UNITED STATES PATENT AND TRADEMARK OFFICE
`
`
`
`
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`
`
`
`
`FORD MOTOR COMPANY
`Petitioner
`
`v.
`
`TRILOGY DEVELOPMENT GROUP, INC.
`Patent Owner
`
`
`
`
`Case IPR2016-01012
`Patent 7,882,057
`
`
`
`
`
`PATENT OWNER TRILOGY DEVELOPMENT GROUP, INC.’S
`
`PRELIMINARY RESPONSE
`
`
`
`Mail Stop “PATENT BOARD”
`Patent Trial and Appeal Board
`U.S. Patent and Trademark Office
`P.O. Box 1450
`Alexandria, VA 22313-1450
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`
`TABLE OF CONTENTS
`
`
`
`I. 
`
`II. 
`
`III. 
`
`B. 
`
`C. 
`
`Introduction ...................................................................................................... 1 
`A. 
`Trilogy is a pioneer in configuration technology, and the
`’057 patent represents a significant advance. ........................................ 2 
`Overview of the ’057 Patent .................................................................. 4 
`B. 
`Overview of Loomans ......................................................................... 10 
`C. 
`D.  Overview of Stahl ................................................................................ 13 
`E. 
`Summary of Argument ........................................................................ 14 
`Ford does not show that Stahl was published. ............................................... 16 
`A. 
`Ford is unclear how it is alleging that Stahl was
`published. ............................................................................................ 17 
`Ford’s evidence of publication by virtue of Stahl’s
`supposed distribution at a conference proceeding is
`insufficient. .......................................................................................... 18 
`Ford’s evidence of publication by virtue of Stahl’s
`supposed printing in the conference’s Proceedings is
`insufficient. .......................................................................................... 21 
`Ford does not show that the cited references suggest every
`feature of the independent claims. ................................................................. 23 
`A. 
`Ford does not show that the cited references teach
`“dividing one or more configuration queries” such that
`“the multiple configuration sub-queries represent the one
`or more configuration queries” as recited in every
`independent claim. ............................................................................... 25 
`1. 
`Ford fails to explain how Stahl’s “interpretation”
`allegedly discloses the claimed “dividing.” .............................. 27 
`Stahl does not disclose dividing such that “the
`multiple configuration sub-queries represent the
`one or more configuration queries” as claimed. ....................... 30 
`Ford provides no reason explaining why a skilled
`artisan would incorporate Stahl’s “interpretation”
`into Loomans’ system. .............................................................. 34 
`Ford does not show that the cited references teach
`“processing of each sub-query using at least one
`
`2. 
`
`3. 
`
`B. 
`
`
`
`- i -
`
`

`
`
`
`
`
`C. 
`
`D. 
`
`B. 
`
`C. 
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`configuration sub-model per sub-query” as recited in
`every independent claim. ..................................................................... 47 
`Ford does not show that the cited references teach that
`“each configuration sub-model includes data to define
`compatibility relationships between parts” as recited in
`every independent claim. ..................................................................... 36 
`Ford does not show that the cited references teach that
`“each configuration sub-model collectively models the
`configurable product” as recited in every independent
`claim. ................................................................................................... 43 
`IV.  Ford does not show that the cited references suggest every
`feature of the dependent claims. .................................................................... 48 
`A. 
`Ford does not show that the cited references teach “using
`different configuration sub-models until a configuration
`validation answer can be determined” as recited in claims
`4, 20, and 34. ....................................................................................... 49 
`Ford does not show that the cited references teach that
`“at least two sub-queries include overlapping
`information” as recited in claims 6, 22, and 36................................... 52 
`Ford does not show that the cited references teach claims
`7, 8, 23, 24, 37 and 38. ........................................................................ 55 
`1. 
`Ford does not show that the cited references teach
`“dividing a consolidated configuration model . . .
`in accordance a predetermined data structure.” ........................ 55 
`Ford does not show that the cited references teach
`“dividing the sub-queries in accordance with the
`sub-model structure.” ................................................................ 57 
`Ford’s obviousness analysis is inadequate, because Ford never
`ascertains the differences between the references and the
`claims. ............................................................................................................ 58 
`VI.  Conclusion ..................................................................................................... 59 
`
`
`
`2. 
`
`V. 
`
`
`
`- ii -
`
`

`
`
`
`
`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`
`EXHIBIT LIST
`
`Description
`Excerpts from Microsoft Dictionary of Computer Terms, Fifth
`Edition (2002).
`Excerpts from IBM Dictionary of Computer Terms, Tenth
`Edition (1994).
`Excerpts from Webster’s New World Dictionary of Computer
`Terms, Sixth Edition (1997).
`Versata website, “About Us”, available at
`http://cpq.versata.com/about-us.
`Ford Starts Firm to Manage Its Web Sites, ComputerWorld,
`Feb. 28, 2008, available at
`http://www.computerworld.com/article/2592825/it-
`management/ford-starts-firm-to-manage-its-web-sites.html.
`McCartney, Laton, “Trilogy Making A Name For Itself,”
`ZDNet, July 28, 2000.
`Versata Software Inc. Company Report, May 18, 2016, via
`Thomson Reuters.
`Field, Tom, Suit Yourself, CIO Magazine, Apr. 15, 1997, Vol.
`10, p. 108.
`“Ford and Trilogy Launch Web Company - Information
`Week,” InformationWeek News, February 23, 2000, available
`at http://www.informationweek.com/ford-and-trilogy-launch-
`web-company/d/d-id/1008183.
`Plaintiff’s Original Complaint for Damages and Injunctive
`Relief, filed in Versata Development Group v. Ford Motor
`Company, Case No. 4:15-cv-00316, (TXED), filed May 7,
`2015; 32 pages.
`
`Exhibit No.
`2001
`
`2002
`
`2003
`
`2004
`
`2005
`
`2006
`
`2007
`
`2008
`
`2009
`
`2010
`
`
`
`- iii -
`
`

`
`
`
`I.
`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`
`Introduction
`Patent Owner Trilogy Development Group, Inc.1 and Petitioner Ford Motor
`
`Company had 19-year business relationship. During most of this time, Ford
`
`licensed Trilogy’s patented Automotive Configuration Management (ACM)
`
`software and used this software as the backbone of its worldwide data
`
`infrastructure. (Ex. 2010, Complaint, pp. 1-2, in Versata Development Group, Inc.
`
`et al. v. Ford Motor Co., (E.D. Tex. filed May 7, 2015).) In 2011, to avoid paying
`
`licensing fees, Ford began a project to replace ACM with internally-developed
`
`software. Trilogy contends that this replacement software is a copy of ACM. (See
`
`generally Ex. 2010.)
`
`Now, Ford requests inter partes review on Trilogy’s U.S. Patent No.
`
`7,882,057 (“the ’057 patent”). The ’057 patent relates to database techniques to
`
`describe different ways that a product, such as an automobile, can be configured.
`
`This Petition is one of two that Ford filed on the ’057 patent, each directed to a
`
`
`1 Trilogy, Inc. acquired Versata, Inc. in 2006, at which point portions of Trilogy
`
`were merged with Versata, Inc. as part of a reorganization. (See Ex. 2007, Versata
`
`Company Report.) For convenience, herein Patent Owner refers to both companies
`
`as Trilogy.
`
`
`
`- 1 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`different group of claims. In this Petition, Ford alleges obviousness of the ’057
`
`
`
`patent’s independent claims 1-16, 18-29, and 31-43 (of which 1, 18, and 31 are
`
`independent) over two references: Loomans (Ex. 1005) and Stahl (Ex. 1006).
`
`As will be addressed in detail below, the Petition should be denied for three
`
`reasons. First, Ford does not show that Stahl is prior art. Second, Ford does not
`
`show that each and every claim element is taught in the art. The cited art lacks
`
`several key claim features, including a feature that the ’057 patent’s original
`
`examiner found to be allowable and a feature that the ’057 specification touts as
`
`driving a significant improvement to the performance of database queries. Third,
`
`Ford’s analysis in its Petition omits a factor set forth in Graham v. John Deere Co.,
`
`(383 U.S. 1 (1966)), and thus is legally insufficient to support a conclusion of
`
`obviousness.
`
`The discussion below first provides some background on Trilogy and its
`
`relationship to Ford, then provides a tutorial on the ’057 patent and prior art, and
`
`finally provides a more detailed summary of the arguments in this Preliminary
`
`Response.
`
`A. Trilogy is a pioneer in configuration technology, and the ’057
`patent represents a significant advance.
`
`Trilogy produces enterprise-level software solutions that, among other
`
`things, help large corporations manage sales of complicated products. (Ex. 2004,
`
`“Versata, About Us.”) One of Trilogy’s most innovative areas of focus is product
`
`
`
`- 2 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`configuration software. (See Ex. 2005, ComputerWorld, p. 2 (Ford press releasing
`
`
`
`describing a joint venture with Trilogy that included “car configuration”).)
`
`Although Trilogy has marketed product configuration solutions in a variety of
`
`industries, this software has been particularly useful in the automotive industry.
`
`(See Ex. 2006, McCartney; see Ex. 2008, Field.)
`
`Trilogy’s configuration technologies are particularly suited to automotive
`
`manufacturing because every vehicle is a complex system involving thousands of
`
`parts, features, and options. Complex software like Trilogy’s can enable
`
`automotive manufacturers to manage the billions of potential permutations that
`
`result from this complexity.
`
`Between 1998 and 2014, Trilogy provided configuration software to Ford.
`
`These solutions included software to support Ford’s online buying site. (Ex. 2006,
`
`“Information Week.”). These solutions also included ACM. Ford used ACM to
`
`configure vehicles and to provide product configuration data to engineering,
`
`manufacturing, sales, marketing, and ordering software that Ford uses around the
`
`world. Before ACM, configuration errors caused Ford substantial waste. ACM
`
`solved these problems.
`
`ACM could not just adapt its pre-existing product configuration solutions to
`
`meet the automotive industry’s needs, because the automotive industry faced
`
`problems that were not present in other industries, such as the personal computer
`
`
`
`- 3 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`configuration described in Loomans and Stahl. (See Ex. 2005.) Similarly, academia
`
`
`
`approached configuration problems differently from the automotive industry, as
`
`illustrated by Stahl. Academia focused on approaches not constrained by the
`
`business processes of the automotive industry; namely, academia depended on the
`
`ability to discern and express why part relationships at a higher level (e.g., attribute
`
`values of those parts and conditions thereupon) were compatible and incompatible.
`
`In contrast, the automotive industry focused on approaches that depended on the
`
`enumeration of every single part relationship. To solve the automotive industry’s
`
`product configuration problems, Trilogy developed the technology that is
`
`described and claimed in the ’057 patent.
`
`B. Overview of the ’057 Patent
`The ’057 patent relates to specifying ways that a product, such as an
`
`automobile, can be configured. The different options available on a product are
`
`described in what the ’057 patent refers to as a configuration model. “A
`
`configuration model,” the ’057 patent explains, “uses, for example, data, rules,
`
`and/or constraints . . . to define compatibility relationships between parts . . .
`
`contained in a specific type of product.” (Ex. 1001, “the ’057 patent,” 1:26-30.) In
`
`accordance with the rules and constraints, “[p]arts may be combined in different
`
`ways . . . to define different instances of” the overall product. (Id., 1:31-33.) “For
`
`example, ‘V6 engine’ or the exterior color ‘red’ can be parts on a vehicle.” (Id.,
`
`
`
`- 4 -
`
`

`
`
`
`
`
`Case IPRR2016-010112 of
`U.S. Pat.
`
`No. 7,8822,057
`
`
`
`1:33-355.) A vehiccle configuuration moodel can iindicate thhat “red” iis the stanndard
`
`
`
`
`
`
`
`
`
`
`
`
`
`color feeature for
`
`a specific
`
`
`
`vehicle, bbut that thee color redd is not coompatible
`
`
`
`
`
`
`
`
`
`“V6 enngine.” Too ask “a qquestion .
`
`
`
`
`
`
`
`. . about t the partss and relaationships
`
`
`
`
`
`
`
`configuuration moddel,” the ’0057 patentt describess using a cconfiguratiion query.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`1:47-488.)
`
`with
`
`in a
`
`(Id.,
`
`
`
`FFor example, the ’057 patent’’s Table 22, reproduuced beloww, illustrattes a
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`simple
`
`
`
`configurattion modell and assoociated queeries. Tab
`
`
`
`
`
`
`
`le 2 lists
`
`
`
`different pparts
`
`product,
`
`
`
`enumerateed as A1,, A2, B1,
`
`
`
`B2, X1,
`
`and X2,
`
`and
`
`
`
`availablle on the
`
`
`
`describees how thee parts are ccompatiblee with one
`
`
`
`
`
`
`
`
`
`another. ““‘S’ repres
`
`
`
`ents ‘standdard’
`
`
`
`and ‘O’ represennts ‘optionnal.’” (Id.,
`
`
`
`
`
`2:15-16;
`
`7:51-66.)
`
`
`
`Table 2 aalso illustrrates
`
`
`
`
`
`examplee configuration queries, such aas “Are A11, B1, and
`
`
`
`
`
`
`
`
`
`
`
`X1 buildaable togethher?”
`
`
`
`
`
`(Id., 7:551-66.)
`
`
`
`
`
`“Solving coonfiguratioon problemms,” the ’0557 patent exxplains, “ooften requirres a
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`signific
`
`
`
`
`ant amounnt of data pprocessing
`
`
`
`capabilitiees.” (Id., 22:38-40.) Inn particulaar, as
`
`
`
`
`
`
`
`
`
`configuuration moodels and
`
`
`
`
`
`queries beecome moore compllex, the daata processsing
`
`
`
`
`
`
`
`
`
`
`
`- 5 -
`
`

`
`
`
`
`
`Case IPRR2016-010112 of
`U.S. Pat.
`
`No. 7,8822,057
`
`
`
`capabiliities needeed to solvee the queriies in a reeasonable aamount off time incrrease
`
`
`
`
`
`
`
`
`
`
`
`
`
`exponenntially, deppicted by eexponentiaal performaance curvee 302 in thee ’057 pateent’s
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`FIG. 3,, reproducced beloww. In FIG. 3, “[t]hee dashed
`
`
`
`
`
`
`
`
`
`
`
`maximuum data pprocessing
`
`
`
`
`
`capabilityy of the pparticular
`
`
`
`used.”
`
`
`
`(Id., 3:3-55.) Thus,
`
`
`
`the compputer systeem could
`
`
`
`
`
`reasonablly “processs []
`
`
`
`
`
`configuuration moodel[s] annd configguration qquer[ies] hhaving
`
`
`
`
`
`
`
`
`
`thhe compleexity
`
`
`
`line 304
`
`
`
`representss the
`
`computer
`
`
`
`system bbeing
`
`
`
`
`
`
`
`represennted by pooint A,” buut “could nnot reasonnably proceess configuuration moodels
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`and connfigurationn queries hhaving a coomplexity
`
`
`
`
`
`
`
`
`
`representeed by pointt B.” (Id.,
`
`
`
`3:1-
`
`8.)
`
`
`
`TThe ’057
`
`
`
`patent’s iinventors
`
`
`
`did not ssolve the
`
`problem
`
`
`
`of processsing
`
`
`
`configuuration moodels and
`
`
`
`
`
`configuraation queriies at poiint B, forr example
`
`
`
`
`
`
`
`, by
`
`
`
`
`
`
`
`
`
`reducing the commplexity oof the ovverall conffiguration
`
`
`
`
`
`system. ((Id., 4:27--29.)
`
`
`
`
`
`- 6 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`Instead, the ’057 patent’s inventors solved this problem by breaking the complex
`
`
`
`model and query into smaller components that could, for example, both be
`
`processed in parallel and would use less memory to arrive at an answer. Thus, the
`
`’057 patent describes “generat[ing] configuration sub-models on the side of the
`
`achievable [exponential] performance curve.” (Id., 4:27-31.) This generation of
`
`configuration sub-models involves “breaking a configuration problem down into a
`
`set of smaller problems, solving them individually and recombining the results into
`
`a single result that is equivalent to [the result of] a conventional [configuration]
`
`interference procedure.” (Id., 4:22-25.) The ’057 patent describes dividing both a
`
`configuration model into sub-models and dividing a configuration query into sub-
`
`queries with respect to its FIG. 4.
`
`
`
`- 7 -
`
`

`
`
`
`
`
`Case IPRR2016-010112 of
`U.S. Pat.
`
`No. 7,8822,057
`
`
`
`
`
`
`
`
`
`Inn FIG. 4, tthe ’057 ppatent illusttrates a “suub-model
`
`
`
`
`
`inference
`
`procedure
`
` 402
`
`
`
`[that] inncludes operations 4004, 406, 4008, and 41
`
`
`
`
`
`
`
`0,” as illu
`
`
`strated bel
`
`low. (Id., 44:46-
`
`
`
`48.) “OOperation
`
`
`
`404 receives, as
`
`an
`
`
`
`inpuut, a connventional,, consoliddated
`
`
`
`
`
`
`
`
`
`configuuration model 412 annd dividess the conssolidated cconfiguratiion model
`
`
`
`
`
`
`
`
`
`
`
`
`
`412
`
`into a
`
`set of
`
`
`
`configuraation subb-models.”
`
`
`
`
`
`(Id., 4::54-57.) FFor exammple,
`
`
`
`
`
`
`
`“operatiion 404 diivides the
`
`
`
`
`
`conventionnal model
`
`
`
`[illustratedd in Tablee 2 above]
`
`
`
`into
`
`the . . .
`
`
`
`configurattion sub-mmodels reprresented inn Table 3[,,] Table 4,, and Tablee 5,”
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`reproduuced beloww. (Id., 8:2--4.)
`
`
`
`
`
`
`
`- 8 -
`
`

`
`
`
`
`
`Case IPRR2016-010112 of
`U.S. Pat.
`
`No. 7,8822,057
`
`
`
`
`
`
`
`WWith the cconsolidateed model
`
`divided
`
`
`
`into sub-mmodels, ““operation
`
`
`
`divides
`
`queries
`
`
`
`
`
`
`
`
`
`ormation tenough infor contain ech togetherQ1 through Qn, whic
`
`
`
`
`
`o represennt the
`
`
`
`originall query 4114.” (Id., 55:24-27.) OOperation
`
`
`
`
`
`
`
`esses the
`408 proce
`
`
`
`sub-queriees to
`
`
`
`generatee sub-answwers. (Id.,
`
`
`
`6:16-18.)
`
`
`
`
`
`
`
`Continuinng the exammple in thhe ’057 paatent,
`
`
`
`
`
`sub-queeries generrated by opperation 4406 and suub-answerss generatedd by operaation
`
`
`
`
`
`
`
`
`
`
`
`
`
`the conveentional, coonsolidatedd configurration querry 414 intoo a set of
`
`
`
`
`
`
`
`
`
`
`
`
`
`406
`
`sub-
`
`
`
`
`
`
`
`408 are illustratedd in Tables 6 and 7, reeproduced
`
`
`
`
`
`
`
`above.
`
`
`
`
`
`
`
`
`
`
`
`FFinally, witth the sub-answers deetermined,, the ’057 ppatent commbined the
`
`
`
`
`
`sub-
`
`
`
`answerss together
`
`
`
`into a colllective annswer at ooperation 4410. (Id.,
`
`
`
`
`
`
`
`10:16-17.)
`
` For
`
`
`
`some tyypes of quueries, “commbination
`
`
`
`
`
`operation
`
`
`
`410 invollves inters
`
`ecting the
`
`. . .
`
`
`
`buildable spaces
`
`
`
`[describedd in the suub-answerss] togethe
`
`
`
`
`
`
`
`r into onee space” uusing
`
`
`
`
`
`overlappping informmation bettween the qqueries. (Idd., 10:18-220.)
`
`
`
`
`
`
`
`
`
`
`
`
`
`- 9 -
`
`

`
`
`
`
`
`Inn this wayy, the ’057
`
`
`
`
`
`Case IPRR2016-010112 of
`U.S. Pat.
`
`No. 7,8822,057
`
`
`
` patent divvides the cconfiguratiion problemm into smmaller
`
`
`
`
`
`
`
`pieces tthat can bee more easily processsed. The smmaller pieeces “do[] nnot changee the
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`exponenntial nature of configguration mmodel and
`
`
`
`
`
`
`
`
`
`query commplexity.”
`
`
`
`(Id., 5:18--21.)
`
`But, as
`
`
`
`
`
`
`
`
`
`
`
`illustratedd in FIG. 55, by breakking into ssmaller pieeces, a queery that wwould
`
`
`
`
`
`
`
`otherwiise be too
`
`
`
`computatioonally exppensive to
`
`
`
`
`
`execute onn the origi
`
`nal model
`
`in a
`
`
`
`reasonaable time aand/or withh availablee system mmemory mmay be divvided, and
`
`
`
`
`
`
`
`
`
`
`
`
`
`sub-
`
`queries
`
`
`
`executed oon the sub
`
`
`
`-models. (IId.) Thus,
`
`
`
`the ’057 ppatent “proovides a waay to
`
`
`
`
`
`
`
`scale quueries to laarger and mmore compllicated connfigurationn models.”
`
`
`
`
`
`
`
`
`
`
`
`
`
`(Id., 5:22--23.)
`
`
`
`
`
`C. Overrview of LLoomans
`
`C T
`
`
`
`
`
`
`
`The princippal reference applieed in Forrd’s grounnd of unppatentabilitty is
`
`
`
`
`
`
`
`
`
`
`
`Loomanns. Loomanns is a U.SS. patent thhat describbes modelinng configuuration of wwhat
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`- 10 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`it calls a complicated entity, such as a computer system. (Ex. 1005, 3:7-53.) The
`
`
`
`complicated entity is modeled using what Loomans calls sub-configuration. Such
`
`“[s]ub-configuration . . . partitions [the] complicated top-level entity (e.g., [a]
`
`computer system []) into a set of configurations of smaller sub-level entities (e.g.,
`
`Drive Bays 1, 2, and 3).” (Id., 4:50-53.) “The top-level entity may be represented
`
`by . . . a parent model, and each [sub-level entity] may be represented by . . . a
`
`child sub-model.” (Id., 1:62-65.) “The sub-models reduce the amount of
`
`redundancy by localizing data and logic needed to represent the shared options.”
`
`(Id., 2:6-8.) This desire to reduce the amount of redundancy is the opposite of
`
`what is taught in the ’057 patent where redundant overlapping data exists in the
`
`sub-queries and sub-models.
`
`To map the parent model to a child sub-model, Loomans uses what it calls a
`
`sub-model mapping set data structure, which it illustrates, for example, as element
`
`410 in its FIG. 4, reproduced below. (Id., 7:44-47.) In FIG. 4, sub-model mapping
`
`set 420 includes an option that represents storage. That option maps to two
`
`available types of storage devices: ABC and XYZ. (Id., 7:54-57.) The ABC and
`
`XYZ storage devices are each represented by sub-models including two mapped
`
`features for the available storage amounts and corresponding power ratings. (Id.,
`
`7:57-66.)
`
`
`
`- 11 -
`
`

`
`
`
`
`
`Case IPRR2016-010112 of
`U.S. Pat.
`
`No. 7,8822,057
`
`
`
`
`
`
`
`TThe configguration mmodel can
`
`
`
`be used
`
`
`
`to determmine whet
`
`
`
`her a prooduct
`
`
`
`configuuration is vvalid. Forr example,, Loomanss describees how its
`
`
`
`
`
`
`
`
`
`
`
`
`
` configuraation
`
`
`
`model is queried tto determinne which storage devvices are coompatible wwith a seleected
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`power ssupply. In tthe examplle, “a valuee (e.g., 40
`
`
`
`
`
`
`
`Amp) . . .
`
`
`
`has been sselected fo
`
`r the
`
`
`
`[p]owerr [s]upplyy option
`
`
`
`
`
`G at thee parent
`
`model.”
`
`
`
`(Id., 8:533-55.) Duuring
`
`
`
`
`
`
`
`
`
`configuuration of tthe ABC ssub-model, Loomanss describess how thatt value forr the
`
`
`
`
`
`
`
`
`
`
`
`power
`
`
`
`supply neeeds to bee passed ddown to thhe ABC ssub-modell to determmine
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`whetherr the value satisfies thhe sub-moodel’s cons
`
`
`
`
`
`
`
`
`
`
`
`
`
`traints. (Idd.) “If a parrticular stoorage
`
`amount
`
`
`
`is selectted (e.g.,
`
`40 GB),
`
`
`
`then thee correspoonding poower ratinng is
`
`
`
`
`
`
`
`
`
`60 Amp)).” (Id., 8:56-58.) Thhe correspponding p
`
`
`
`
`
`
`
`
`
`ower ratinng is
`
`
`
`
`
`
`
`
`
`the selecteed power ssupply in thhe power ssupply subb-model, wwhich
`
`
`
`
`
`
`
`determined (e.g.,
`
`ed against
`compar
`
`
`
`- 12 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`in this case is 40 Amps. The power supply’s 40 Amps are inadequate to support a
`
`
`
`60 Amp storage device. (Id., 8:59-60.) Thus, Loomans determines that the 40 GB
`
`storage is not buildable with the 40 Amp power supply. (Id., 8:60-63.)
`
`In this way, Loomans describes modeling a product into a hierarchy
`
`involving a parent model and child sub-models, and traversing the hierarchy to
`
`determine whether a configuration is valid.
`
`D. Overview of Stahl
`Stahl is a non-patent document that Ford alleges was published at a
`
`conference that took place in Bled, Slovenia in June 2000. Ford relies on Stahl for
`
`the claim features related to dividing a query into sub-queries.
`
`Stahl describes a product customization approach involving two steps. (Ex.
`
`1006, p. 4.) In the first step, Stahl represents a query as a tree, as illustrated in its
`
`FIG. 3, reproduced below. Stahl answers this query by interpreting the root node as
`
`the actual problem. In FIG. 3, the root node in the query searches for a PC with a
`
`price of 2000 Deutsche Marks. (See id., pp. 4-5.) In addition to interpreting the root
`
`node of the query tree, Stahl “interpret[s] [] different leaf nodes of the [same]
`
`query as sub-problems that must be solved to solve the overall problem, i.e., the
`
`configuration of the required PC.” (Id., p. 5.) Dependencies may exist between
`
`sub-problems due to technical constraints between components. (Id.) Using a tree
`
`query structure, Stahl retrieves a product from a product database that is similar to
`
`
`
`- 13 -
`
`

`
`
`
`
`the one
`
`
`
` requestedd. (Id.) In tthis way,
`
`
`
`Case IPRR2016-010112 of
`U.S. Pat.
`
`No. 7,8822,057
`
`
`
`al of sed retrievamilarity-basStahl perfoforms “sim
`
`
`
`
`
`the bestt available base produuct.” (Id.)
`
`
`
`
`
`
`
`(IId., Fig. 3.))
`
`
`
`
`
`OOnce the baase producct is retrievved, Stahl’’s second
`
`
`
`
`
`
`
`
`
`
`
`productt using what it calls an adaaptation cyycle. The
`
`
`
`
`
`
`
`
`
`
`
`
`
`step custommizes the
`
`base
`
`
`
`adaptationn cycle iss an
`
`
`
`
`
`“iterativve proceduure which pperforms thhe necessaary adaptatiion of the rretrieved bbase-
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`productt if it does nnot fulfill aall customeer demandds.” (Id., p. 5.)
`
`
`
`
`
`
`
`
`
`
`
`
`
`Summary of AArgument
`
`
`
`that
`
`
`
`four reasoons. First,
`
`
`
`Ford failss to show
`
`E.
`
`
`E F
`
`
`
`Ford’s Petition is defficient for
`
`
`
`Stahl wwas publishhed, let alonne that Staahl is a prinnted publiccation undder 35 U.S..C. §
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`312(a)(33)(A) and
`
`
`
`associatedd rules. Sommetimes FFord seemss to allege
`
`
`
`
`
`
`
`that Stahl
`
`was
`
`
`
`publisheed by distrribution at
`
`
`
`
`
`a conferennce in Sloovenia, andd other timmes Ford seeems
`
`
`
`
`
`
`
`
`
`
`
`to allegge that Staahl was puublished inn the confference’s PProceedinggs book. FFord
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`providees no evidence to suppport eitherr allegationn.
`
`
`
`
`
`
`
`
`
`
`
`- 14 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`Second, Ford does not establish that Stahl and Loomans teach or suggest key
`
`
`
`claim features related to the ’057 patent’s sub-queries and sub-models as recited in
`
`the independent claims. A key to the claimed invention is “dividing” the query into
`
`sub-queries. That dividing feature is what persuaded the examiner to allow the
`
`’057 patent during the original prosecution. For this feature, Ford points to Stahl’s
`
`“interpretation” of a query. But interpretation is not the same as dividing, and, even
`
`it were, the analysis from Dr. Greenspun—Ford’s own expert—shows that the
`
`alleged divided sub-queries still do not represent the original input configuration
`
`query as claimed. Moreover, Stahl does not process each alleged sub-query “using
`
`at least one configuration sub-model per sub-query” as claimed. Instead, Stahl
`
`requires that additional logic be present outside of the alleged sub-queries in order
`
`to process them. For the sub-models, Ford heavily relies on Loomans. But because
`
`Loomans’ sub-models at best only describe a part’s features and its compatibility
`
`with the overall product, Loomans’ sub-models do not “define compatibility
`
`relationships between parts” as claimed. Still further, the claimed “compatibility
`
`relationships between parts” is merely a list, without requiring a hierarchical
`
`structure. In contrast, Loomans determines whether parts are compatible by
`
`modeling a product into a hierarchy involving a parent model and child sub-
`
`models, and traversing the hierarchy to determine whether a configuration is valid.
`
`Because Loomans vests some of the product configuration with a parent model and
`
`
`
`- 15 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`not with its sub-model, Loomans’ sub-models do not “collectively model[] the
`
`
`
`configurable product” as claimed.
`
`Third, Ford’s theories are lacking not only for the independent claims, but
`
`also for the dependent claims. For example, Ford fails to establish that the
`
`references render obvious certain claim features relating
`
`to overlapping
`
`information (e.g., claim 5) and sub-model structures (e.g., claim 7).
`
`Fourth and finally, Ford fails to conduct the requisite factual inquiries in any
`
`obviousness analysis. In particular, Ford fails to ascertain the differences between
`
`the prior art and the claims, leaving Patent Owner and the Board to guess at its
`
`theories.
`
`These four points are addressed in the four argument sections (Sections II,
`
`III, IV, and V) below.
`
`II.
`
`Ford does not show that Stahl was published.
`
`Because the Petition does not satisfy the relevant statute and rules requiring
`
`a petitioner to identify the printed publication relied upon, the Board need go no
`
`further than this Section II to deny the Petition. The PTO rules require that the
`
`petition must identify copies of “patents and printed publications relied upon for
`
`each ground.” 37 C.F.R. § 42.104(b)(2). The Petition does not live up this rule,
`
`failing to identify how Stahl was published and failing to provide the requisite
`
`
`
`- 16 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`evidence to show that Stahl is a printed publication within the meaning of 35
`
`
`
`U.S.C. § 103.
`
`The Petition’s reliance on a few incomplete and contradictory citations to
`
`Stahl shows that Ford did not clearly identify the printed publication that Ford
`
`claims Stahl is a copy of as required by the statute and rules. Ford sometimes
`
`points to a Slovenian conference and sometimes points to a Proceedings title for
`
`the conference. Whether Ford is relying on distribution at the conference or in the
`
`Proceedings title, Ford fails to provide any evidence showing that either was a
`
`printed publication. Each point is addressed in turn.
`
`Ford is unclear how it is alleging that Stahl was published.
`
`A.
`The IPR statute states that a petition may only be considered if “the petition
`
`identifies . . . copies of patents and printed publications that the petitioner relies
`
`upon.” 35 U.S.C. § 312(a)(3)(A). Implementing the statute, the PTO rules provide
`
`that the petition must identify “the patents or printed publications relied upon.” 37
`
`C.F.R. § 42.104(b)(2). Here, Ford relies on a document marked Ex. 1006, referred
`
`to as Stahl, but the Petition is unclear as to how Ford is alleging Stahl was
`
`published. In other words, Ford does not clearly identify the printed publication
`
`that Ford claims Stahl is a copy of, as required by the statute and rules.
`
`For example, Ford’s exhibit list cites to the 13th International Bled
`
`Electronic Commerce Conference for Ex. 1006, and, at the end of Ex. 1006, Ford
`
`
`
`- 17 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`attached what appears to be an associated conference pamphlet. But, after the
`
`
`
`citation to the conference, the exhibit list has a parenthetical: “(available at
`
`_____________.)” (Pet., p. iii.) That parenthetical suggests that Ford was perhaps
`
`not relying on the conference presentation as Stahl’s publication, but some
`
`unspecified source where the document was available.
`
`Moreover, in the Petition, Ford claims: “Stahl was published in the
`
`Proceedings of the 13th International Bled Electronic Commerce Conference in
`
`2000.” (Pet., p. 1.) This suggests that Ford was relying not on the conference itself,
`
`but a Proceedings book that Ford claims was published at some unspecified time in
`
`2000. Thus, contrary to the relevant statute and rules, Ford does not clearly identify
`
`the printed publication that Ford claims Stahl is a copy of, and the Petition should
`
`be denied for that reason alone.
`
`B.
`
`Ford’s evidence of publication by virtue of Stahl’s supposed
`distribution at a conference proceeding is insufficient.
`
`Though not entirely clear from the Petition, Ford may be alleging that (1)
`
`Ex. 1006 to Stahl was available at a conference called the 13th International Bled
`
`Electronic Commerce Conference in Bled, Slovenia on June 19-21, 2000, and (2)
`
`that its availability at the conference makes Ex. 1006 a printed publication. If that
`
`is what Ford is alleging, Ford lacks sufficient evidence on both counts.
`
`With respect to the first allegation, Ford provides no testimony that Ex. 1006
`
`was available at the alleged Slovenian conference. Moreover, Ex. 1006 on its face
`
`
`
`- 18 -
`
`

`
`
`
`Case IPR2016-01012 of
`U.S. Pat. No. 7,882,057
`appears to include two separate documents. In particular, pages 1-9 appear to be a
`
`
`
`paper, concluding with a bibliography. But starting at page 10, the page numbering
`
`and headings that were present earlier vanish and what appears be a conference
`
`program begins. There is no mention at all of the conference program in the
`
`Petition. Moreover, Ford does not tie this conference program to the paper, or
`
`make any claims about what the program shows. As a result, Ford fails to show
`
`that the paper itself was actually presented at the conference.
`
`This failure to prove that Ex. 1006 was presented at the conference is fatal.
`
`For example, in Norian Corp. v. Stryker Corp., the Federal Circuit held that a
`
`document was not a printed publication because there was no evidence that the
`
`document was actually available at the conference, even though general practice
`
`was for presenters at the conference to give abstracts to interested attendees. 363
`
`F.3d 1321, 1330 (Fed. Cir. 2004). Indeed, the Board has held that evidence that a
`
`presentation occurred