`
`______________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`______________
`
`
`APPLE, Inc.,
`
`Petitioner
`
`v.
`
`QUALCOMM INCORPORATED,
`
`Patent Owner
`______________
`
`Case IPR2018-01281
`
`U.S. Patent No. 8,768,865
`______________
`
`
`
`DECLARATION OF JOHN VILLASENOR
`
`
`
`
`
`
`1
`
`Apple Inc. v. Qualcomm Incorporated
`IPR2018-01281
`Qualcomm Ex. 2004
`Page 1 of 30
`
`
`
`A. Qualifications
`1. My name is John Villasenor. I am a professor at UCLA. I have been retained
`
`by Qualcomm Incorporated to provide opinions in the inter partes review
`
`proceedings IPR2018-01281 (the ‘1281 proceeding) and IPR2018-01282 (the
`
`‘1282 proceeding) challenging U.S. Patent No. 8,768,865 (the ‘865 Patent).
`
`2. My work focuses on
`
`innovative, high-performance communications,
`
`networking, media processing, and computing technologies and their broader
`
`implications. Since well before the respective priority dates of the ‘865 Patent, I
`
`have performed research in areas including image processing, machine learning,
`
`and delivering content to mobile devices. For example, I have also done research
`
`in machine learning, with substantial experience developing algorithms that adapt
`
`in response to changing characteristics in the environment as reflected, for
`
`example, through data measured through sensors. In addition, I have performed
`
`research in relation to mobile devices since the 1990s. This research included
`
`multiple aspects of mobile devices, including wireless communications, sensing
`
`information (such as orientation) on mobile devices, and methods for delivering
`
`content to mobile devices, including considerations such as the selection of type of
`
`content for transmission to the mobile device.
`
`
`
`2
`
`Page 2 of 30
`
`
`
`3.
`
`I received my B.S. in Electrical Engineering from the University of Virginia
`
`in 1985, and M.S. and Ph.D. in Electrical Engineering from Stanford University in
`
`1986 and 1989, respectively.
`
`4. While at Stanford, I concentrated my research on digital signal processing
`
`and communications.
`
`5.
`
`Between 1990 and 1992, I worked for the Jet Propulsion Laboratory in
`
`Pasadena, CA, where I helped to develop techniques for imaging and mapping the
`
`earth from space. Since 1992, I have been on the faculty of the Electrical
`
`Engineering Department of the University of California, Los Angeles (UCLA).
`
`Between 1992 and 1996, I was an Assistant Professor; between 1996 and 1998, an
`
`Associate Professor; and since 1998, I have been a full Professor.
`
`6.
`
`For several years starting in the late 1990s, I served as the Vice Chair of the
`
`Electrical Engineering Department at UCLA. I also hold an appointment in the
`
`Department of Public Policy within the UCLA School of Public Affairs. In
`
`addition, I teach in the UCLA Anderson School of Management.
`
`7.
`
`Since joining the UCLA faculty in 1992, my research has addressed
`
`software, algorithms, hardware, networking, protocols, and other aspects of
`
`systems and devices that acquire, store, process, transmit, and display information.
`
`8.
`
`I am an inventor on approximately 20 issued and pending U.S. patents in
`
`areas including signal processing, data compression, communications, and
`
`
`
`3
`
`Page 3 of 30
`
`
`
`cybersecurity. I have published over 150 articles in peer-reviewed journals and
`
`academic conference proceedings.
`
`9.
`
`In addition to my work at UCLA, I am a nonresident senior fellow at the
`
`Brookings Institution in Washington, D.C. Through Brookings I have examined a
`
`wide range of topics at the technology/policy intersection including cybersecurity,
`
`wireless mobile devices and systems, intellectual property, financial inclusion for
`
`“unbanked” populations, digital media policy, “drones,” critical infrastructure
`
`security, driverless cars, and digital currencies and emerging payment methods. I
`
`have published articles and commentary related to technology policy in venues
`
`including Billboard, the Brookings Institution, the Chronicle of Higher Education,
`
`Fast Company, Forbes, the Huffington Post, the Los Angeles Times, Scientific
`
`American, Slate, and the Washington Post.
`
`10. My attached curriculum vita (Ex. 2006) details my over expertise and
`
`experience in the field of computer graphics and image processing.
`
`B. Materials reviewed
`11.
`I have reviewed each Petition submitted in each of the ‘1281 and ‘1282
`
`proceedings, as well the Patent Owner Preliminary Response submitted in each of
`
`the ‘1281 and ‘1282 proceedings.
`
`12.
`
`I have reviewed the ‘865 Patent that is included as Exhibit 1001 in each of
`
`the ‘1281 and ‘1282 proceedings, U.S. Provisional Application No. 61/434,400,
`
`
`
`4
`
`Page 4 of 30
`
`
`
`which incorporated by reference by the ‘865 Patent, that is included as Exhibit
`
`2001 in each of the ‘1281 and ‘1282 proceedings, and the prosecution history for
`
`the ‘865 Patent, excerpts of which are included as Exhibit 1002 in each of the
`
`‘1281 and ‘1282 proceedings.
`
`13.
`
`I have reviewed each declaration of Dr. Allen, which is Ex. 1003 in the
`
`‘1281 proceeding and Ex. 1021 in the ‘1282 proceeding.
`
`14.
`
`I have reviewed each Institution Decision entered by the Panel in each of the
`
`‘1281 and ‘1282 proceedings.
`
`15.
`
`I have reviewed the transcript of Dr. Allen’s deposition conducted as part of
`
`the ‘1281 and ‘1282 proceedings.
`
`16. A complete listing of the documents I reviewed is as follows:
`
`Proceeding
`
`Pr/ Ex Title
`
`1281 Proceeding
`1281 Proceeding
`1281 Proceeding
`1281 Proceeding
`1281 Proceeding
`1281 Proceeding
`1281 Proceeding
`1281 Proceeding
`
`1281 Proceeding
`1282 Proceeding
`1282 Proceeding
`1282 Proceeding
`1282 Proceeding
`1282 Proceeding
`
`2 Petition
`6 Patent Owner Preliminary Response
`7 Institution Decision
`1001 U.S. Pat. No. 8,768,865
`1002 Excerpts of ‘865 Patent Prosecution History
`1003 Allen Declaration
`1005 Wang
`2001 U.S. Provisional Application No. 61/434,400
`(Incorporated by Reference by the ‘865 Patent)
`2003 Deposition of James Allen (April 25, 2019)
`2 Petition
`6 Patent Owner Preliminary Response
`7 Institution Decision
`1011 Louch (U.S. 8,676,224)
`1021 Allen Declaration
`
`
`
`5
`
`Page 5 of 30
`
`
`
`C. Legal Principals
`17.
`I understand in this IPR proceeding claim terms, the BRI standard applies,
`
`which means 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.
`
`18.
`
`I understand that for a prior art reference to anticipate a claim, elements of a
`
`claim must be disclosed within the reference either expressly or inherently. For a
`
`claim element to be inherent in a reference, I understand that the claim element
`
`must be necessarily present in the reference. I also understand that the reference
`
`must clearly and unequivocally disclose the claimed invention or direct those
`
`skilled in the art to the invention without any need for picking, choosing, and
`
`combining various disclosures not directly related to each other by the teachings of
`
`the cited reference.
`
`D. Overview of the ‘865 Patent
`19. The ‘865 Patent generally concerns methods and apparatuses relating to
`
`machine learning on a mobile device, such as a smartphone or tablet. The ‘865
`
`Patent addresses the problem of making sense of large amounts of multi-
`
`dimensional data, as may be generated and/or processed by a mobile device. Such
`
`multi-dimensional data may be used to identify user behavior or actions. As the
`
`number of sensors and/or other data sources in a mobile device increases, it
`
`
`
`6
`
`Page 6 of 30
`
`
`
`becomes more difficult to process and analyze sensor and device data in order to
`
`identify relevant patterns. When faced with a large multi-dimensional dataset,
`
`efficiency of the associated analysis becomes extremely important—particularly on
`
`a mobile device with limited resources. The ‘865 Patent describes this problem:
`
`A popular and rapidly growing market trend in sensor-enabled
`technology includes, for example, intelligent or smart mobile
`communication devices that may be capable of understanding what
`associated users are doing (e.g., user activities, intentions, goals, etc.)
`so as to assist, participate, or, at times, intervene in a more meaningful
`way. Integration of an ever-expanding variety or suite of embedded or
`associated sensors that continually capture, obtain, or process large
`volumes of incoming information streams may, however, present a
`number of challenges. These challenges may include, for example,
`multi-sensor parameter tracking, multi-modal information stream
`integration, increased signal pattern classification or recognition
`complexity, background processing bandwidth requirements, or the
`like, which may be at least partially attributed to a more dynamic
`environment created by user mobility. Accordingly, how to capture,
`integrate, or otherwise process multi-dimensional sensor information
`in an effective or efficient manner for a more satisfying user
`experience continues to be an area of development.
`
`Ex. 1001 at 1:42-60. The patent also notes:
`
`As alluded to previously, continually tracking or monitoring all or
`most varying parameters or variables that may be associated with a
`multi-dimensional stream of sensor
`information may be a
`computationally intensive, resource-consuming, at times intractable,
`or otherwise less than efficient or effective approach for pattern
`matching or recognition.
`Id. at 7:58-63.
`
`
`
`7
`
`Page 7 of 30
`
`
`
`20. The ‘865 Patent teaches an efficient multi-stage approach processing and
`
`analyzing this multi-dimensional data, thereby saving device resources and
`
`improving the performance of the mobile device. See, e.g., id. at 8:50-54 (“This
`
`may make pattern matching more tractable or otherwise allow for a more effective
`
`or efficient pattern recognition since a pattern matching process is performed in a
`
`remaining or reduced set of variables.”). By using this multi-stage approach, the
`
`mobile device avoids the shortcomings of the prior art requiring “continually
`
`tracking or monitoring all or most varying parameters.” See, e.g., id. at 7:58-63.
`
`21.
`
` One exemplary embodiment is disclosed in Figure 4 of the ‘865 Patent:
`
`
`
`8
`
`Page 8 of 30
`
`
`
`
`In the context of the ‘865 Patent claims, a mobile device monitors its various
`
`22.
`
`sensor and application data sources and detects a condition or an event of interest.
`
`See, e.g., id. at 7:40-45; 8:54-60; 9:7-11; 14:60-65. This can be considered a first
`
`stage.
`
`23. The mobile device also identifies a “first pattern” based on this detected
`
`condition. See, e.g., id. at Fig. 4, 14:66-15:5. The “first pattern may comprise a
`
`distinct signal-related pattern having one or more varying parameters or variables
`
`of interest that may be representative or otherwise correspond to” the “detected
`
`condition.” See id. at 15:1-5.
`
`
`
`9
`
`Page 9 of 30
`
`
`
`24. The mobile device then “fix[es] a subset of varying parameters associated
`
`with said first pattern.” See id. at Fig. 4. As will be discussed below, this “fixing”
`
`step sets the scope of analysis for subsequent pattern recognition. Specifically, the
`
`mobile device may then “initiat[e] a process to attempt recognition of a second
`
`pattern.” See id. This can be considered a second stage. By fixing varying
`
`parameters of the first pattern, during that second stage, the mobile device may
`
`attempt to recognize a second pattern that occurs when the fixed varying
`
`parameters match the first pattern. See id. at 13:23-37.
`
`25. Figure 5 of the ‘865 Patent illustrates an exemplary mobile device with a
`
`number of sensors that may be monitored for patterns indicative of user behavior
`
`or action:
`
`
`
`10
`
`
`
`Page 10 of 30
`
`
`
`26. Like most modern mobile devices, the device depicted in Figure 5 has
`
`several sensors, including, in this example, one or more accelerometers 514, an
`
`ambient light sensor 516, a proximity sensor 518, and other sensors 520 “such as a
`
`gyroscope, magnetometer, microphone, camera, GPS, WiFi, Bluetooth™-enabled
`
`devices, etc. to facilitate or otherwise support one or more processes associated
`
`with operating environment 500.” See id. at 17:28-42. The ‘865 Patent explains
`
`that “parameters,” “varying parameters,” or “variables” are “derived” from these
`
`monitored sensors. See id. at 9:63-10:3. The patent uses these three terms
`
`interchangeably. Below, for shortness, I may use only one term, but use of any
`
`such one term should be considered a reference to all three terms. The ‘865 Patent
`
`also teaches that other “variables” may be derived from non-sensor sources, such
`
`as indications user behavior or action, including time of day, day of the week, a
`
`state or action of an application, and actions taken on the mobile device (e.g.
`
`silencing a ringer, muting a call, or sending a text message). See id. at 8:1-6,
`
`14:60-65. The variables have values that change over time. See, e.g., id. at Figure
`
`2.
`
`The ‘865 Patent describes examples of “variables” that are derived using different
`
`levels of analysis, and one variable may be used to derive another variable. For
`
`example, the patent describes “acceleration” as an example of a “variable.” Id. at
`
`4:8-13. The patent also describes “motion state” as another variable and explains
`
`
`
`11
`
`Page 11 of 30
`
`
`
`“acceleration” may be used to determine values of the variable “motion state” such
`
`as “driving,” explaining: “an acceleration vibration may, for example, indicate that
`
`a user is driving or walking.” Id. at 7:30-32.
`
`27. A person of ordinary skill in the art would understand that “fixing”
`
`parameters, in the context of the ‘865 Patent, refers to setting the scope of analysis
`
`to enable pattern recognition of additional patterns when there is a pattern in the
`
`fixed parameters.
`
`28. The ‘865 Patent describes the function of “fixing” on several occasions, each
`
`consistent with the understanding described above. First, the ‘865 Patent explains:
`
`At least one subset of variables of interest may be fixed, as discussed
`above, and 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.
`
`Id. at 13:19-22.
`
`29. Next, the ‘865 Patent explains:
`
`By way of example but not limitation, an application processor
`associated with a mobile device may observe what other variables
`have patterns if a motion state corresponds, for example, to "driving,"
`as one possible illustration.
`Id. at 13:23-26. The patent goes on to explain that what was described in the
`
`quotation above is “fixing one variable associated with or corresponding to
`
`‘driving.’” Id. at 13:36-37.
`
`
`
`12
`
`Page 12 of 30
`
`
`
`30. The ‘865 Patent references U.S. Provisional Application No. 61/434,400 and
`
`states “the entire disclosure of which is hereby incorporated by reference.” Id. at
`
`1:7-10. I understand that this means that the contents of that provisional
`
`application are part of the ‘865 Patent disclosure. That provisional application
`
`states:
`
`■ Monitor variables individually for patterns
`■ Fix one subset of variables and identify patterns in a second subset
`of variables when there is a pattern in the fixed subset of variables
`
`■ E.g., observe what other variables have patterns when motion
`state corresponds to “driving”
`Ex. 2001 at 2001.015.
`
`31. The ‘865 Patent also uses the term “associating.” A person of ordinary skill
`
`in the art would understand that in the computer science context, the word
`
`“associating” refers to a basic building-block step that may be used in any number
`
`of contexts to, along with other steps, achieve any number of goals. This is also
`
`how the term is used in the ‘865 Patent. For example, the ‘865 Patent describes
`
`“associating” as part of multiple different processes:
`
` “An example of context labeling may include associating a specific
`
`accelerometer pattern with the context ‘surfing,’ for example, by providing
`
`other context, such as a camera view, location corresponding to a beach,
`
`detecting ‘wetness’, or the like.” Ex. 1001 at 14:13-17 (emphasis added).
`
`
`
`13
`
`Page 13 of 30
`
`
`
` “In some instances, a subset may be fixed, for example, by associating
`
`parameters or variables with a particular, distinct, or otherwise suitable
`
`pattern to represent a certain detected condition or event, as one possible
`
`example.” Id. at 15:9-12 (emphasis added)
`
`A person of ordinary skill in the art would understand that “context labeling” and
`
`“fixing” are not the same processes.
`
`32. A person of ordinary skill in the art would understand that the ‘865 Patent
`
`describes “associating” as a substep of “fixing,” that “associating” does not, on its
`
`own, accomplish “fixing.” As discussed above, the patent explains that by “fixing”
`
`“motion state” to “driving,” “a mobile device may observe what other variables
`
`have patterns if a motion state corresponds, for example, to ‘driving,’ as one
`
`possible illustration.” Id. at 13:23-26, 13:36-37. But a mobile device would not
`
`able to “observe what other variables have patterns if a motion state corresponds,
`
`for example, to ‘driving,’” if all that had been done was “associating” “motion
`
`state” with “driving.” Rather, that association must be used to set the scope of
`
`analysis to enable pattern recognition of additional patterns when “motion state” =
`
`“driving.”
`
`33. The concept of fixing can be illustrated through annotation of Figure 2 of the
`
`‘865 Patent. In the first annotation below, the scope of analysis is shown as the
`
`area inside the red box.
`
`
`
`14
`
`Page 14 of 30
`
`
`
`‘865 Patent Figure 2 (Annotated to Show “Fixing”)
`34. The scope of analysis shown in the red box corresponds to where “Motion
`
`State” = “Driving.” As discussed above, the ‘865 Patent uses “fixing” to describe
`
`this step of setting the scope of analysis. The substep of “associating” “Motion
`
`State” with “Driving” is part of this process, but on its own does not accomplish
`
`fixing.
`
`35. The next annotation below shows a second pattern found within the red box
`
`as a result of pattern recognition efforts after the “fixing” process. Specifically,
`
`based on the “fixing … by associating …,” the system can then attempt to
`
`recognize a second pattern when motion state corresponds to “driving.” Ex. 2001
`
`at 2001.015. This example second pattern is “Motion State” = “Driving” +
`
`“Change in Location” from “SSID_3” to “SSID_1.”
`
`
`
`15
`
`Page 15 of 30
`
`
`
`‘865 Patent Figure 2 (Annotated to Show “Recognize Second Pattern”)
`36. The patent also explains that by “fixing,” “a set of variables associated with
`
`a multi-dimensional sensor information stream may be advantageously reduced.” I
`
`note that “fixing” is not reducing variables, but a reduction in variable can result
`
`from “fixing.” For example, within the scope of analysis set in the fixing step, the
`
`variable “Brightness Level” remains unchanged. The system may be able to
`
`disregard that variable as a result. This is shown in the annotation below.
`
`
`
`16
`
`Page 16 of 30
`
`
`
`‘865 Patent Figure 2 (Annotated to Show “Variable Disregarded”)
`
`E. Claim Construction
`i.
`“Pattern”
`37. A person of ordinary skill in the art would understand the BRI of “pattern,”
`
`as used in the ‘865 Patent, to be: “a collection of one or more pairs of varying
`
`parameters and corresponding parameter values, as well as the relationship
`
`between each pair (where the relationship may be implicit).” The claims refer to
`
`both a “first pattern” and a “second pattern.” I understand the Petitioner proposed
`
`a construction of “a collection of one or more parameter values.” Pet. at 11-13.
`
`This construction is incomplete because it does not explicitly state that the pattern
`
`includes not only parameter values, but the linked parameter.
`
`
`
`17
`
`Page 17 of 30
`
`
`
`38. One example pattern in the patent is: “location X AND motion state Y.” Id.
`
`at 13:8-13. Here, “location” and “motion state” are variables, with “X” and “Y”
`
`being the corresponding values. “AND” is a logical representation that indicates
`
`that the pattern requires both pairs. Figure 3 of the patent illustrates another
`
`example of a pattern:
`
`
`
`39.
`
`In this example, “context_type” refers to “variables” and “context_value”
`
`refers to the linked variable value. In this example, while not written down, it
`
`would be understood that logically each pair of variables and corresponding values
`
`are required to make up the pattern. That is, the pattern may be represented as:
`
`“SoundIntensity = Loud AND PeriodicMovement = Running.”
`
`
`
`18
`
`Page 18 of 30
`
`
`
`40. The patent also discusses “identifying” patterns. A person of ordinary skill
`
`in the art would understand that identifying a pattern means identifying all of the
`
`elements I discussed above that make up the pattern.
`
`ii.
`“Fixing … by Associating …”
`41. Consistent with the descriptions of “fixing” and “associating” above, a
`
`person of ordinary skill in the art would understand that “fixing a subset of varying
`
`parameters associated with said first pattern by associating at least one parameter
`
`of said subset of varying parameters with said first pattern to represent said at least
`
`one detected condition” means “setting the scope of pattern recognition analysis to
`
`where a subset of varying parameters match parameter values associated with said
`
`first pattern by associating at least one parameter of said subset of varying
`
`parameters with said first pattern to represent said at least one detected condition.”
`
`This term is found in each of the challenged independent claims (Claims 1, 21, or
`
`46). As discussed above, the ‘865 Patent consistently describes the function of
`
`“fixing” parameters as setting the scope of analysis to enable pattern recognition of
`
`additional patterns when there is a pattern in the fixed parameters.
`
`42.
`
`I understand that Petitioner has asserted that “[t]his phrase is broad enough
`
`to encompass ‘associating at least one parameter of a subset of varying parameters
`
`with the first pattern to represent at least one detected condition.’” I do not agree
`
`that a person of ordinary skill in the art would interpret the claim language in this
`
`
`
`19
`
`Page 19 of 30
`
`
`
`way, which takes the original language and removes “fixing a subset of varying
`
`parameters associated with said first pattern,” for several reasons.
`
`43. First, as I have discussed above, “associating at least one parameter of a
`
`subset of varying parameters with the first pattern to represent at least one detected
`
`condition,” on its own, does not accomplish what the specification repeatedly
`
`describes “fixing” as accomplishing. Rather, “associating” is only a substep that
`
`on its own does not accomplish “fixing.” Therefore, Petitioner’s proposed
`
`construction is contrary to the description of “fixing” in the patent.
`
`44. Second, Petitioner’s construction—“associating at least one parameter of a
`
`subset of varying parameters with the first pattern” is indistinguishable from the
`
`separately-recited step of “identifying a first pattern based, at least in part, on said
`
`at least one detected condition.” As discussed above, “identifying a first pattern”
`
`includes identifying each of the variables included in the first pattern and linking
`
`those variables to corresponding values.
`
` But Petitioner’s construction—
`
`“associating at least one parameter of a subset of varying parameters with the first
`
`pattern”—also just covers that same “linking” activity and nothing more. That is,
`
`“associating at least one parameter of a subset of varying parameters with the first
`
`pattern” is already accomplished as part of identifying the first pattern because
`
`identifying the first pattern includes linking parameter values of the first pattern
`
`with the parameters of the first pattern. A person of ordinary skill in the art would
`
`
`
`20
`
`Page 20 of 30
`
`
`
`not interpret the “fixing” step in a way that made it duplicative of the “identifying”
`
`step.
`
`F.
`45.
`
`Level of ordinary skill in the art
`In my opinion, a person of ordinary skill in the art relevant to the ’865 Patent
`
`at the time of its invention would have had a Bachelor’s of science degree in
`
`electrical engineering, computer science, computer engineering, or a closely-
`
`related field, and at least 2 years of work or research experience in the field of
`
`machine learning or a closely related field. More work experience could
`
`compensate for less education, and vice versa. At the time of the earliest filing
`
`date to which the ‘865 Patent claims priority, I was at least a person of ordinary
`
`skill.
`
`46.
`
`I note that the Allen Declaration defines a person of ordinary skill in the art
`
`as having a Bachelor of Science degree in either computer science or electrical
`
`engineering, together with at least two years of study in an advanced degree
`
`program in artificial intelligence, machine learning, or pattern recognition, or
`
`comparable work experience.
`
`47. To the extent there are any differences in these two definitions of one of
`
`ordinary skill in the art, I do not believe that there is a meaningful change of
`
`outcome using one definition of a person of ordinary skill in the art or the other.
`
`
`
`21
`
`Page 21 of 30
`
`
`
`G. Opinions about Wang and Petitioner’s mapping to the challenged
`claims
`I have reviewed Wang et al, “A Framework of Energy Efficient Mobile
`
`48.
`
`Sensing for Automatic User State Recognition” (“Wang”). I have also reviewed
`
`Petitioner’s mapping of “GROUND-1A” asserting that “Wang anticipates Claims
`
`1-4, 15-17, 21-23, 28, 29, 46, 47.” I understand that Petitioner also asserted other
`
`grounds alleging obviousness of certain claims based on Wang in view of
`
`additional references. My opinions relate to Petitioner’s mapping of Wang to
`
`challenged claims. I have not been asked to address those additional references. I
`
`note the additional grounds relate solely to dependent claims. For the reasons
`
`described below, Wang does not anticipate the challenged independent claims
`
`(Claims 1, 21, or 46), or dependent Claims 4 or 23. The claims addressed in
`
`GROUND-1B and GROUND-1C are not anticipated based on the fact that they
`
`depend on claims I address.
`
`i.
`Overview of Wang
`49. Wang describes “a novel design framework for an Energy Efficient Mobile
`
`Sensing System (EEMSS).” Ex. 1005 at 1. Wang describes a system that uses the
`
`current state and detection of “state transitions” to recognize user states: “EEMSS
`
`uses hierarchical sensor management strategy to recognize user states as well as to
`
`detect state transitions.”
`
`
`
`22
`
`Page 22 of 30
`
`
`
`50.
`
`“The core component of EEMSS is a sensor management scheme which
`
`defines user states and state transition rules by an XML styled state descriptor.” Id.
`
`at 2. “In essence, the state descriptor consists of a set of state names, sensors to be
`
`monitored, and conditions for state transition.” Id. at 3. Wang explains that the
`
`“conditions for state transition” are used to determine when to transition from one
`
`state to another. For example, Wang explains: “If the state transition criteria is has
`
`been satisfied, the user will be considered as entering a new state (denoted by
`
`“<NextState>” in the descriptor).” Id. at 4. Thus, Wang discloses EEMSS using
`
`the fact that the device is in a given state in combination with “state transition
`
`criteria” being satisfied to define a next state.
`
`51. Wang also explains that EEMSS achieves “energy savings” by defining a
`
`different set of sensors to be used depending on the current state. Id. at 3. Wang
`
`describes this as the “sensors to be monitored” specified in the XML styled state
`
`descriptor file. Id. The sensors to be monitored reflect what sensors need to be
`
`monitored to detect a state transition to a next state. Thus, the selection of those
`
`sensors is based on the defined next state or states in the system design. It does not
`
`reflect what sensors were needed to detect the current state.
`
`52. Wang also includes a Table 1. Wang states: “Table 1 illustrates the set of
`
`user states to be recognized by EEMSS and three characteristic features that define
`
`
`
`23
`
`Page 23 of 30
`
`
`
`each of these states. The three features are the location, motion and background
`
`sound information.” Id. at 5. Table 1 is reproduced below:
`
`
`
`Id. at 6.
`
`53. While Table 1 describes a way to “define” states, Wang does not disclose
`
`using those definitions in the EEMSS system. That is, Wang defines states in two
`
`different ways—based on the groups of “State Features” and linked values shown
`
`in Table 1, but separately based on the current state and a state transition—but only
`
`describes EEMSS using one the latter definition. For example, Table 1 defines
`
`state “Vehicle” as “Location” = “Keep on Changing” and “Motion” = “Moving
`
`Fast.” Id. at 6. But Wang does not describe EEMSS using that definition. Instead,
`
`Wang explains that EEMSS defines “Vehicle” as the combination of being in the
`
`state “Walking” and a state transition being detected: “If a significant amount of
`
`increase is found on both user speed and recent distance of travel, a state transition
`
`will happen and the user will be considered riding a vehicle.” Id. at 5. It is this
`
`
`
`24
`
`Page 24 of 30
`
`
`
`information that is provided in the XML state descriptor file. See id. at 3. A
`
`person of ordinary skill in the art would understand that the use of a current state
`
`plus a state transition criteria to define a next state is an alternative to defining a
`
`state based on a collection of state features and values.
`
`54. Further, the EEMSS system is incompatible with defining a state based on a
`
`collection of state features and values. As discussed above, Wang states that
`
`EEMSS uses “an XML-format state descriptor as system input that contains all the
`
`states to be automatically classified as well as sensor management rules for each
`
`state.” Id. at 3. Wang also illustrates this arrangement in Figure 3, reproduced
`
`below. As discussed above, the XML state descriptor file does not include “state
`
`features” and values as listed in Table 1. Thus, it is not possible for EEMSS to use
`
`that information.
`
`Figure 3 of Wang (Highlighted)
`
`
`
`
`
`25
`
`Page 25 of 30
`
`
`
`55.
`
`I understand that Petitioner identifies the state features shown in Table 1 for
`
`“walking” as the claimed “first pattern” and the state features shown in Table 1 for
`
`“vehicle” and “working” as two alternative claimed “second patterns.” See Pet. at
`
`18, 31, 35. In each situation, based on the definitions of those states, it is not
`
`possible for the identified “first pattern” to overlap in time with the identified
`
`“second pattern.” This is because Table 1 defines each of these three patterns
`
`using the “State Feature” “Motion” having a different value—“Moving Slowly”
`
`for “Walking,” “Moving Fast” for “Vehicle,” and “Still” for “Working.” Because
`
`“Motion” cannot have multiple values simultaneously, these three states cannot
`
`overlap in time.
`
`56. The fact that the states Petitioner asserts are associated with the “first
`
`pattern” and “second pattern” cannot overlap in time is also apparent from Figure 2
`
`of Wang, which shows a “sequential flow chart.” As shown in Figure 2, below, to
`
`reach the “Vehicle” state, a transition from the “Walking” state, i.e, “Walking” is
`
`no longer true, is detected. Addressing Figure 2, Wang explains: “If a significant
`
`amount of increase is found on both user speed and recent distance of travel, a state
`
`transition will happen and the user will be considered riding a vehicle. id. at 5.
`
`
`
`
`
`26
`
`Page 26 of 30
`
`
`
`
`
`Id. at 5.
`
`57. As a result of this lack of overlap, even if EEMSS used patterns of state
`
`features to detect states (which it does not), EEMSS could never detect what
`
`Petitioner identifies as a “second pattern” in data where what Petitioner identifies
`
`as the “first pattern” is true.
`
`ii. Wang does not disclose “identifying a first pattern” or
`“associating … with said first pattern”
`independent challenged claim (Claims 1, 21, or 46) recites
`
`58. Each
`
`“identify[ing] a first pattern based, at least in part, on said at least one detected
`
`condition.” Petitioner asserts that the claimed “first pattern” corresponds in Wang
`
`to a collection of “state features” and corresponding values as recited in Table 1 of
`
`Wang: “‘Location’ = ‘Keep on changing’