`
`
`
`
`
`
`
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`BAYERISCHE MOTOREN WERKE AKTIENGESELLSCHAFT and
`BMW OF NORTH AMERICA, LLC,
`Petitioners,
`
`v.
`
`PAICE LLC and THE ABELL FOUNDATION, INC.,
`Patent Owners.
`
`
`
`
`
`
`
`
`
`
`
`
`Case IPR2020-01386
`Patent 7,237,634
`
`
`
`
`
`
`
`
`
`
`
`
`PATENT OWNERS’ SUR-REPLY
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`TABLE OF CONTENTS
`
`
`
`
`
`B.
`
`C.
`
`A. Ground 1 – Severinsky in View of Nii Does Not Render Claim 33
`Obvious ................................................................................................. 1
`1. The prior art does not teach or suggest “varying said setpoint
`accordingly” ................................................................................ 2
`2. BMW’s reasons to combine are flawed .................................... 10
`Ground 2 – Severinsky in View of Quigley Does Not Render Claim
`33 Obvious .......................................................................................... 16
`Ground 3 – Severinsky in View of Graf Does Not Render Claim 33
`Obvious ............................................................................................... 21
`1. Severinsky in view of Graf does not teach or suggest “varying
`said setpoint accordingly” ......................................................... 21
`2. Severinsky in view of Graf does not teach or suggest
`“monitoring patterns of vehicle operation over time” .............. 21
`D. Grounds 4-9 ......................................................................................... 23
`E.
`Grounds 7, 10, and 11 – Severinsky in View of Ma Does Not Render
`Claims 49, 105, and 188 Obvious ....................................................... 23
`Grounds 12-14 ..................................................................................... 28
`
`
`F.
`
`
`
`i
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`EXHIBIT LIST
`
`Exhibit No.
`PAICE 2001
`
`PAICE 2002
`
`PAICE 2003
`
`PAICE 2004
`PAICE 2005
`PAICE 2006
`
`PAICE 2007
`PAICE 2008
`PAICE 2009
`
`PAICE 2010
`PAICE 2011
`PAICE 2012
`PAICE 2013
`PAICE 2014
`PAICE 2015
`PAICE 2016
`PAICE 2017
`PAICE 2018
`PAICE 2019
`
`PAICE 2020
`
`Description
`Patent Owners’ Preliminary Response to Petition for IPR in
`IPR2015-0722 Dated August 10, 2015
`Patent Owner’s Preliminary Response to Petition in IPR2015-
`00787 dated August 10, 2015
`Patent Owner’s Preliminary Response to Petition in IPR2015-
`00791 dated August 10, 2015
`Statutory Disclaimer
`Reserved
`Response to Interrogatory 27, BMW Responses to PAICE 1st
`Set of Interrogatories [1-28] dated May 6, 2020
`Paice/Toyota Complaint dated June 8, 2004
`Paice/Toyota Amended Complaint dated July 3, 2007
`Scheduling Order [Docket No. 36] from 1:19-cv-03348-SAG
`(USDC-DMD) dated February 25, 2020
`Docket Navigator Statistics – Top Patents by Number of IPRs
`Ex. J to BMW Invalidity Contentions dated June 8, 2020
`Reserved
`IPR2017-00232 Petition dated November 14, 2016
`UK Patent Application GB 2,318,105 Cover Page
`Printout of http://www.paicehybrid.com/licensing-agreements/
`Declaration of Mahdi Shahbakhti, Ph.D.
`Curriculum Vitae of Mahdi Shahbakhti, Ph.D.
`Bosch Gasoline-engine Management
`Selected Pages From John Heywood, Internal Combustion
`Engines Fundamentals
`Selected Pages From Merhdad Ehsani et al, Modern Electric,
`Hybrid Electric, and Fuel Cell Vehicles
`PAICE 2021 Matthew Cuddy et al., Analysis of the Fuel Economy Benefit
`of Drivetrain Hybridization
`Selected Pages From Draft Technical Assessment Report:
`Midterm Evaluation of Light-Duty Vehicle Greenhouse Gas
`Emission Standards and Corporate Average Fuel Economy
`Standards for Model Years 2022-2025
`Selected Pages From Assessment of Fuel Economy of Fuel
`Economy Technologies for Light Duty Vehicles
`
`PAICE 2022
`
`PAICE 2023
`
`ii
`
`
`
`
`
`PAICE 2024
`
`PAICE 2025
`PAICE 2026
`PAICE 2027
`PAICE 2028
`
`PAICE 2029
`PAICE 2030
`PAICE 2031
`PAICE 2032
`PAICE 2033
`PAICE 2034
`
`PAICE 2035
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`Selected Pages From Richard Stone, Introduction to Internal
`Combustion Engines
`Reserved
`Reserved
`Reserved
`Guzzella et al., Introduction to Modeling Control of Internal
`Combustion Engine Systems
`Reserved
`Reserved
`Reserved
`Reserved
`Guzzella et al., Vehicle Propulsion Systems
`Bumby, J.R. et al., “Optimisation and control of a hybrid
`electric car,” IEE PROCEEDINGS, Vol. 134, Pt. D, No. 6
`(Nov. 1987), 373-87 (“Bumby II”)
`June 16, 2021 Deposition Transcript of Dr. Gregory Davis in
`IPR2020-00994
`
`iii
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`I.
`
`Introduction
`BMW’s reply confirms that all challenged claims of the ’634 patent are
`
`patentable. Despite BMW’s repeated attempts to backfill its deficient Petition,
`
`significant gaps remain in the prior art. BMW’s reply continues to beat the same
`
`drum about “bodily incorporation.” But obviousness is BMW’s burden, and
`
`BMW’s incorrect, oversimplified, and unexplained reasons to combine fail to
`
`explain how and why a POSA would combine its proposed references.
`
`II. BMW’s Grounds Fail To Demonstrate Obviousness
`A. Ground 1 – Severinsky in View of Nii Does Not Render Claim 33
`Obvious
`Claim 33 specifically requires “varying said setpoint”—a parameter the
`
`hybrid controller uses for selecting operating modes—based on “monitoring
`
`patterns of vehicle operation over time.” BMW cannot show that either Severinky
`
`or Nii vary anything resembling a “setpoint” or even contemplate how the
`
`controller could vary such a control parameter based on monitored vehicle patterns
`
`over time. Severinsky does not vary any alleged “setpoint,” and Nii is concerned
`
`with setting the engine at a constant output for battery charging.
`
`In lieu of evidence, BMW offers sweeping statements about how pattern
`
`information is good for enhancing vehicle efficiency. But the claims are not so
`
`broad, and BMW cannot satisfy its burden by disregarding key claim elements.
`
`1
`
`
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`1.
`
`The prior art does not teach or suggest “varying said
`setpoint accordingly”
`Neither Severinsky nor Nii discloses a hybrid controller that varies the
`
`“setpoint,” and BMW cannot fill this gap with unsupported expert testimony.
`
`BMW does not dispute that Nii fails to disclose a “setpoint” and that Nii’s
`
`teachings are unrelated to selecting between electric motor and engine propulsion
`
`modes. Moreover, common sense and record evidence belie BMW’s assertion that
`
`Seversinky discloses varying the claimed setpoint.
`
`First, BMW again incorrectly argues that Severinsky varies the setpoint by
`
`“operating its engine ‘outside its most fuel efficient operating range, on occasion.’”
`
`BMW Reply, 9 (quoting BMW1013, 18:23-25.) BMW confuses the claimed
`
`“setpoint” and the actual operating point of the engine. POR, 25-26; PAICE2016,
`
`¶¶75-77. The “setpoint” is a parameter used by the controller to select the mode
`
`for operating the engine or motor. (PAICE2035 (Davis Tr.), 36:17 – 37:2, 38:1-7.)
`
`A shown by Figure 9, the controller compares road load to the setpoint to select
`
`motor propulsion mode or engine propulsion mode. POR, 25-26; PAICE2035,
`
`14:13 – 15:5. It does not compare road load to the actual operating point of the
`
`engine as Dr. Davis admitted. (PAICE2035, 49:12-17.)
`
`The operating point of the engine—which BMW confuses with the
`
`“setpoint”—is the amount of torque the engine is actually producing at a given
`
`time. POR, 26-27. The claimed “setpoint,” however, is the control parameter for
`
`2
`
`
`
`
`triggering the engine to come on. Once the engine is on, it can operate at any point
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`between 1% and 100% MTO independent of the claimed “setpoint,” which BMW
`
`does not dispute. Dr. Davis admitted that the 30% MTO setpoint (in red below) in
`
`Figure 7 of the ’634 patent stays fixed while the actual operating point of the
`
`engine, i.e., the “Engine Torque Output” (shown in the dashed lines) fluctuates
`
`above and below the setpoint. (PAICE2035, 17:14 – 18:6; see also id., 15:9 –
`
`16:1, 16:10 – 17:13; PAICE2016, ¶77.)
`
`
`
`(BMW1001, Fig. 7(a) (annotated).)
`
`Moreover, the claims distinguish between the operating point of the engine
`
`(in blue) and the setpoint (in red).
`
`55. The method of claim 33, further comprising: operating the engine
`at torque output levels less than SP under abnormal and transient
`conditions to satisfy drivability and/or safety considerations.
`
`3
`
`
`
`
`BMW1001, 63:18-20 (emphasis added). Severinsky’s choice to operate the engine
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`below the 60% MTO setpoint does not mean that Severinsky varies the setpoint. It
`
`means Severinsky operates the engine inefficiently in certain circumstances (i.e.,
`
`below 60% MTO) because other factors require that the engine stay on. POR, 25.
`
`Indeed, the engine can turn on (and operate) for many reasons such as safety
`
`considerations (as contemplated in claim 55) or because the battery state of charge
`
`is too low. (PAICE2016, ¶65.) This is not evidence of varying a torque-based
`
`setpoint. As Dr. Shahbakhti explained, hybrid control systems arbitrate between
`
`different algorithms for turning the engine on/off. (Id., ¶72.) Indeed, even the
`
`Board previously found that Severinsky uses two distinct control algorithms—a
`
`torque algorithm and “speed-responsive hysteresis” algorithm—for turning the
`
`engine on/off. (BMW1059, 17.)
`
`Second, BMW’s argument that Severinsky’s separate on/off speed
`
`thresholds are evidence of varying the setpoint is incorrect. Regardless of the fact
`
`that these are speed thresholds (as opposed to a torque “setpoint”), they are never
`
`varied—they are “factory-set parameters” as BMW admits. Reply, 4.
`
`Claim 33 requires a “monitoring patterns of vehicle operation over time and
`
`varying the SP accordingly.” The word “accordingly” indicates that varying the
`
`setpoint occurs based on vehicle monitoring over time. In other words, the
`
`variation takes place in response to the vehicle’s previous activity. BMW,
`
`4
`
`
`
`
`however, provides no evidence that Severinsky’s controller selects between the
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`engine on and engine off speed thresholds during vehicle operation. BMW admits
`
`these speed thresholds are “factory-set parameters” (Reply, 4)—they are written
`
`into the source code and do not change. (PAICE2016, ¶¶60-61.)
`
`Moreover, Severinsky’s separate on/off speed thresholds are not evidence of
`
`any type of varying. Severinsky’s upper speed range (i.e., “30-35 mph”) and lower
`
`speed range (i.e., “20-25 mph”) are two separate speed thresholds much like a
`
`thermostat may turn an air conditioner on at 75 degrees and turn the air conditioner
`
`off at 74 degrees. POR, 21; PAICE2016, ¶¶61-62. BMW’s insinuation that
`
`Severinsky’s controller “switch[es] between two (or more) values that represent
`
`the ‘setpoint’ at any given time” is wrong. Reply, 11. Severinsky does not switch
`
`between setpoints. Severinsky’s controller always uses the upper “30-35 mph”
`
`speed threshold to turn the engine on and always uses the lower “20-25 mph” to
`
`5
`
`
`
`
`turn the engine off.1 (PAICE2016, ¶¶61-62.)2 In other words, Severinsky uses
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`different thresholds for different purposes. There is no variation or dynamic
`
`“switching” of the different thresholds to speak of.
`
`Claim 39 of the ‘634 patent (which depends from claim 33) in fact
`
`separately recites the use of two fixed setpoints, SP and SP2:
`
`39. The method of claim 33, further comprising: monitoring the RL
`over time; wherein said operating the internal combustion engine to
`propel the hybrid vehicle is performed when: … or the RL>a second
`setpoint (SP2), wherein the SP2 is a larger percentage of the MTO
`than the SP.
`
`(BMW1001, 61:42-48 (emphasis added).)3
`
`
`1 BMW falsely states that Patent Owners concede that “Severinsky uses ‘two
`
`separate speed thresholds’ for turning off the engine.” Reply, 11. But Patent
`
`Owners stated that Severinsky discloses a speed threshold for turning the engine
`
`on and a speed threshold for turning the engine off. POR, 21.
`
`2 Dr. Davis admitted that the “30-35 mph” threshold is for turning the engine on
`
`and the “20-25 mph” threshold is for turning the engine off. (PAICE2035, 7:2-7,
`
`8:24 – 9:7.)
`
`3 BMW’s assertion that the ’634 patent uses “hysteresis” to “vary[] said setpoint”
`
`is false. Consistent with claim 39, the ’634 patent uses different setpoints for
`
`6
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`Finally, BMW’s reply relies on speed thresholds in an attempt to show
`
`
`
`varying the claimed torque setpoint. POR, 21-23. BMW’s reply shies away from
`
`the single 60% MTO value that the Board found is the “setpoint,” and its new and
`
`baseless assertions that Severinsky’s “speed-responsive hysteresis” is a
`
`“misnomer” and “‘speed thresholds’ are also torque thresholds by their very
`
`nature” have no foundation in reality, much less Severinsky’s disclosure. Reply,
`
`11. But BMW does not dispute that speed and torque are independent variables or
`
`Dr. Shahbakhti’s example of how the instantaneous torque may vary between -80
`
`and 180 Nm for a single fixed speed. (PAICE, ¶¶70-71.)
`
`
`different purposes. (BMW1001, 41:4-47.) This is not evidence of varying the
`
`setpoint. These two setpoints are written into source code during the vehicle
`
`design phase and are not based on monitoring vehicle operation over time.
`
`(PAICE2016, ¶¶61-62.)
`
`7
`
`
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`
`
`(Id., ¶¶70-71.) It makes no sense to have a single torque/speed setpoint for two
`
`independent variables that vary drastically. BMW provides no cogent explanation
`
`for why Severinsky would not use parallel control systems to separately control for
`
`separate variables.
`
`Severinsky’s speed thresholds are written in units of speed (miles per hour),
`
`not units of torque. BMW cannot rewrite Severinsky, and Dr. Davis’s identical
`
`statements are “[u]ntethered to any supporting evidence” and cannot support an
`
`obviousness finding. TQ Delta, LLC v. CISCO Sys., Inc., 942 F.3d 1352, 1362
`
`(Fed. Cir. 2019) (rejecting expert’s “ipse dixit declaration” that is “[u]ntethered to
`
`8
`
`
`
`
`any supporting evidence…”).4 The Board previously found Severinsky’s 60%
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`MTO value is the claimed “setpoint” (BMW1059, 13-14), and BMW has failed to
`
`show that Severinsky’s “speed-responsive hysteresis” would vary this value.5
`
`Patent Owners are not “relitigating” anything by pointing out that
`
`Severinsky’s “speed-responsive hysteresis” is separate from the torque “setpoint”
`
`(and related control algorithm) that the Board previously identified. The Board did
`
`not find that Severinsky’s speed thresholds were the claimed “setpoint.” The
`
`Board instead stated “that Severinsky may teach an additional hysteresis feature as
`
`
`4 BMW’s new argument that speed thresholds are torque thresholds goes beyond
`
`the proper scope of the Reply. BMW relied on “speed-responsive hysteresis” for a
`
`claim about torque and should have presented argument attempting to connect
`
`these different concepts in its Petition. TPG, 74 (Nov. 2019); Ariosa Diagnostics
`
`v. Verinata Health, Inc., 805 F.3d 1359, 1367 (Fed. Cir. 2015).
`
`5 BMW relies on a secondary reference (Frank) to argue it would be obvious for
`
`Severinsky to use two different setpoints rather than a single 60% MTO setpoint,
`
`calling into question BMW’s new theory that Severinsky’s separate on/off speed
`
`thresholds are really two torque setpoints in disguise. Petition, 47-49, 51 (“It
`
`would have been a simple design choice to apply Frank’s hysteresis strategies to
`
`Severinsky’s 60% of MTO torque ‘setpoint’….”).
`
`9
`
`
`
`
`a way of controlling unintended engine starts during temporary dips in speed does
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`not preclude Severinsky from also teaching the use of a torque value….”
`
`(BMW1059, 17). In other words, the Board found that Severinsky uses two
`
`distinct control algorithms for turning the engine on/off and the “speed-responsive
`
`hysteresis” algorithm did not preclude use of the torque algorithm.
`
`2.
`BMW’s reasons to combine are flawed
`BMW’s defense of its reasons to combine—consisting of vacillations and
`
`contradictions—solidify the weakness of its combination. BMW’s mystifying
`
`discussion of its proposed combination has left even BMW confused as to how it
`
`intends to modify Severinsky based on Nii’s “pattern information.” BMW both
`
`argues that “Petitioners do not propose that average power consumption be used to
`
`turn Severinsky’s engine on or off” (Reply, 7) and that “Nii’s … average power
`
`requirement could [] be used to adjust Severinsky’s setpoint for turning off the
`
`engine.” Id. (emphasis added). BMW cannot carry its burden of providing an
`
`“explanation as to how or why the references would be combined to produce the
`
`claimed invention.” TriVascular, Inc. v. Samuels, 812 F.3d 1056, 1066 (Fed. Cir.
`
`2016).
`
` BMW cannot rehabilitate its defective reason to combine. For all the
`
`many things that BMW does say, BMW’s silence is more revealing. BMW does
`
`not dispute that:
`
`10
`
`
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
` Nii does not disclose a “setpoint” or selecting between electric motor
`
`propulsion and engine propulsion;
`
` Nii’s engine is always decoupled from the road wheels and the instantaneous
`
`torque demand;
`
` Nii only uses “pattern information” to set the generator output for battery
`
`charging;
`
` Nii’s “pattern information” does not include instantaneous vehicle
`
`requirement information; and
`
` Nii does not teach any “pattern information” other than average power.
`
`BMW’s claim that “knowing precisely how a vehicle will actually be operated is
`
`the ‘holy grail’ for fine-tuning hybrid vehicle efficiency” fails to paper over these
`
`cracks. Reply, 3. The claims do not broadly claim fine-tuning a hybrid vehicle,
`
`and Nii does teach anything useful for “fine-tuning” Severinsky’s alleged setpoint.
`
`BMW’s “bodily incorporation” criticism rings hollow. BMW does not
`
`dispute that Nii solely uses its “pattern information” to set a generator output for
`
`battery charging in a series hybrid, which has nothing to do with varying the
`
`claimed “setpoint” for operating mode selection in a parallel hybrid. Nor does
`
`BMW articulate what “pattern information” a POSA would use from Nii other than
`
`average power. If BMW could have identified any relevant disclosures in Nii, it
`
`would have. Yet, BMW still cannot cogently explain its combination.
`
`11
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`First, BMW’s philosophical debate about whether it is possible to combine
`
`
`
`series and parallel hybrid vehicles is inapposite. Patent Owners have not argued
`
`that series and parallel hybrid vehicles can never be combined. And BMW’s
`
`assertion that its combination is sound because series and parallel hybrid vehicles
`
`are both “concerned with enhancing hybrid vehicle efficiency” is tantamount to
`
`saying a POSA would add the rotor blades of a helicopter to a hot air balloon
`
`because both are concerned with flying. Reply, 5. This oversimplified analysis
`
`cannot demonstrate obviousness.
`
`BMW’s discussion about operating modes in series and parallel is also
`
`irrelevant. Nii has no operating modes related to selecting different sources of
`
`vehicle propulsion. Moreover, BMW’s discussion about operating modes is
`
`supported solely by Dr. Davis’s testimony that it is possible for series and parallel
`
`vehicles to use a “Max. SOC-PPS control strategy” (Reply, 5 (citing BMW1088,
`
`¶¶38-44)), which “emphasizes maintaining the SOC.” (PAICE2020, 244.)6 Yet,
`
`
`6 BMW’s criticism of Dr. Shahbakhti’s use of the Ehsani textbook from 2005 is
`
`unwarranted. Dr. Shahbakhti uses the Ehsani textbook to illustrate how series and
`
`parallel hybrid architectures are configured and not to form opinions about the
`
`state of the art in 1998. BMW also improperly relies on Sundance, Inc. v.
`
`DeMonte Fabricating Ltd., 550 F.3d 1356, 1363 (Fed. Cir. 2008) to suggest that
`
`12
`
`
`
`
`BMW does not assert that “Max. SOC-PPS control strategy” has any relevance to
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`the claimed setpoint or that Severinsky or Nii use any control strategy resembling
`
`“Max. SOC-PPS.” BMW’s detour into other hypothetical combinations not before
`
`the Board is irrelevant.7
`
`Second, BMW’s attempt to identify similarities between Severinsky and Nii
`
`is based on the false assertion that Nii’s engine output is based on the
`
`instantaneous torque requirements. Reply, 6. Nii uses “travel patterns” to isolate
`
`
`Dr. Shahbakhti’s age makes his testimony less reliable. But Sundance is about
`
`excluding a patent law expert (with no technical expertise) from opining on matters
`
`of infringement and invalidity. Id. at 1365.
`
`7 BMW’s misrepresents that Dr. Shahbakhti provided “counter-testimony.” (Reply,
`
`6.) Dr. Shahbakhti previously opined that it would have been obvious to modify a
`
`parallel hybrid by adding braking grids connected to the traction motor for
`
`absorbing dynamic braking energy when the battery cannot receive all of the
`
`energy. BMW1098, ¶75 (“a [POSA] would have been motivated to modify Kira’s
`
`drive system to incorporate a dynamic braking grid as described by Kumar.”) This
`
`opinion has nothing to do with Dr. Shahbakhti’s present opinion that a POSA
`
`would not modify Severinsky’s “setpoint” based on Nii’s alleged “pattern
`
`information.”
`
`13
`
`
`
`
`the steady state engine operation from the driver demand, which is possible in
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`series vehicles where the engine is always decoupled from the wheels.
`
`(PAICE2016, ¶98.) This solution is antithetical to Severinsky, which switches
`
`operating modes based on the instantaneous demands of the driver. POR, 32.
`
`BMW’s claim that Nii’s engine changes based on the instantaneous torque
`
`requirement (Reply, 6) is at odds with the primary purpose of Nii’s invention.
`
`(PAICE2035, 22:5-7 (Dr. Davis admitting that “an overall goal [of Nii] might be to
`
`try and just operate the engine as [sic] single speed, single load operating point”).)8
`
`Indeed, even Dr. Davis admitted that engines in series vehicles (like Nii) are “not
`
`mechanically connected to the wheels and the engine is therefore controlled
`
`independently of driving conditions.” (BMW1008, ¶71 (emphasis added).)
`
`Third, BMW’s confusing explanation of how Severinsky could use Nii’s
`
`“average power requirement” information reiterates the incompatibility of the
`
`references. BMW again improperly conflates the claimed “setpoint” with the
`
`actual operating point of the engine by asserting that a POSA would use “average
`
`power consumption … to vary the ‘setpoint’ at which the engine is operated.”
`
`POR, 7; see also id. (arguing Severinsky’s “engine would be turned on … while
`
`
`8 The fact that Nii can turn the engine off under high SOC does not make Nii’s
`
`controller dependent on road load.
`
`14
`
`
`
`
`adjusting its setpoints to improve engine efficiency” ). As discussed above in
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`Section II.A.1, the claimed “setpoint” is not the same as the actual operating point
`
`of the engine, and BMW’s reliance on this false premise underscores the
`
`deficiency of its combination.
`
`BMW’s alternative, inconsistent hypothetical—that Nii’s “recognition of
`
`vehicle travel patterns or average power requirements could each be used to adjust
`
`Severinsky’s setpoint for turning off the engine during hysteresis sooner …”
`
`(Reply, 7)—is unsupported and unintelligible. BMW does not explain whether this
`
`“average power requirement” is low, high, or somewhere in between. Nor does
`
`BMW explain what “vehicle travel patterns” it is referring to or how either would
`
`increase fuel economy.
`
`BMW cannot make sense of its proposed combination because Nii’s
`
`“average power requirement” is not useful in adjusting Severinsky’s setpoint.
`
`BMW does not dispute that instantaneous torque cannot be derived from average
`
`power or other averages. POR, 35-40. Instead, BMW explains how time averages
`
`are calculated, which is a mathematical operation that is not in dispute. Reply, 8-9
`
`(arguing high instantaneous torque values would result in a high average value).
`
`As the POR explains, the instantaneous torque requirement could be very high
`
`15
`
`
`
`
`during periods of operation even if there is a low average power requirement (and
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`vice versa), and BMW has no rebuttal on this point. Reply, 8-9.9
`
`Finally, BMW’s suggestion that the ’634 patent uses average values is false.
`
`The ’634 patent identifies a pattern where the road load “var[ies] between 0 and
`
`50% of MTO,” i.e., it moves above and below the setpoint, and adjusts the setpoint
`
`to 60% MTO accordingly. (BMW1001, 40:53-61.)10
`
`B. Ground 2 – Severinsky in View of Quigley Does Not Render
`Claim 33 Obvious
`BMW’s improper reply arguments cannot salvage its deficient ground.
`
`As the Board’s ID recognized, BMW’s “proposed combination appears to lack a
`
`rational underpinning.” ID, 40. The Board correctly explained that “Quigley and
`
`Dr. Davis’s testimony does not reveal evidentiary support for Petitioner’s
`
`contention that Quigley varies a setpoint that constitutes a ‘torque’ value.” ID, 40.
`
`Quigley’s “optimal operation” on which BMW relies (Petition, 26) has nothing to
`
`
`9 Dr. Davis’s hypothetical that Severinsky would turn the engine off “using the
`
`average power requirement” fails to address this problem. (BMW1088, ¶53.)
`
`10 BMW’s reliance on the ’634 patent’s separate discussion about “fluctuat[ions]”
`
`of road load cannot be considered an average as BMW claims. Reply, 8 (citing
`
`BMW1001, 41:14-29.)
`
`16
`
`
`
`
`do with adjusting the setpoint.11 And the Board remarked that BMW only
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`mentioned “Severinsky’s hysteresis mode in passing to support its contention.”
`
`Id.12
`
`In reply, BMW admits that its discussion of Severinsky’s hysteresis “was
`
`perhaps more brief” (Reply, 18) and backfills the gaps in the Petition. Id., 17-19.
`
`The reply is not the place for BMW to take a second shot at establishing a prima
`
`facie case of obviousness, and BMW’s reliance on 27 paragraphs from Dr. Davis’s
`
`reply declaration (BMW1088, ¶¶72-98) to remedy its deficient ground is improper.
`
`TPG, 73; Acceleration Bay, LLC v. Activision Blizzard Inc., 908 F.3d 765, 775
`
`(Fed. Cir. 2018). BMW’s belated attempt to explain its combination underscores
`
`the Board’s observation that BMW’s ground lacks a “rational underpinning.”
`
`Notwithstanding BMW’s improper reply arguments, Ground 2 remains
`
`lacking because Severinsky does not vary the “setpoint.” BMW fails to address
`
`
`11 Quigley states that its “intelligent controller” is for determining how to balance
`
`use of the engine and electric motor (when both the engine and motor are already
`
`on) and not for deciding when to employ either the electric motor or the engine.
`
`POR, 45-46.
`
`12 Severinsky’s “hysteresis mode” does not vary a setpoint and thus does not
`
`provide the missing rational underpinning. See Section II.A.
`
`17
`
`
`
`
`how a POSA would “vary[] said setpoint.” As set forth in Section II.A.1,
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`Severinsky’s “speed-responsive hysteresis” has nothing to do with the varying the
`
`alleged torque-based setpoint. (PAICE2016, ¶¶57-77.)
`
` Notwithstanding BMW’s improper reply arguments, there remains no
`
`reason to combine Severinsky and Quigley. First, BMW’s generic obviousness
`
`analysis—that a POSA would incorporate Quigley’s “intelligent” controller to
`
`“optimize[]” Severinsky’s setpoint to make Severinsky more efficient—cannot
`
`sustain an unpatentability finding. Reply, 15-16. Such vague assertions fail to
`
`sufficiently explain how or why a POSA could combine Severinsky and Quigley.
`
`TriVascular, 812 F.3d at 1066.
`
`BMW fails to identify a single “pattern” in Quigley that would be useful for
`
`improving Severinsky. In making its “bodily incorporation” argument, BMW
`
`criticizes Patent Owners for focusing on the “parameters” disclosed in Quigley
`
`(i.e., total distance and duration). But BMW’s Petition asserted that Quigley’s
`
`observation of “habitual usage characteristics” is for the purposes of “estimation of
`
`upcoming journey parameters [which] can allow ‘optimal operation with respect to
`
`18
`
`
`
`
`exhaust emissions and fuel consumption.’” Petition 27 (quoting Ex. 1054, 1).13
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`BMW further argued that a POSA would incorporate this same “optimal
`
`operation” into Severinsky. Id. Patent Owners can only respond to BMW.
`
`BMW does not explain how any of Quigley’s “parameters” would be useful
`
`to Severinsky, especially in regard to adjusting the alleged setpoint (i.e., 60%
`
`MTO). BMW does not dispute that these parameters have nothing to do with road
`
`load, the instantaneous torque required to propel the vehicle. Moreover, BMW
`
`presents no reason why Severinsky would use any of these parameters to modify
`
`the setpoint. Petition, 18.14 Ultimately, BMW does not explain how Severinsky
`
`would use any teaching from Quigley to improve emissions or fuel economy.
`
`Moreover, Dr. Davis admitted in IPR2020-01299 that “the only way you can
`
`intelligently modify the controller is based on an understanding of the load
`
`
`13 See also id., 29 (“A POSA would only need to modify Severinsky’s controller
`
`logic to use the information reflecting the monitored journey parameters as taught
`
`by Quigley….”).
`
`14 Total duration and distance could only be used to calculate average vehicle
`
`speed. And BMW does not dispute that average values (such as average speed) is
`
`not useful for “energy management” in “complex propulsion systems” such as
`
`Severinsky’s parallel hybrid. (Ex. 2033 (Guzzella) at 38.)
`
`19
`
`
`
`
`requirements of the vehicle” (IPR2020-01299, Ex. 1088, ¶17), and Quigley is
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`completely silent on that point.
`
`In Reply, BMW argues for the first time that Quigley “would have tracked
`
`information such as speed, engine speed, road speed and road load.” Reply, 21.15
`
`Yet, BMW fails to explain what would constitute the “pattern” under BMW’s
`
`interpretation, much less identify any disclosures in Quigley to support the notion
`
`that Quigley’s “patterns” include this information. Nor does BMW explain how
`
`Severinsky would use this information to optimize its controller. BMW argues that
`
`it does not have to identify any specific “patterns” in Quigley because the claims
`
`are not so specific. Reply, 22. But it is BMW’s burden to explain how a POSA
`
`would combine the prior art. It is not sufficient to show that separate claim
`
`elements were known. In re Kahn, 441 F.3d 977, 986 (Fed. Cir. 2006) (“[M]ere
`
`identification in the prior art of each element is insufficient to defeat the
`
`patentability of the combined subject matter as a whole.”).
`
`
`15 Quigley does not track these inputs and instead uses a “GPS data logger.”
`
`(BMW1054, 3.)
`
`20
`
`
`
`
`
`Attorney Docket: 36351-0018IP1
`Case IPR2020-01386
`
`C. Ground 3 – Severinsky in View of Graf Does Not Render Claim
`33 Obvious
`1.
`
`Severinsky in view of Graf does not teach or suggest
`“varying said setpoint accordingly”
`As with Grounds 1 and 2, none of the prior art “var[ies] the SP [setpoint].”
`
`Severinsky does not vary a “setpoint” (see section II.A.1), and “Graf’s ‘operating
`
`points’ are ‘the points at which the engine