every replacement part for which all-time
requirements are to be estimated having actual sales data
available for the re-indexed years. Moore describes searching for
the decay function which best describes actual demand for the
K - 1 years of sales decline as starting by fitting a particular curve
and each point. This produces N - K ellipses from which the best
fitting ellipse must be identified for comparison with the best
fitting parabola and line. Moore B-207.
09. Moore describes the best fitting ellipse being identified by first
calculating for each curve, the log of predicted sales. These can
easily be transformed into estimated yearly sales which are then
compared with actual sales in years 2, ··· , K and an error term
developed. That ellipse of the N - K generated which produces
the least estimation error is designated as the best fitting ellipse.
This same process is repeated to determine the best fitting
parabolic and linear decay functions. Finally, the single best
fitting curve, i.e., the ellipse, parabola or line with the smallest
error term, is selected and used to estimate future years’ sales as
well as the year in which all demand will vanish. Id.
Appeal 2013-003374
Application 12/637,593

6

ANALYSIS
Claims 1–18 rejected under obviousness type double patenting
We summarily affirm this rejection as Appellants do not contest it.
Claims 1–18 rejected under 35 U.S.C. § 103(a) as unpatentable over Moore
and Official Notice
As to claim 1 limitation [2], the Specification suggests the limitation
should be interpreted as reciting retrieving a group phase-out profile
associated with the phased-out product, and that profile has been generated
by sorting and aggregating as recited.2 See Spec. para. 25. .
We are not persuaded by the Appellants’ argument that
the phase-out profile is 1) an aggregation of phase-out profiles
of group members and 2) must serve as the basis for the
demand forecast for the phased-out product. This means that
the demand of forecast for the phased-out product depends on
the individual phase-out profiles of each member of the group.
Inasmuch as what the Office has characterized as the “groups”
(effectively high, medium and low cost replacement parts) are
not aggregated in any manner they cannot satisfy the limitations
of claim l. Rather a particular part is identified as falling into
the high, medium or low cost part family and then has an
elliptical, parabolic or linear decay function applied based on its
own historical demand. Thus, the demand for windshields and
the demand for transmission parts have no effect on each
other’s respective forecast demand. Since these two products
are explicitly in the same elliptical phase-out group this is

2 Appellants might consider inserting a comma after the phrase “associated
with the phased-out product” to make this clear.
Appeal 2013-003374
Application 12/637,593

7

fundamentally inconsistent with what Appellants have claimed.
It is submitted that the utter lack of aggregation (there is no
combination in any manner of the respective demand curves of
group members) of what the Office has characterized as the
individual phase-out profiles; there is no formation of a group
phase-out profile as claimed. Moreover, there is absolutely no
dependence in the forecast for one phase-out product group to
another.
App. Br. 9. Much of this argument pertains to the profile. The actual steps
recited do not affect or depend on the profiles other than to retrieve and
forecast a demand based on them. Claim 1 has three steps of retrieving data
about a product, retrieving some profile data, and forecasting demand based
on these two sets of data. It is uncontested that Moore does this much.
Although the second step (i.e., limitation [2]) recites a limitation
regarding how the profile was generated, such generation would have
occurred prior to the start of the steps in the claim. Any such steps are
outside the scope of the claim. The non-transitory medium of claim 1 does
not recite instructions for generating the profile. The second step is merely
that of retrieving the profile, which as a primitive computer operation, is
independent of how the data were generated. The third step recites
forecasting using the profile, but as no implementation is recited, the third
step as recited is independent of how the profile was generated.
As such, actual historical data for the first step and a profile with a
constant multiplier used to perform a scalar multiplication of the historical
data would be within the scope of the recited steps. Put another way, the
recitation of how the profile is generated is aspirational rather than
functional. As such the manner of generation is deserving of no patentable
weight. See King Pharm., Inc, 616 F.3d at 1279) (“The rationale behind
Appeal 2013-003374
Application 12/637,593

8

this line of cases is preventing the indefinite patenting of known products by
the simple inclusion of novel, yet functionally unrelated limitations.”)
But even granting patentable weight to the limitation, we find the
limitation is drafted with sufficient breadth to encompass Moore within its
scope. The Examiner found that Moore described grouping automotive parts
into generic groups such as major engine and transmission parts, body
panels, and similar categories. The Examiner found that Moore then found
that after translating sales data onto a logarithmic scale, each such group in
turn exhibited a sales curve that was geometrically elliptical, parabolic, or
linear. As the claim does not recite or narrow the implementation of the
recited profiles, the data Moore relied on to represent sales for each
individual part, summaries by part, summaries by category, and summaries
by geometric sales trait all fit within the scope of such a profile.
Thus, the sorting of parts by part type and by larger category and then by
geometric profile each is within the scope of sorting a plurality of phased out
products into groups. As each of these summary levels is aggregated into
the next larger such grouping, such aggregation is within the scope of
aggregating a phase-out profile for each group member. Finally, fitting the
data for a particular category onto one of the geometric curves requires
regression which is also a form of aggregation.
As to the contention that “the demand of forecast for the phased-out
product depends on the individual phase-out profiles of each member of the
group,” this is simply an inherent result of such aggregation. As to the
contention that “there is no reasonable interpretation where forecasting
demand for the phase-out product is based on the group phase-out profile
Appeal 2013-003374
Application 12/637,593

9

(defined as the aggregation of phase-out profiles for each group member)”
(App. Br. 9), this is not what is recited. Instead, the claim recites generating
the group phase-out profile by sorting products into groups and aggregating
a phase-out profile for each group member. Every individual product,
product genus, and product category and even geometric description genus is
a member of a group of phased-out products. Further, the limitation at issue
does not recite or narrow how the generation results from aggregating a
phase-out profile for each group member. In particular, the generation is not
necessarily itself an aggregation function; the generation merely in some
unspecified manner relies on some aggregation. The Examiner’s finding
that curve fitting results from ordinary regression is within such a scope.
As to separately argued claims 4 and 11, reciting retrieving a standard
deviation value for the phased out product, the Examiner found that Moore
described doing so for computing stock replenishment quantities. We are
not persuaded by the Appellants’ argument that the claim requires retrieval
of a standard deviate of a phase out profile. Claim 4 recites retrieving a
standard deviation for a product, not of a profile. Clearly retrieving a
standard deviation for the purpose of stock replenishment is within the scope
of being for the replenished phased out product.
As to separately argued claim 9, reciting generating the profile if a
minimum number of data requirements are met, the Examiner found that
Moore described setting a minimum of 1. Ans. 8–9. We are not persuaded
by the Appellants’ argument that the Examiner failed to make a finding.
Appeal 2013-003374
Application 12/637,593

10

We are persuaded by the Appellants’ argument that the Examiner made
no findings as to claims 12, 13, and 15–18. Thus, the Examiner has not
presented a prima facie case as to these claims.
The remaining claims are argued on the basis of claim 1.
CONCLUSIONS OF LAW
The rejection of claims 1–18 under obviousness type double patenting is
proper.
The rejection of claims 1–11 and 14 under 35 U.S.C. § 103(a) as
unpatentable over Moore and Official Notice is proper.
The rejection of claims 12, 13, and 15–18 under 35 U.S.C. § 103(a) as
unpatentable over Moore and Official Notice is improper.
DECISION
The rejection of claims 1–18 is affirmed.
No time period for taking any subsequent action in connection with this
appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R.
§ 1.136(a)(1)(iv) (2011).

AFFIRMED

tkl