Paper 25
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`______________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`______________
`
`
`APPLE, INC.,
`
`Petitioner
`
`v.
`
`QUALCOMM INCORPORATED,
`
`Patent Owner
`______________
`
`Case IPR2018-01282
`
`U.S. Patent No. 8,768,865
`______________
`
`PATENT OWNER’S SUR-REPLY TO PETITIONER’S REPLY
`
`
`98623272.1
`
`

`

`TABLE OF CONTENTS
`
`Case IPR2018-01282
`Patent 8,768,865
`
`Page
`
`I. 
`II. 
`
`3. 
`
`4. 
`
`5. 
`
`INTRODUCTION .......................................................................................... 1 
`CLAIM CONSTRUCTION ........................................................................... 4 
`A. 
`“Pattern” ............................................................................................... 4 
`B. 
`“Fixing … by Associating …” ............................................................. 6 
`Petitioner Cannot Justify Deleting the “Fixing” Phrase .......... 6 
`1. 
`2. 
`Qualcomm’s Construction Sets Forth the Meaning of
`“Fixing” Used by the Specification ........................................... 7 
`Petitioner Identifies No Legitimate Criticism of
`Qualcomm’s Construction ......................................................... 9 
`Subsequent Pattern Recognition Efforts May Indicate that
`Fixing Did Not Occur .............................................................. 12 
`Qualcomm’s Construction Does Not Contradict Its
`Litigation Positions .................................................................. 13 
`“Initiating a process to attempt a recognition of a second
`pattern…” ........................................................................................... 15 
`PETITIONER’S NEW “FIXING … BY ASSOCIATING …”
`THEORY RELIES ON FUNCTIONALITY LOUCH DOES NOT
`ACTUALLY DISCLOSE ............................................................................. 16 
`A. 
`Louch Does Not Define, Nor Would a POSITA Understand There
`to Be, Any Portion of a “Duration” Pattern as a Separate Pattern ..... 17 
`Even if a Portion of a “Duration” Pattern Were Deemed a First
`Pattern, Louch Does Not Disclose Setting a Scope of Analysis in
`Any Way ............................................................................................. 19 
`IV.  PETITIONER’S TWO “FIRST” AND “SECOND” PATTERN
`THEORIES FAIL AS INCONSISTENT WITH THE CLAIMS AND
`GROUNDED ON FUNCTIONALITY LOUCH DOES NOT
`ACTUALLY DISCLOSE ............................................................................. 19 
`A. 
`Petitioner’s “Learning Mode” Theory is Contrary to Dr. Allen’s
`Testimony and the Plain Language of the Challenged Claims .......... 20 
`
`III. 
`
`C. 
`
`B. 
`
`98623272.1
`
`- i -
`
`

`

`B. 
`
`V. 
`
`Case IPR2018-01282
`Patent 8,768,865
`Petitioner’s “Duration” Pattern Theory Relies on Functionality
`Louch Does Not Actually Disclose .................................................... 20 
`PETITIONER’S THEORY OF RECOGNIZING A SECOND
`PATTERN IN A REDUCED SET OF VARYING PARAMETERS
`RELIES ON FUNCTIONALITY LOUCH DOES NOT ACTUALLY
`DISCLOSE ................................................................................................... 21 
`VI.  PETITIONER’S THEORY OF CAPTURING SNAPSHOTS RELIES
`ON FUNCTIONALITY LOUCH DOES NOT ACTUALLY
`DISCLOSE, AND STILL IS NOT IN RESPONSE TO DETECTING
`A CONDITION ............................................................................................ 23 
`VII.  CONCLUSION ............................................................................................. 25 
`
`
`
`
`
`
`
`
`98623272.1
`
`- ii -
`
`

`

`TABLE OF AUTHORITIES
`
`Case IPR2018-01282
`Patent 8,768,865
`
` Page(s)
`
`Cases
`Hockerson-Halberstadt, Inc. v. Avia Group Int’l, Inc.,
`222 F.3d 951 (Fed. Circ. 2000) ........................................................................... 11
`In re Translogic Tech., Inc.,
`504 F.3d 1249 (Fed. Cir. 2007) .......................................................................... 10
`
`
`
`
`
`
`
`98623272.1
`
`- iii -
`
`

`

`PATENT OWNER’S EXHIBIT LIST
`
`Case IPR2018-01282
`Patent 8,768,865
`
`Description
`No.
`2001 U.S. Provisional Application No. 61/434,400 (Incorporated by Reference
`by Ex. 1001)
`
`2002
`
`2003
`
`2004
`
`omitted
`
`Transcript of Deposition of James Allen (April 25, 2019)
`
`omitted
`
`2005 Declaration of John Villasenor
`
`2006
`
`CV of John Villasenor
`
`2007 As-Filed Claims of U.S. Patent No. 8,676,224 to Louch (Ex. 1011)
`
`
`
`
`
`98623272.1
`
`- iv -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`
`I.
`
`INTRODUCTION
`Petitioner’s Reply does not—and cannot—dispute the key factual points that
`
`demonstrate that Petitioner’s construction of “fixing … by associating …” is wrong
`
`(and Qualcomm’s is correct). Moreover, for each of the four separate groups of
`
`claim elements that Qualcomm demonstrated Louch does not disclose, Petitioner’s
`
`Reply identifies no actual disclosure of Louch of the functionality Petitioner asserts
`
`meets the claims. Instead, Petitioner describes functionality that could have been
`
`designed, but Louch does not actually disclose.
`
`First, Petitioner does not dispute its own expert’s testimony that “the
`
`specification uses ‘associating’ as a substep of the process of fixing” and that
`
`“associating” can be used to “achieve many different goals.” Ex. 2003 at 49:25-
`
`50:9; 51:20-23. These facts, not in the record at the time of the Institution Decision,
`
`lead to the conclusion the recited “associating …” need not result in “fixing ….” As
`
`the Institution Decision explained, the plain claim language “limits the claim to
`
`fixing the parameters only by the action of associating.” Decision at 17 (emphasis
`
`omitted). Petitioner’s construction does not “limit[] the claims to fixing” and is,
`
`therefore, wrong.
`
`Thus, it is necessary to assign meaning to “fixing”—and Qualcomm is the
`
`only party offering any meaning. Petitioner does not—and cannot—identify any
`
`portion of the ’865 Patent Specification that is inconsistent with Qualcomm’s
`
`98623272.1
`
`- 1 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`construction. Petitioner instead repeatedly mischaracterizes Qualcomm’s Response
`
`as not providing sufficient cites to the specification—a contention easily refuted.
`
`The specification repeatedly explains that the system is able to identify patterns that
`
`co-occur with the fixed varying parameters as a result of the fixing. This goal is
`
`achieved by “setting the scope of analysis” as recited in Qualcomm’s construction
`
`and is not by merely “associating.”
`
`Second, to address both the “fixing … by associating …” under Qualcomm’s
`
`construction and the “second pattern” of dependent Claims 3, 22, and 27, Petitioner
`
`asserts that Louch’s disclosure of a “duration” pattern includes two separate patterns.
`
`But Petitioner concedes that in the embodiment of Louch in which a “duration”
`
`pattern is used, “the first pattern is not used to trigger a control action.” Reply at 22.
`
`In fact, Petitioner cites to no disclosure of Louch even mentioning any such “first
`
`pattern” or describing stages of recognizing a “duration” pattern. Moreover,
`
`Petitioner fails to address citations in Qualcomm’s Response from Louch, the ’865
`
`Patent, Dr. Villasenor, and Petitioner’s own expert, Dr. Allen, contradicting
`
`Petitioner’s arguments.
`
`Third, dependent Claims 4 and 23 each recite that “said second pattern is
`
`[recognized in] / [associated with] a reduced set of varying parameters. Petitioner
`
`does not—and cannot—identify any actual disclosure in Louch related to limiting
`
`the varying parameters in which a pattern, including any purported “second pattern,”
`
`98623272.1
`
`- 2 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`is recognized. See Response at 27-28. Petitioner instead describes functionality that
`
`could exist, but Louch does not actually disclose.
`
`Fourth, dependent Claims 5, 24, and 48 each recite “captur[ing] a snapshot …
`
`in response to said detection of said at least one condition.” Petitioner does not—
`
`and cannot—identify any actual disclosure in Louch related to taking snapshots, let
`
`alone in response to detection of a condition. See Response at 38-39. Instead,
`
`Petitioner argues that any sensor capture process necessarily includes taking
`
`snapshots “continuously.” Id. Even if Louch could be interpreted as including such
`
`functionality, Petitioner does not—and cannot—explain how continuously capturing
`
`snapshots could meet the plain claim language requiring capturing a snapshot “in
`
`response to said detection of said at least one condition.”
`
`Accordingly, for multiple independent reasons, the patentability of all
`
`Challenged Claims should be affirmed.
`
`98623272.1
`
`- 3 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`
`II. CLAIM CONSTRUCTION1
`A.
`“Pattern”
`Qualcomm seeks construction of this term solely to clarify that “identifying a
`
`pattern” includes linking varying parameters in the pattern to their corresponding
`value (such as Parameter “Sound Intensity” = Value “Loud” and
`Parameter “PeriodicMovement” = Value “Running”). Petitioner
`
`does not—and cannot—dispute this point. See Reply at 4-5; see also Ex. 2003 at
`
`56:20-23, 58:9-15 (Dr. Allen agreeing to same). While Qualcomm believes it is also
`
`accurate that a pattern includes the logical relationship between each pair (such as
`AND), the Board need not reach this dispute as it is not relevant to any issue to be
`
`decided.2
`
`
`1 Qualcomm notes for the Board’s convenience that Sections II.A-B of this sur-reply
`
`are the same as the corresponding sections of the 1281 sur-reply. Section II.C is
`
`specific to this IPR.
`
`2 Petitioner’s arguments that Dr. Villasenor has taken inconsistent positions (Reply
`
`at 5) is irrelevant given Petitioner does not dispute the relevant portion of the
`
`construction. It is also incorrect for the reasons Dr. Villasenor explained at his
`
`deposition: “[T]o eliminate any chance of [] ambiguity, or at least eliminate any
`
`chance in any ambiguity regarding what I was saying, I provided an expanded
`
`98623272.1
`
`- 4 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`Because Petitioner does not—and cannot—dispute that “identifying” a pattern
`
`includes linking parameters to parameter values, Petitioner has no legitimate basis
`
`to dispute that that Petitioner’s construction of “fixing … by associating …” renders
`
`that step
`
`indistinguishable from
`
`the separately-recited “identifying” step.
`
`Specifically, Dr. Allen testified that “defining a link between two data items” “would
`
`certainly be one way of associating.” Ex. 2003 at 48:9-13. But that is exactly what
`
`identifying a pattern also encompasses. Thus, “associating at least one parameter of
`
`said subset of varying parameters with said first pattern” and “identifying a first
`
`pattern” are indistinguishable under Petitioner’s construction, demonstrating the
`
`incorrectness of Petitioner’s “fixing … by associating …” construction.
`
`Petitioner offers no legitimate explanation. Petitioner attempts to dismiss the
`
`exercise as illegitimate because “Qualcomm uses its re-definition of ‘pattern’”
`
`(Reply at 6), but Petitioner does not dispute the relevant portion of that definition.
`
`Petitioner also points to the Institution Decision’s conclusion that Petitioner had
`
`mapped “identifying” and “associating” to different functionality. See Reply at 5.
`
`But the issue addressed in the Institution Decision—whether Petitioner mapped to
`
`the same functionality—is a different issue. This issue here is that Petitioner’s
`
`
`definition in Exhibit 2005; but to me they mean the same thing.” Ex. 2026 at 137:8-
`
`19; see also 136:2-20.
`
`98623272.1
`
`- 5 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`proposed construction of “fixing …” is so broad that it is indistinguishable from the
`
`separately-recited “identifying a pattern.” For example, the Institution Decision
`
`notes that Petitioner pointed to learning patterns in a training phase as “identifying.”
`
`Decision at 21. Such learning of patterns, which includes linking of a varying
`
`parameter to a parameter value, would also be “associating” under Petitioner’s
`
`construction of “fixing … by associating ….”
`
`B.
`
`“Fixing … by Associating …”
`1.
`Petitioner Cannot Justify Deleting the “Fixing” Phrase
`The Institution Decision held that the language “fixing … by associating …”
`
`“limits the claim to fixing the parameters only by the action of associating as
`
`recited, rather than by other actions.” Decision at 17 (emphasis in original). The
`
`Decision did not suggest that it is permissible to remove “fixing …” as a limitation;
`
`rather, the Decision did not accept Qualcomm’s argument, under the record at the
`
`time, that Petitioner’s construction actually removed “fixing.” See id. Now, the
`
`evidentiary record has substantially evolved, and the unanimous evidence shows that
`
`Petitioner’s construction does remove “fixing ….”
`
`Petitioner does not—and cannot—dispute the testimony of its own expert that
`
`“the specification uses ‘associating’ as a substep of the process of fixing.” Ex. 2003
`
`at 51:20-23; see also Reply at 15 (“Even if, in a vacuum, ‘associating’ could be a
`
`sub-step of ‘fixing,’ …”). Nor does Petitioner even address Dr. Allen’s testimony
`
`98623272.1
`
`- 6 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`that “associating” is an action that “can be used in many different applications” to
`
`“achieve many different goals.” Ex. 2003 at 49:25-50:9. As Dr. Villasenor
`
`explained, “fixing …” and “associating …” are not synonymous, nor is “associating
`
`…” a special case of “fixing ….” Ex. 2004 at ¶¶ 32, 43.
`
`Thus, the undisputed record demonstrates that “associating” need not result in
`
`fixing. Accordingly, if Petitioner is permitted to delete the 12-word “fixing …”
`
`phrase, the claims no longer require “fixing.” Ex. 2004 at ¶ 43. Such a deletion
`
`would be improper under any circumstanced, but particularly here where the ’865
`
`Patent emphasizes the role of “fixing” in achieving the goals of the invention. See,
`
`e.g., Ex. 1001 at 8:45-60.
`
`2. Qualcomm’s Construction Sets Forth the Meaning of “Fixing”
`Used by the Specification
`Qualcomm’s construction of “fixing” as “setting the scope of analysis” stands
`
`as the only construction that actually assigns meaning to “fixing.”3 While Petitioner
`
`criticizes Qualcomm’s construction, it fails to offer any alternative meaning for
`
`“fixing” beyond “associating.”
`
`
`3 Qualcomm has thus addressed the Institution Decision’s observation that
`
`Qualcomm had not proposed an alternative construction. Decision at 20.
`
`98623272.1
`
`- 7 -
`
`

`

`The correctness of Qualcomm’s construction—and
`
`Case IPR2018-01282
`Patent 8,768,865
`incorrectness of
`
`Petitioner’s—can be seen by substituting both into the specification passages that
`
`describe what “fixing” enables. For example, the ’865 Patent states that the act of
`
`“fixing one variable associated with or corresponding to ‘driving’” results in “an
`
`application processor associated with a mobile device” being able to “observe what
`
`other variables have patterns if a motion state corresponds, for example, to
`
`‘driving.’” Ex. 1001 at 13:36-37, 13:23-26.
`
`Substituting in Qualcomm’s construction, it remains a true statement that
`
`“setting the scope of analysis to where motion state is equal to driving” enables
`
`an application processor to “observe what other variables have patterns if a motion
`
`state corresponds, for example, to ‘driving.’” Id. at 13:23-26. In contrast, merely
`
`“associating the varying parameter motion state with the parameter value
`
`driving” would not enable an application processor to “observe what other variables
`
`have patterns if a motion state corresponds, for example, to ‘driving.’” Id.
`
`Petitioner’s construction is also consistent with the fact that “associating” is a
`
`subset of “fixing” because the scope of analysis can be set using the “association”
`
`of “motion state” with “driving.” But, again, merely performing the “associating”
`
`without using the “associating” to set the scope of analysis would not assist
`
`subsequent pattern recognition.
`
`98623272.1
`
`- 8 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`Additional specification passages describe “fixing” as enabling this same
`
`result. See, e.g., id. at 13:19-22 (“one or more patterns in a second subset of variables
`
`may be identified, for example, if there is a pattern in the fixed subset of variables”),
`
`Ex. 2001 at 15 (“observe what other variables have patterns when motion state
`
`corresponds to “driving””). In each case, “fixing” sets the scope of analysis to where
`
`the fixed parameters match the fixed parameter values.
`
`3.
`
`Petitioner Identifies No Legitimate Criticism of Qualcomm’s
`Construction
`Petitioner does not identify any portion of the specification that is inconsistent
`
`with Qualcomm’s construction. See Reply at 8-13. Instead, Petitioner makes a
`
`series of arguments that either provide no legal basis to dispute Qualcomm’s
`
`construction or are factually incorrect.
`
`First, Petitioner incorrectly asserts that Qualcomm relies solely on the
`
`provisional application and annotations of Figure 2. Reply at 8. To the contrary,
`
`Qualcomm cites to and analyzes numerous instances in the specification that use the
`
`word “fixing” in the same context as the claim language. See Response at 10-11
`
`(citing ’865 Patent4 at 13:19-22, 13:36-27, 13:23-26). Moreover, given that the ’865
`
`
`4 Qualcomm’s citation on page 10 to 13:19-22 should have referred to “Ex. 1001”
`
`rather than “id.” Given that citations to the ’865 Patent are by column and line and
`
`98623272.1
`
`- 9 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`Patent incorporates its provisional application by reference (Ex. 1001 at 1:7-10),
`
`Petitioner is incorrect to suggest text from that provisional application is any less
`
`part of the specification than any other passage.
`
`Second, Petitioner argues that the specific words “setting the scope of pattern
`
`recognition analysis” do not appear in the specification. See Reply at 8. Petitioner
`
`cites no case law that the words of a construction must appear in the specification,
`
`nor does such a requirement have any logical basis. Under the broadest reasonable
`
`interpretation standard, claim terms generally are given their ordinary and customary
`
`meaning, as would be understood by one of ordinary skill in the art in the context of
`
`the entire disclosure. In re Translogic Tech., Inc., 504 F.3d 1249, 1257 (Fed. Cir.
`
`2007). Qualcomm’s construction does so.
`
`Third, Petitioner argues that consistent use of the term in the specification is
`
`not enough because “a repeated and consistent description is insufficient to create a
`
`clear disavowal of claim scope.” Reply at 12. This argument is irrelevant because
`
`Qualcomm has not asserted any claim scope disavowal or sought to narrow the
`
`claims to less than the plain and ordinary meaning in view of the entire disclosure.
`
`
`citations to the provisional application are by page, that citation—and those that
`
`follow—are plainly directed to column and line numbers of the ’865 Patent.
`
`98623272.1
`
`- 10 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`Petitioner has offered no alternative plain and ordinary meaning of “fixing.” It seeks
`
`to remove the term from the claim.
`
`Fourth, Petitioner criticizes Dr. Villasenor’s annotations of Figure 2 of the
`
`’865 Patent. Reply at 9-11. To be clear, Qualcomm does not suggest that Figure 2
`
`provides independent support for the meaning of “fixing.” Rather, Dr. Villasenor
`
`simply used the illustration of Figure 2 as a background upon which to illustrate the
`
`concept of “fixing” as described in the specification passages discussed above. See
`
`Ex. 2004 at ¶ 33. In any event, Petitioner’s criticisms of the annotations are all based
`
`on attempts to interpret the spacing of elements on the timeline of Figure 2 in a way
`
`that Dr. Villasenor explained is contrary to how a POSITA would interpret the
`
`figure: “And, again, this is an illustrative figure, it’s not one that’s intended to be
`
`analyzed with calipers measuring tenth of a millimeter distances. I think it’s clear
`
`to one who’s skilled in the art.” Ex. 1026 at 58:20-24; see also Hockerson-
`
`Halberstadt, Inc. v. Avia Group Int’l, Inc., 222 F.3d 951, 956 (Fed. Circ. 2000) (“it
`
`is well established that patent drawings do not define the precise proportions of the
`
`elements and may not be relied on to show particular sizes if the specification is
`
`completely silent on the issue.”). Petitioner offers no evidence to the contrary.
`
`Fifth, Petitioner argues that Qualcomm’s construction requires recognizing a
`
`second pattern in Claim 1. See Reply at 7. As Qualcomm already explained—and
`
`Petitioner does not address—“Patent Owner does not suggest that Claim 1 requires
`
`98623272.1
`
`- 11 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`performing pattern recognition—that is recited separately in dependent Claim 3.”
`
`Response at 19. Rather, fixing is setting the scope of analysis such that, when a
`
`pattern recognition process is later initiated, that later analysis is performed within
`
`the scope set by “fixing.”
`
`Sixth, Petitioner argues that the phrase “setting the scope of pattern
`
`recognition analysis” renders the claim indefinite because “Qualcomm’s approach
`
`introduces antecedent basis error.” Reply at 7. Qualcomm disagrees that this phrase
`
`leads to an antecedent basis issue in the same way “the side of a box” does not
`
`introduce any antecedent basis issues. In any event, this is not a basis to
`
`substantively challenge the construction. If the Board believes “setting a scope of
`
`pattern recognition analysis” is more appropriate, Qualcomm is not opposed to that
`
`modification of the construction.
`
`4.
`
`Subsequent Pattern Recognition Efforts May Indicate that
`Fixing Did Not Occur
`Petitioner argues that Qualcomm improperly reads a “co-occurring” limitation
`
`into the claim. Reply at 11-14. This does not accurately state Qualcomm’s position.
`
`As Qualcomm has stated, and as reflected in the proposed construction, “fixing” sets
`
`the scope of subsequent pattern recognition analysis to where the fixed varying
`
`parameters match their fixed value(s). It logically follows that, during the time the
`
`scope of pattern recognition analysis is set, only patterns that co-occur with the fixed
`
`98623272.1
`
`- 12 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`varying parameters having their fixed value(s) can be recognized. This is because
`
`pattern recognition efforts would not include analyzing an instance of data where the
`
`fixed values are not met.
`
`Again, Qualcomm does not assert that “fixing” itself involves pattern
`
`recognition. But, where a reference describes a pattern recognition analysis that
`
`identifies patterns that include a value other than the fixed value(s) of a fixed
`
`varying parameter (i.e., does not co-occur with the fixed pattern), then it can be
`
`concluded that fixing did not occur. That is, the scope of pattern recognition was
`
`not set to where the fixed values are true for that pattern recognition analysis.
`
`The specification’s description of identifying patterns that “co-occur” after
`
`“fixing” further supports Qualcomm’s construction. Petitioner incorrectly asserts
`
`that “Qualcomm does not cite a single passage of the as-issued ’865 patent
`
`describing a co-occurrence requirement.” Reply at 12. To the contrary, Qualcomm
`
`cites multiple portions of the ’865 Patent describing identifying a second pattern that
`
`“co-occurs” with the fixed varying parameters as a result of the “fixing.” Response
`
`at 8-9 (citing Ex. 1001 at 8:7-11, 15:21-25).
`
`5. Qualcomm’s Construction Does Not Contradict Its Litigation
`Positions
`Petitioner also argues that Qualcomm has contradicted its litigation positions.
`
`Reply at 14. To the contrary, Qualcomm disputed Petitioner’s construction of
`
`98623272.1
`
`- 13 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`“fixing” throughout the district court claim construction process, stating:
`
`“Qualcomm believes Apple’s construction fails to give meaning to all words of the
`
`claim, which may be relevant to validity.” Ex. 1024 at 86. Moreover, Qualcomm’s
`
`construction is consistent with Qualcomm’s description of the meaning of “fixing”
`
`in the district court. For example, in opposition to a Section 101 motion filed by
`
`Petitioner, Qualcomm again emphasized that Petitioner was ignoring the “fixing”
`
`requirement of the claim. See, e.g., D.I. 1295 at 10 (“By performing this ‘fixing,’ a
`
`‘mobile device may then search for a pattern that co-occurs’ with the condition
`
`represented by the fixed parameters.”) (citing Ex. 1001 at 8:41-45).
`
`Finally, Petitioner asserts that Qualcomm’s infringement contentions are
`
`inconsistent with its claim construction, but offers no actual explanation of this
`
`inconsistency. Reply at 14. Qualcomm’s infringement contentions are consistent
`
`with its construction of fixing. On page 53 of the claim chart (page 90 of Ex. 1025),
`
`Qualcomm identifies headphone status (i.e., whether headphones are plugged in) as
`
`an example of a varying parameter with plugged in as an example of a value that can
`
`be fixed. Then, on page 68 of the claim chart (page 105 of Ex. 1025), Qualcomm
`
`explains that a second pattern representing a user listening to music at the gym is
`
`
`5 To the extent the Board would like Qualcomm to file this document as an exhibit,
`
`Qualcomm can do so.
`
`98623272.1
`
`- 14 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`identified based on the first identified, and co-occurring, pattern of “a user desiring
`
`to listen to audio upon plugging in headphones” that includes the varying parameter
`
`headphone status fixed to the value of plugged in.
`
`C.
`
`“Initiating a process to attempt a recognition of a second
`pattern…”
`Petitioner argues, for the first time in its Reply, that “the BRI” of “first
`
`pattern” and “second pattern” as used in Claim 3 is “broad enough to encompass
`
`instances where the first pattern and the second pattern are the same.” Reply at 20.
`
`Petitioner makes this attorney argument without elaboration, including any
`
`analysis of the intrinsic record. Id. Nothing in the ’865 Patent suggests that “first
`
`pattern” and “second pattern” refer to multiple instances of the same pattern, and
`
`such an interpretation would be contrary to the intrinsic record.
`
`The specification consistently explains recognizing a “second” pattern refers
`
`to recognizing something other than the first pattern and uses different language to
`
`refer to another occurrence of the “first” pattern. For example, in describing the
`
`functionality claimed in Claim 3, the patent explains that as part of “a process to
`
`attempt a recognition of a second pattern”
`
`For example, in the next or otherwise suitable occurrence of a first
`
`pattern (e.g., a condition or event-related pattern, etc.), a “snapshot” of
`
`98623272.1
`
`- 15 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`a second pattern co-occurring with a first pattern and representative
`
`of associated sensors’ behavior may be electronically captured.
`
`Ex. 1001 at 15:18-25 (emphasis added). It is apparent from this passage that the
`
`’865 Patent uses “occurrence” to refer to multiple instances of the same pattern. It
`
`is likewise apparent that “first” and “second” are used to refer to different patterns.
`
`See also id. at 13:23-26 (describing an attempt to recognize “what other variables
`
`have patterns” when the first pattern is present).
`
`Claim 3 recites: “initiating a process to attempt a recognition of a second
`
`pattern in connection with said monitoring said input signals based, at least in part,
`
`on said first identified pattern.” Id. at 21:17-20 (emphasis added). Claim 3 does
`
`not use the term “occurrence” used to indicate multiple instances of the same pattern.
`
`Thus, a POSITA would understand that the context of the ’865 Patent is
`
`different than the context of the claims in Louch. In contrast, Louch Claim 16
`
`explicitly recites “a substantial match between the first movement pattern and the
`
`second movement pattern,” leaving no doubt that those terms refer to occurrences of
`
`data, not separate patterns. Ex. 1011.
`
`III. PETITIONER’S NEW “FIXING … BY ASSOCIATING …” THEORY
`RELIES ON FUNCTIONALITY LOUCH DOES NOT ACTUALLY
`DISCLOSE
`Petitioner now asserts that “fixing … by associating …” is met under
`
`Qualcomm’s construction where the system of Louch uses “a first pattern (e.g.,
`
`98623272.1
`
`- 16 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`position of the device) plus a time duration” (i.e., a “duration” pattern) “to determine
`
`a control action.” Reply at 18. This theory suffers from at least two independent
`
`flaws. First, Louch does not disclose any such “first pattern” when a “duration”
`
`pattern is used to detect a control action, nor would a POSITA understand any such
`
`partial pattern to be separately-defined. Second, even if Louch disclosed detecting
`
`Petitioner’s purported “first pattern” as part of detecting a “duration” pattern, Louch
`
`includes no disclosure regarding the scope of analysis used when confirming a
`
`“duration.”
`
`A. Louch Does Not Define, Nor Would a POSITA Understand There
`to Be, Any Portion of a “Duration” Pattern as a Separate Pattern
`Petitioner’s theory fails because Louch never discloses the alleged “first
`
`pattern” in an embodiment in which a “duration” pattern is used. As an initial matter,
`
`Petitioner does not—and cannot—dispute that Petitioner may not rely on alternative
`
`embodiments to cobble together a first and second pattern. Compare Response at
`
`33-36 with Reply at 21-23. Further, Petitioner concedes that in the embodiment of
`
`Louch in which a “duration” pattern is used, “the first pattern is not used to trigger
`
`a control action.” Reply at 22.
`
`To be clear, Louch never suggests that a system would define two patterns—
`
`one with and one without a duration component. Ex. 2006 at ¶ 56. To the contrary,
`
`not only Dr. Villasenor, but also Louch, the ’865 Patent, and Dr. Allen are all in
`
`98623272.1
`
`- 17 -
`
`

`

`Case IPR2018-01282
`Patent 8,768,865
`agreement that a “duration” pattern is a single pattern that is either met or not. Louch
`
`explains: “if the state change does not exceed a certain amount of time … an
`
`assumption can be made that the state change is temporal, and therefore no control
`
`action will be triggered.” Ex. 1011 at 5:13-17. The ’865 Patent likewise explains
`
`“unless a pattern lasts or exists for a minimum or threshold amount of time,” it may
`
`not “be indicative of an actual pattern.” Ex. 1001 at 10:17-21. And, Dr. Allen agrees
`
`that a “pattern with a duration constraint would have to hold for that amount of time
`
`before it matched.” Ex. 2003 at 34:25-35:7.
`
`Petitioner fails to address, let alone distinguish, these passages—all of which
`
`were cited in Qualcomm’s Response. Compare Response at 29-30 with Reply at 17-
`
`19, 21-23. Instead, Petitioner points to deposition testimony of Dr. Villasenor
`
`related not to what Louch actually discloses, but to how a pattern could be defined
`
`and how a duration pattern could be recognized. See Reply at 22-23. This does not
`
`help Petitioner demonstrate the actual disclosure of Louch. In fact, in the same line
`
`of questioning Petitioner cites, Dr. Villasenor made clear that whether one could, in
`
`the abstract, define an intermediate state, that is not what Louch discloses: “[I]f
`
`duration is a criterion, then the way Louch is presented then you wouldn’t declare
`
`that you’re in the, quote, state until you actually have also assessed the duration.”
`
`Ex. 1026 at 123:23-124:4.
`
`98623272.1
`
`- 18 -
`
`

`

`B.
`
`Case IPR2018-01282
`Patent 8,768,865
`Even if a Portion of a “Duration” Pattern Were Deemed a First
`Pattern, Louch Does Not Disclose Setting a Scope of Analysis in
`Any Way
`Petitioner’s mapping fails for a second, independent reason. It is not enough
`
`to show Louch recognizing a “first pattern” and a “second pattern” that “co-occurs”
`
`with the “first pattern.” Petitioner must show that, prior to attempting to recognize
`
`the second pattern, Louch “set[s] the scope of pattern recognition analysis to where
`
`a subset of varying parameters match parameter values associated with said first
`
`pattern.” Again, Lo