`Fitbit’s Reply
`
`
`
`
`IN THE UNITED STATES PATENT AND TRADEMARK OFFICE
`
`––––––––––
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`––––––––––
`
`FITBIT, INC. and WAHOO FITNESS LLC,
`Petitioners,
`v.
`
`BLACKBIRD TECH LLC d/b/a BLACKBIRD TECHNOLOGIES,
`Patent Owner.
`
`––––––––––
`
`Case IPR2017-02012
`Patent 6,434,212
`
`––––––––––
`
`PETITIONER FITBIT’S REPLY TO PATENT OWNER’S RESPONSE
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`
`I.
`
`Introduction
`
`The Board’s Institution Decision correctly found that claims 2 and 5 were
`
`obvious in view of Amano, and that claim 6 is obvious in view of the combination
`
`of Amano and Kato. Blackbird’s arguments are centered on the data processing
`
`limitations of each challenged claim. In finding that claims 2 and 5 were not
`
`anticipated by Amano, the Board appears to have accepted Blackbird’s argument
`
`that the Amano device does not calculate distance traveled by the user, but instead
`
`determines the user’s speed based upon data taken from a single step every 30
`
`seconds. (ID at 18; POR at 17-20). Blackbird advances this argument again in its
`
`Patent Owner Response, and argues that because Amano’s invention relates to
`
`calculating blood oxygen levels, its other teachings should be ignored. (POR at
`
`22).
`
`Blackbird’s “single step” argument is wrong and contradicted by the explicit
`
`teachings of Amano. As explained by Fitbit’s expert (and ignored by Blackbird in
`
`its Response), Amano teaches performing calculations using an overall time
`
`interval of data (30 seconds is provided as one example) in order to determine
`
`pitch (number of steps run per unit time), as well as the distance traveled by a
`
`runner during that time interval. (Ex. 1005 at ¶106).
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`-1-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Blackbird’s Patent Owner Response is silent on how its single step theory is
`
`possible given the fact that Amano teaches performing the calculations using
`
`signal processing techniques, such as Fast Fourier Transform (“FFT”). The
`
`unrebutted explanation by Fitbit’s expert establishes that these techniques analyze
`
`data from an extended time sample, and Amano’s reference to these techniques is
`
`inconsistent with Blackbird’s assertion that Amano only teaches performing
`
`calculations for a single step. Indeed, Blackbird’s expert admitted during his
`
`deposition that signal processing techniques disclosed by Amano (such as wavelet
`
`analysis) would use a time period longer than a single step, and that a POSITA
`
`who wanted to implement these signal processing techniques would be motivated
`
`to use a longer period of data to allow for more accurate step count and stride rate
`
`calculations.
`
`While the Board found at the Institution Decision stage that Amano
`
`discloses a single step calculation, it is clear that Amano’s calculation is not based
`
`upon only a single step, but an interval of data (such as 30 seconds). As explained
`
`by Fitbit’s expert, Amano teaches the claimed distance calculation of multiplying
`
`the number of steps counted by a stride length that varies in accordance with a
`
`stride rate. For these reasons, the Board should find that the challenged claims are
`
`not only obvious, but anticipated by Amano.
`
`
`
`
`
`
`
`
`
`
`-2-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Blackbird’s second argument, that Amano’s teachings may be rejected
`
`because the invention is used to calculate blood oxygen levels, mischaracterizes
`
`the extensive teachings of Amano for the operation of exercise devices and is legal
`
`error. The Federal Circuit has consistently held that it is the teachings contained
`
`within a prior art reference that is relevant and not merely the claimed invention.
`
`With respect to the Amano and Kato ground, Blackbird re-raises multiple
`
`alternate arguments regarding the combination, including an alleged lack of
`
`motivation to combine and the supposed lack of similarity of the references. (See
`
`PPOR at 43-54, POR at 25-40, ID at 24-28). These arguments lack merit and fail
`
`to address Fitbit’s explanation and evidence showing that a POSITA would have
`
`been motivated to substitute Kato’s stride length calculations that do not use the
`
`specific user’s own stride length measurements with Amano’ calculations that are
`
`based on user-specific stride length measurements. As explained below,
`
`Blackbird’s arguments do not hold up to scrutiny, are not supported by its own
`
`expert’s testimony, and are contradicted by the teachings of Amano and Kato.
`
`The Board should also be wary of relying on Blackbird’s expert. He was
`
`unable to answer basic questions, including being unable to provide definitions to
`
`terms included in his declaration, such as fitness trackers, health trackers, and
`
`pedometers. (See, e.g., Ex. 1021 at 15:18-16:3; 22:20-23:5; 24:10-24; 25:15-26:1;
`
`
`
`
`
`
`
`
`
`
`-3-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`29:1-18; 33:7-25). He struggled to explain what specific calculation was required
`
`by the claims. (Ex. 1021 at 63:5-74:16). And Dr. Caloyannides refused to take a
`
`position on what knowledge a POSITA possessed, claiming that such knowledge
`
`would depend on the specific (hypothetical) individual. (Ex. 1021 at 118:25-
`
`120:9). Such a flawed fundamental position should render his entire declaration
`
`without merit.
`
` Finally, Blackbird’s constitutional arguments against
`
`the
`
`application of inter partes review should be rejected.
`
`II. Amano anticipates and renders obvious “calculat[ing] a distance
`traveled by multiplying a number of steps counted by a stride length” as
`claimed.
`
`In the Institution Decision, the Board found that the limitation that requires
`
`“calculat[ing] a distance traveled by multiplying a number of steps counted by a
`
`stride length that varies in accordance with a stride rate” is rendered obvious by
`
`Amano. (ID at 18). In making this finding, the Board credited Dr. Choudhury’s
`
`(Fitbit’s expert) explanation that Amano calculates a distance by multiplying the
`
`user’s pitch (i.e., the number of steps per unit time) by the user’s stride length, and
`
`that Amano adjusts the stride length according to the user’s pitch. (ID at 17-18
`
`(citing Ex. 1005 at ¶106)). In mathematical terms, Amano performs the following
`
`calculation:
`
`
`
`
`
`
`
`
`
`
`-4-
`
`
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`This is “calculat[ing] a distance traveled by multiplying a number of steps
`
`counted by a stride length that varies in accordance with a stride rate.” And even if
`
`Amano did not expressly disclose the claimed calculation, it would have been
`
`obvious based on this teaching. Indeed, Blackbird’s expert admitted that the basic
`
`relationship between stride length and stride rate (applied by Amano and the ’212
`
`Patent) was well known. For example, he stated “I mean, common sense shows
`
`that. People have been running the same way for thousands of years.” (Ex. 1021
`
`at 82:21-83:3).
`
`The Board did not originally institute on the Amano anticipation ground and
`
`the Institution Decision suggests that the Board accepted Blackbird’s arguments
`
`that are re-raised in the Response: that Amano only calculates a speed and not a
`
`distance (POR at 16-17), and that Amano calculates a speed based on “a single
`
`step” instead of “a number of steps” (POR at 17-21). Both of Blackbird’s
`
`arguments are wrong and are premised on a mischaracterization of Amano.
`
`Furthermore, the claimed calculation is obvious in view of Amano.1 The Board
`
`
`
`1 Furthermore, while Blackbird’s Response argues that Amano does not teach the
`
`specific method of calculating distance set forth in claims 2 and 5, Blackbird’s
`
`expert admitted that the challenged claims do not set forth an exact calculation, and
`
`
`
`
`
`
`
`
`
`
`-5-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`should find that Amano anticipates claims 2 and 5 because Amano discloses
`
`calculating a distance for the processing time interval, and Amano’s use of FFT
`
`signal processing (and other similar techniques) in performing the calculation
`
`demonstrates that Amano performs a calculation based on a time window of data
`
`and is not based on a speed detected during a single step.
`
`A. Amano discloses “calculat[ing] a distance traveled” based on a
`time interval that is longer than a single step.
`
`Amano teaches and discloses “calculat[ing] a distance traveled” over an
`
`extended time interval. Blackbird is therefore wrong when it asserts that Amano
`
`discloses only calculating speed over a time interval. In particular, as set forth in
`
`the Petition, Amano teaches determining the pitch of a runner, which can be
`
`detected based upon processing carried out by CPU 201 on the body motion signal
`
`from body motion detector 104. (Petition at 35-36; Ex. 1003 at 11:52-55). Amano
`
`teaches that the CPU performs this calculation by using FFT processing of the
`
`body signal motion, and that CPU 201 can be setup to execute calculations at
`
`specific time intervals, with a time interval of 30 seconds being used as an
`
`example. (Ex. 1003 at 12:4-7; 13:25-26). The specification further explains that
`
`
`
`that necessary aspects of the claimed calculation must be implied from the claim
`
`language. (Ex. 1021 at 71:16-74:9).
`
`
`
`
`
`
`
`
`
`
`-6-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`when CPU 201 performs the calculation, it calculates “the distance run by the test
`
`subject.” (Ex. 1003 at 12:15-18).
`
`Dr. Caloyannides (Blackbird’s expert) argues that Amano calculates a speed
`
`based on Amano’s disclosure that “the user’s pitch…is sampled at time intervals
`
`and this sampling is used ‘to calculate the distance run by the test subject per unit
`
`of time.’” (Ex. 2001 at ¶38). Rather than take this disclosure at face value (i.e.,
`
`that Amano discloses calculating the distance the user ran during a time interval),
`
`Dr. Caloyannides opines that Amano merely discloses calculating a speed. (Ex.
`
`2001 at ¶38; Ex. 1021 at 142:15-143:6 (testifying that Amano has no teaching of
`
`“determining the distance that is traveled by a person that is walking”)).
`
`Blackbird and its expert are also wrong to assert that “Amano does not
`
`disclose calculating distance at all.” (POR at 2; Ex. 1021 at 142:15-
`
`143:6)(emphasis added). This assertion is directly contradicted by Amano,
`
`including by the following teaching:
`
`On the other hand, if an altitude difference is present,
`
`then in step Sa103, CPU 201 first determines the slope
`
`from the aforementioned altitude distance and the
`
`distance run during the 30 seconds.
`
`(Ex. 1003 at 18:40-44)(emphasis added).
`
`
`
`
`
`
`
`
`
`
`-7-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Thus, as made clear by the overall teachings of Amano as well as the
`
`specific language from Column 18, Amano teaches calculating the distance run
`
`over a particular time interval (30 seconds in this example), and does not merely
`
`perform a speed calculation based upon data consisting of a single step every 30
`
`seconds.
`
`When confronted with the passage from Column 18 of Amano quoted
`
`above, which shows that the disclosed device calculates the distance run during a
`
`time interval, Dr. Caloyannides (who stated he had considered the passage
`
`previously) chose to ignore the specific language regarding the determination of
`
`the distance run over the time interval:
`
`Q.
`
`So let's turn to column 18 of Amano.
`
`A. All right.
`
`Q. On line 40 it says, On the other hand, if an altitude
`
`difference is present, then in step Sa103, CPU 201 first
`
`determines the slope from the aforementioned altitude
`
`distance and the distance run during the 30 seconds.
`
`A. Yes. Again distance over time. That's the
`
`velocity. Distance run over 30 seconds, that's a velocity
`
`measurement.
`
`(Ex. 1021 at 165:2-11; see also Ex. 1003 at FIG. 12 (stating that the calculation
`
`display processing is executed every 30 seconds)). As the challenged claims
`
`
`
`
`
`
`
`
`
`
`-8-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`require calculating “a” distance traveled – calculating a distance traveled by the
`
`user in 30 seconds meets this claim limitation.2
`
`Even more telling is that Dr. Caloyannides made no attempt to determine
`
`how the device disclosed in Amano allegedly calculates speed or distance. At his
`
`deposition, Dr. Caloyannides admitted that he did not consider how Amano works
`
`because he had concluded that the ’212 Patent is directed to solving a different
`
`problem:
`
`Q. …It's your opinion that Amano calculates speed of
`
`the user every 30 seconds; correct?
`
`A. Yes.
`
`Q. And it's your opinion that it does that by an
`
`instantaneous calculation; correct?
`
`A. Yes, presumably, yes, yes.
`
`Q.
`
`There's no measurement of a – where was the user
`
`
`
`2 Blackbird’s Response appears to argue that the challenged claims require
`
`calculating the overall actual distance a runner has traveled (POR at 17, 25)(“actual
`
`distance traversed”, “as opposed to accurate total distance determination”), such a
`
`requirement is absent from the challenged claims. Instead, the claims merely
`
`require calculating “a distance traveled” – something that Amano explicitly
`
`discloses and explains in detail.
`
`
`
`
`
`
`
`
`
`
`-9-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`
`at some time zero, and where – and how far did the user
`
`travel to the stopping point?
`
`A. Oh, I don't know exactly. That's his algorithm. I
`
`did not go into that because I felt that he is addressing a
`
`different problem, does not do what the '212 patent
`
`does, and I did not take the time to see just exactly how
`
`does he do what he does which does not relate to the
`
`patent.
`
`Q. Okay. So in calculating speed, you can't say one
`
`way or the other whether Amano accomplishes the
`
`instantaneous speed calculation based upon measuring a
`
`distance traveled by the user in some small amount of
`
`time?
`
`A.
`
`I don't know over what period of time he actually
`
`averages the speed. I do not know that.
`
`Q.
`
`So you don't know if it's averaging the speed over
`
`one second?
`
`A.
`
`I don't know whether he average over one second,
`
`five or ten. I think that's a design parameter that he can
`
`select while he's running his machine.
`
`Q.
`
`I mean, is it possible that he's averaging the
`
`speed over 30 seconds?
`
`A.
`
`It's possible.
`
`Q. Okay.
`
`(Ex. 1021 at 155:15-156:21)(emphasis added).
`
`
`
`
`
`
`
`
`
`
`-10-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Dr. Caloyannides clearly erred by refusing to consider the overall teachings
`
`of the Amano reference. The Federal Circuit has consistently held that it is the
`
`teaching contained within a prior art reference that is relevant, not merely the
`
`claimed invention. See In re Applied Materials, Inc., 692 F.3d 1289, 1298 (Fed.
`
`Cir. 2012) (“A reference must be considered for everything that it teaches, not
`
`simply the described invention or a preferred embodiment.”).
`
`B. Amano’s use of FFT signal processing means that the disclosed
`calculation uses a time window of data and not just a single step.
`
`Amano’s use of FFT processing and other signal processing techniques
`
`(such as wavelet) also demonstrates that Blackbird’s single step theory is wrong.
`
`Signal processing techniques including FFT are used in Amano for detecting heart
`
`beats and body motions. (Ex. 1003 at Figs. 21A, 21B, 22A, 22B, 7:18-37; 11:52-
`
`60; 22:22-23:12; 23:29-24:4)(using 16 seconds as an example of the duration of an
`
`FFT processing interval). Amano discloses that the user’s steps and distance are
`
`calculated using FFT signal processing, which means that Amano performs the
`
`calculations using a time window of data and are not based upon data for a single
`
`step. (Ex. 1003 at 11:52-60, 23:35-57)(describing how steps are identified in the
`
`output of FFT processing applied to body motion sensor 302). In her declaration,
`
`Dr. Choudhury explains Amano’s disclosure that the body motion detector and
`
`CPU use FFT signal processing to detect the user’s steps and to determine the
`
`
`
`
`
`
`
`
`
`
`-11-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`user’s pitch while running. (Ex. 1005 at ¶¶93-94). Dr. Choudhury further explains
`
`how these calculations are used to calculate the distance traveled. (Ex. 1005 at
`
`¶¶106-108). Blackbird asked Dr. Choudhury to explain how Amano calculates the
`
`user’s pitch (i.e., stride rate), and she explained that Amano discloses using FFT
`
`signal processing and that FFT requires a time window of data:
`
`Q. Yes. What is your understanding in terms of how
`
`the calculation is done?
`
`A.
`
`There is signal, acceleration signal that is
`
`generated by movement of -- related to body motion
`
`which is capturing, is correlated with the steps. And
`
`based on a time window of acceleration data, FFT is
`
`done to estimate pitch.
`
`Q. You said a time window. What were you referring
`
`to when you said there was a time window?
`
`A.
`
`So it mentions Fast Fourier transform, FFT
`
`processing. For FFT processing, one would need to use
`
`a time window of data.
`
`(Ex. 2003 at 45:2-14)(emphasis added).
`
`Blackbird specifically asked Dr. Choudhury about its contention that Amano
`
`determines pitch based upon a single step, which she refuted:
`
`Q. Okay. Does Amano determine pitch based
`
`on a single step?
`
`
`
`
`
`
`
`
`
`
`-12-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`
`A.
`
`That would not be my understanding.
`
`Q. Why is that not your understanding?
`
`A.
`
`Because it's using FFT. You cannot get -- for that,
`
`you would need to have a time window of data. And
`
`usually, the process of how FFT works is you need at
`
`least multiple samples to get -- accurately estimate the
`
`rate of change.
`
`(Ex. 2003 at 45:21-46:4)(emphasis added).
`
`C. Blackbird’s expert did not address the fact that Amano uses FFT
`signal processing in its calculations.
`
`Despite Amano’s clear disclosure that FFT processing is used to perform the
`
`relevant calculations and the fact that Dr. Choudhury relied on the FFT disclosures
`
`in her declaration, Blackbird’s Response and expert declaration are silent on FFT
`
`processing and signal processing in general. In fact, FFT is not even mentioned in
`
`either document.
`
`In his deposition, Blackbird’s expert admitted that “signal processing in
`
`copious amounts” is relevant to the ’212 Patent and professed to having expertise
`
`in such techniques. (Ex. 1021 at 21:1-12, 124:14-25). Despite this, not only did
`
`Dr. Caloyannides’ declaration fail to address the FFT-related opinions from Dr.
`
`Choudhury’s declaration, he admitted that he did not even consider Dr.
`
`
`
`
`
`
`
`
`
`
`-13-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Choudhury’s deposition testimony when forming his own opinions. (Ex. 1021 at
`
`11:11-16).
`
`Nevertheless, Dr. Caloyannides made a number of admissions that support
`
`Dr. Choudhury’s FFT and signal processing-related opinions. Amano explains that
`
`it is not limited to using FFT processing, and that “it is also acceptable to employ
`
`MEM analysis, wavelet analysis or the like.” (Ex. 1003 at 15:66-16:4). Under
`
`questioning, Dr. Caloyannides admitted that wavelet analysis is a type of signal
`
`processing that could be used in the field of pedometers, health trackers and fitness
`
`trackers at the time of the invention. (Ex. 1021 at 126:3-11). Dr. Caloyannides
`
`also explained that such techniques use data from a time domain, and that a person
`
`of ordinary skill would prefer to have data for a longer time rather than a shorter
`
`time period. (See Ex. 1021 at 131:1-21). Dr. Caloyannides also admitted that a
`
`POSITA would prefer to have more data – rather than less – for performing any
`
`signal processing technique:
`
`Q.
`
`So a person of ordinary skill in the art using a
`
`technique such as wavelet processing would prefer to use
`
`as much data that is available?
`
`A. Yes.
`
`Q. Okay. And would that same consideration apply
`
`for maximum entropy techniques?
`
`
`
`
`
`
`
`
`
`
`-14-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`
`A.
`
`For any techniques. The more raw data you have,
`
`the better.
`
`(Ex. 1021 at 132:16-25). Finally, Dr. Caloyannides confirmed that too short
`
`of a time period can produce unreliable information:
`
`Q.
`
`-- can you perform this analysis on a single -- the
`
`single smallest observation or measurement?
`
`A. When you say "this analysis," which one is this
`
`analysis?
`
`Q. A wavelet analysis.
`
`A. Well, you can perform some kind of an analysis in
`
`a sample which is very short in time. It would be not
`
`very meaningful. . . .
`
`
`
`So, yes, you can perform an analysis. It might be
`
`laughable because of it's known limitations. So because
`
`of that, people usually do not perform an analysis of
`
`something which is so short that it makes no sense to
`
`make an analysis.
`
`(Ex. 1021 at 127:23-128:19).
`
`Thus, Blackbird’s expert confirms what Fitbit’s expert said – that signal
`
`processing techniques analyze data based upon a time window of data and that a
`
`POSITA would have not understood Amano’s disclosure to mean that the relevant
`
`calculation was performed on a single step. Because it is clear that Amano does
`
`
`
`
`
`
`
`
`
`
`-15-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`not perform calculations based upon a single step, but instead uses a time window
`
`of data to determine a distance traveled, the Board should find that claims 2 and 5
`
`are not only obvious but also anticipated by Amano.
`
`III. The combination of Kato and Amano renders the challenged claims
`obvious.
`
`Blackbird raises two general theories with respect to the Kato/Amano
`
`ground for claim 6: (1) that Kato does not teach calculating a distance traveled by
`
`multiplying the number of steps counted by a stride length that varies according to
`
`a rate at which steps are taken, and (2) that Fitbit has not shown that a POSITA
`
`would have been motivated to combine the teachings of Amano and Kato.
`
`Blackbird’s first argument is a rehash of an argument from its Preliminary
`
`Patent Owner Response that the Board properly rejected in the Institution
`
`Decision. (PPOR at 43-46; ID at 24). As explained in the Petition and the
`
`Institution Decision, Kato teaches or at least suggests multiplying the number of
`
`steps counter by the step counter by a stride length. (Petition at 71-73; ID at 24).
`
`Additionally, Blackbird’s argument that Kato does not teach the claimed distance
`
`calculation is irrelevant because the claimed calculation is taught by Amano. As
`
`explained above, Amano also teaches the distance calculation required by claim 6.
`
`Blackbird’s second argument is also unavailing. Essentially, Blackbird
`
`argues that a POSITA would not have been motivated to combine the Amano and
`
`
`
`
`
`
`
`
`
`
`-16-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Kato references because the references are not sufficiently similar, that a POSITA
`
`would not have been motivated to substitute Amano’s teaching that the device
`
`should employ user-specific (e.g., customized) data instead of the generic data
`
`employed by Kato, and that a POSITA would not have viewed Amano’s teaching
`
`that the device should employ user-specific data as a way to provide more accurate
`
`measurements. (POR at 24-40).
`
`While Blackbird argues that Amano and Kato are “not especially similar,” a
`
`review of the references reveals that this position has no merit. (POR at 30). As
`
`set forth in the Petition, a POSITA would have been motivated to combine the
`
`teachings of Amano and Kato for numerous overlapping reasons. At a high level,
`
`both references are within the field of exercise science and both references
`
`recognize that a user’s stride length varies with the user’s stride rate. (Petition at
`
`58). More specifically, both references are concerned with accounting for changes
`
`in stride lengths and how such changes affect speed and distance calculations.
`
`(Id.) And even more specifically, a POSITA would have been motivated to use
`
`Amano’s data processing steps to solve the underlying problem being address in
`
`Kato – calculating a distance traveled by the user while taking into account varying
`
`stride lengths and rates – and doing so with user-specific stride measurements
`
`(taught by Amano) within the Kato system. (Petition at 74-75).
`
`
`
`
`
`
`
`
`
`
`-17-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Blackbird’s expert fails to provide any explanation for why a POSITA
`
`would think that using user-specific stride length measurements would not result in
`
`more accurate stride length calculations as opposed to deriving the user’s stride
`
`length from the user’s height. As explained in Dr. Choudhury’s declaration, Kato
`
`calculates the user’s stride length based upon the user’s stride rate and user’s
`
`height. (Ex. 1005 at ¶81). The formula Kato uses for this calculation is derived
`
`from observing 142 test subjects that walked on a treadmill. (Ex. 1004 at 4:7-49).
`
`In contrast, Amano teaches measuring the specific user’s stride length at different
`
`stride rates, and then interpolating between the user’s stride length-stride rate data
`
`points to calculate the user’s stride length at other stride rates. (Ex. 1005 at ¶109).
`
`Thus, modifying Kato’s distance calculation to use Amano’s calculation (one that
`
`is based on the user’s own actual stride length) would have been desirable because
`
`it would allow a device to more accurately determine the stride length of a
`
`particular user. (Petition at 75, Ex. 1005 at ¶150).
`
` Faced with these well-articulated motivations to combine, Blackbird tries to
`
`deflect by arguing that Kato’s teaching of a relationship between stride rate, stride
`
`length, and height precludes a POSITA from looking to Amano’s teaching. (POR
`
`at 33-37). But this argument only reinforces Fitbit’s position: that a POSITA
`
`looking at Kato’s use of a formula to calculate the user’s stride length based upon
`
`
`
`
`
`
`
`
`
`
`-18-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`the user’s height and stride rate would have been motivated to replace the
`
`calculation with Amano’s calculation that uses actual measurements of the user’s
`
`stride length. (Petition at 75). Blackbird is also wrong in stating that Amano does
`
`not consider the relationship of stride length, stride rate, and height. (See Ex. 1003
`
`at 16:6-14)(determining stride length indirectly obtained from height).
`
`IV. The claimed “step counter” must be mounted on the waist, chest, or leg.
`
`The Institution Decision incorrectly states that the “Petitioner proposes ‘step
`
`counter’ should be interpreted as proffered in the District Court litigation — ‘a
`
`device that collects data to generate step count…’” and that the parties agree to this
`
`construction.3 (ID at 7-8). In its Petition, Fitbit proposed a narrower construction
`
`for “step counter” that is “a device mounted to the chest, waist, or leg that counts
`
`the number of steps a user takes.” (Petition at 27). In contrast, Blackbird proposed
`
`in the district court the broader construction that “step counter” means “a device
`
`that collects data to generate step counter.” (Id.; PPOR at 15).
`
`Fitbit requests that the Board adopt Fitbit’s proposed construction for “step
`
`counter” or, in the alternative, find that it is unnecessary to construe the term “step
`
`
`3 Fitbit believes this confusion may have been the result of another entity,
`
`TomTom, filing IPRs on the ’212 Patent close in time to Fitbit’s IPR. See
`
`IPR2017-02023 and IPR2017-02025.
`
`
`
`
`
`
`
`
`
`
`-19-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`counter” to decide this inter partes review because Fitbit has demonstrated for each
`
`ground of invalidity that the “step counter” is mounted to the chest, waist, or leg.
`
`(See Petition at 28). Although Blackbird argued in its Preliminary Response that
`
`Amano Grounds 1 and 2 do not disclose or teach a “step counter” mounted to the
`
`chest, waist, or leg (PPOR at 31-36 and 40-42), the Board considered and rejected
`
`Blackbird’s arguments in the Institution Decision. (See ID at 20-22). In addition,
`
`Blackbird did not raise the issue its Response, and Blackbird’s expert did not offer
`
`any opinions that are based on Fitbit’s construction of “step counter.” (Ex. 1021 at
`
`102:2-18).
`
`V. This inter partes review proceeding was properly instituted under the
`governing statutes.
`
`35 U.S.C. §314(a) sets the threshold test for instituting inter partes review: a
`
`determination by the Board “that the information presented in the petition under
`
`section 311 and any response filed under section 313 shows that there is a
`
`reasonable likelihood that the petitioner would prevail with respect to at least 1 of
`
`the claims challenged in the petition.” The Board’s timely Institution Decision
`
`found “that the Petitioner has demonstrated a reasonable likelihood of prevailing
`
`on the challenge to claims 2 and 5 as obvious over Amano” (ID at 22) and that
`
`“independent claim 6 would have been obvious over Kato and Amano” (ID at 28).
`
`The Institution Decision gave the required notice of the Board’s determination that
`
`
`
`
`
`
`
`
`
`
`-20-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`the threshold for institution as required by §314(c) was met by the petition. By
`
`operation of the plain-language of the statute, the Board’s decision effectively
`
`instituted inter partes review on all grounds presented in the petition because the
`
`Supreme Court has held that the Board can only institute the petition in its entirety
`
`or not at all. See SAS Inst., Inc. v. Iancu, 138 S. Ct. 1348, 1354-56 (2018); see also
`
`PGS Geophysical AS v. Iancu, 891 F.3d 1354, 1360 (Fed. Cir. 2018) (interpreting
`
`SAS “to require a simple yes-or-no institution choice” for all challenges); Eset,
`
`LLC v. Finjan, Inc., IPR2017-01738, Paper 28 (PTAB Aug. 10, 2018) (denying
`
`rehearing request to partially institute or terminate because “petitioner’s petition,
`
`not the Director’s discretion, is supposed to guide the life of the litigation”)
`
`(quoting Adidas AG v. Nike, Inc., 894 F.3d 1256 (Fed. Cir. 2018)).
`
`Contrary to Blackbird’s argument, 37 C.F.R. §42.5 authorized the Board to
`
`issue its later order permitting consideration of the petition’s anticipation ground
`
`because the Supreme Court’s decision that the Board may only grant or deny
`
`institution of a petition in full is an intervening change in law that could not have
`
`been contemplated by the existing regulations. In addition, the Federal Circuit has
`
`indicated that the Board must consider all challenged claims and all challenged
`
`grounds in the petition. See BioDelivery Scis. Int’l, Inc. v. Aquestive Therapeutics,
`
`
`
`
`
`
`
`
`
`
`-21-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`Inc., 898 F.3d 1205, 1210 (Fed. Cir. 2018) (“[E]ven if a prior action did not appear
`
`unlawful at the time, this does not insulate it from corrective action.”).
`
`VI. The Constitution does not bar the application of the inter partes review
`process to the ’212 Patent.
`
`Blackbird’s argument that the application of inter partes review to a patent
`
`that issued before the enactment of the America Invents Act (AIA) lacks merit for
`
`three reasons. First, inter partes review only provides prospective relief. In
`
`Landgraf v. USI Film Products, the Supreme Court noted that “[w]hen the
`
`intervening statute authorizes or affects the propriety of prospective relief,
`
`application of the new provision is not retroactive.” 511 U.S. 244, 273 (1994).
`
`Inter partes review provides only prospective relief – a decision that one or more
`
`claims are unpatentable, and therefore is not a retroactive application of law.
`
`Second, inter partes review does not qualify as a retroactive application of law
`
`because the intervening change in law that created inter partes review merely
`
`effected procedural changes over what was previously available under the law. As
`
`Landgraf explained, changes of procedure do not effect a retroactive change
`
`“[b]ecause rules of procedure regulate secondary rather than primary conduct.”
`
`511 U.S. at 275. Applied here, the substance of an inter partes review is the same
`
`as the substance of a pre-existing inter partes reexamination–i.e., the Patent Office
`
`can re-consider patentability using printed prior art under 35 U.S.C. §§ 102 and
`
`
`
`
`
`
`
`
`
`
`-22-
`
`
`
`
`
`
`
`
`IPR2017-02012
`Fitbit’s Reply
`
`103. The differences in proce