`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`______________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`______________
`
`
`FORD MOTOR COMPANY
`
`Petitioner,
`
`v.
`
`PAICE LLC & ABELL FOUNDATION, INC.
`
`Patent Owner.
`
`______________
`
`U.S. Patent No. 7,104,347 to Severinsky et al.
`
`IPR Case No. IPR2014-00579
`
`______________
`
`
`
`
`
`DECLARATION OF DR. GREGORY DAVIS IN SUPPORT OF
`REPLY TO THE PATENT OWNER’S RESPONSE
`
`
`Page 1 of 52
`
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`
`Table of Contents
`
`
`
`Updated Exhibit List .............................................................................................................. 3
`
`I.
`
`Introduction ................................................................................................................. 6
`
`II. Motivation to Combine .............................................................................................. 6
`
`III. The Bumby References disclose a setpoint that is used for determining
`when to operate the engine in claims 1 and 23 ..................................................... 23
`
`A.
`
`B.
`
`Constant-power example illustrates how the Bumby references
`determine when to operate the engine/motor based on a torque
`setpoint ........................................................................................................... 23
`
`Even though the Bumby references consider “Demand Power”
`input, the Bumby references still use the “lower torque bound”
`setpoint to operate the engine/motor .......................................................... 26
`
`IV. The Bumby references disclose a first electric motor required by claim 1
`and claim 8 ................................................................................................................. 35
`
`A.
`
`B.
`
`Prior Art, including the AMPhibian Paper confirms that a
`conventional starter motor could accept current from the claimed
`battery ............................................................................................................... 35
`
`It was well-known to connect a conventional starter motor to a
`high-voltage battery ....................................................................................... 39
`
`V.
`
`Bumby references disclose efficiently using the engine to charge the
`battery ......................................................................................................................... 40
`
`A.
`
`B.
`
`C.
`
`
`
`The Bumby references do not teach away from the “battery
`charging mode” .............................................................................................. 40
`
`The Bumby references disclose operating the engine above
`setpoint to charge the battery, as in claims 1 and 23 ................................ 44
`
`The Bumby references disclose operating the engine to charge the
`battery when RL is less than SP, as in claim limitation [23.10] ................ 45
`
`
`
`Page 2 of 51
`
`
`
`FORD 1140
`
`

`

`
`
`
`
`Exhibit
`No.
`1101
`1102
`
`1103
`
`1104
`
`1105
`
`1106
`
`1107
`
`1108
`1109
`
`1110
`
`1111
`
`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`Updated Exhibit List
`
`Description
`U.S. Patent No. 7,104,347
`’347 Patent File History
`
`Bumby, J.R. et al. “Computer
`modelling of the automotive energy
`requirements for internal combustion
`engine and battery electric-powered
`vehicles” - IEE Proc. A 1985(5)
`Bumby, J.R. et al. “Optimisation and
`control of a hybrid electric car” - IEE
`Proc. A 1987, 134(6)
`Bumby, J.R. et al. “A Hybrid Internal
`Combustion Engine/Battery Electric
`Passenger Car for Petroleum
`Displacement” – Proc Instn Mech
`Engrs Volume 202 (D1), 51-65
`Bumby, J.R. et al. “A Test-Bed
`Facility for Hybrid IC Engine-Battery
`Electric Road Vehicle Drive Trains” -
`Trans Inst Meas & Cont 1988 Vol.
`10(2)
`Bumby, J.R. et al. “Integrated
`Microprocessor Control of a Hybrid
`i.c. Engine/Battery-Electric
`Automotive Power Train” - Trans
`Inst Meas & Cont 1990 Vol. 12:128
`Declaration of Gregory W. Davis
`U.S. Patent Trial and Appeal Board
`January 3, 2014 Decision (Appeal
`No. 2011-004811)
`Plaintiff Paice LLC’s Reply Claim
`Construction Brief (Case No. 2:04-
`cv-00211)
`Plaintiff Paice LLC’s Claim
`Construction Brief (Case No. 2:04-
`cv-00211)
`
`Date
`
`n/a
`n/a
`
`Sept. 1985
`
`Identifier
`The ’347 Patent
`’347 Patent File
`History
`Bumby I
`
`Nov. 1987
`
`Bumby II
`
`Jan. 1988
`
`Bumby III
`
`Apr. 1, 1988
`
`Bumby IV
`
`Jan. 1, 1990
`
`Bumby V
`
`
`Jan. 3, 2014
`
`
`n/a
`
`March 8, 2005
`
`Mar. 29, 2005
`
`
`
`
`
`Page 3 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`
`Exhibit
`No.
`1112
`
`1113
`
`1114
`
`1115
`
`1116
`
`1117
`
`1118
`1119
`
`1120
`1121
`1122
`
`1123
`1124
`
`1125
`
`1126
`
`1127
`
`1128
`
`1129
`
`Identifier
`
`Date
`Sept. 28, 2005
`
`June 25, 2008
`
`Aug. 1, 2008
`
`Dec. 5, 2008
`
`Nov. 14, 2013
`
`Dec. 16, 2013
`
`
`
`
`
`
`
`
`
`
`
`
`
`Description
`Claim Construction Order (Case No.
`2:04-cv-00211)
`Plaintiff Paice LLC’s Opening Claim
`Construction Brief (Case No. 2:07-
`cv-00180)
`Plaintiff Paice LLC’s Reply Brief on
`Claim Construction (Case No. 2:07-
`cv-00180)
`Claim Construction Order (Case No.
`2:07-cv-00180)
`Plaintiff Paice LLC and Abell
`Foundation, Inc.’s Opening Claim
`Construction Brief (Case No. 1:12-
`cv-00499)
`Plaintiff Paice LLC and Abell
`Foundation, Inc.’s Responsive Brief
`on Claim Construction (Case No.
`1:12-cv-00499)
`Curriculum Vitae of Gregory Davis
`Innovations in Design: 1993 Ford
`Hybrid Electric Vehicle Challenge
`1996 Future Car Challenge
`1997 Future Car Challenge
`History of the Electric Automobile –
`Hybrid Electric Vehicles
`Hybrid Vehicle for Fuel Economy
`Hybrid/Electric Vehicle Design
`Options and Evaluations
`Challenges for the Vehicle Tester in
`Characterizing Hybrid Electric
`Vehicles
`Electric and Hybrid Vehicles
`Program
`Technology for Electric and Hybrid
`Vehicles
`Strategies in Electric and Hybrid
`Vehicle Design
`Hybrid Vehicle Potential Assessment Sept. 30, 1979 Declaration Ex.
`
`
`Feb. 1994
`
`Feb. 1997
`Feb. 1998
`1998
`
`
`Feb. 24-28,
`1992
`April 9-11,
`1997
`
`Declaration Ex.
`Declaration Ex.
`
`Declaration Ex.
`Declaration Ex.
`Declaration Ex.
`
`Declaration Ex.
`Declaration Ex.
`
`Declaration Ex.
`
`April 1995
`
`Declaration Ex.
`
`Feb. 1998
`
`Declaration Ex.
`
`Feb. 1996
`
`Declaration Ex.
`
`Page 4 of 51
`
`
`
`FORD 1140
`
`

`

`
`
`Exhibit
`No.
`1130
`
`1131
`
`1132
`
`1133
`
`1134
`1135
`
`1136
`1137
`
`1138
`
`1139
`
`1140
`1141
`
`1142
`
`1143
`
`1144
`
`Description
`Final Report Hybrid Heat Engine /
`Electric Systems Study
`Transactions of the Institute of
`Measurements and Control: A
`microprocessor controlled gearbox
`for use in electric and hybrid-electric
`vehicles
`Propulsion System Design of Electric
`Vehicles
`Propulsion System Design of Electric
`and Hybrid Vehicles
`Bosch Handbook
`Design Innovations in Electric and
`Hybrid Electric Vehicles
`U.S. Patent No. 6,209,672
`Introduction to Automotive
`Powertrains (Davis Textbook)
`Yamaguchi article: Toyota Prius,
`Automotive Engineering
`International
`60/100095 Provisional Application
`
`Dr. Davis Reply Declaration
`Deposition Transcript of Mr.
`Hannemann IPR2014-00579
`Exhibit 12 from the April 8, 2015
`deposition of Mr. Hannemann
`Deposition Transcript of Mr.
`Hannemann IPR2014-00571
`U.S. Patent No. 5,285,862
`
`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`Date
`June 1, 1971
`
`Identifier
`Declaration Ex.
`
`Sept. 1, 1988 Declaration Ex.
`
`1996
`
`Declaration Ex.
`
`Feb. 1997
`
`Declaration Ex.
`
`Oct. 1996
`Feb. 1995
`
`Declaration Ex.
`Declaration Ex.
`
`Apr. 3, 2001
`
`
`Declaration Ex.
`Declaration Ex.
`
`Jan. 1998
`
`Declaration Ex.
`
`Declaration Ex.
`
`Filed Sept. 11,
`1998
`Davis Reply
`
`April 7-8, 2015 Hannemann
`Dep.
`
`
`April 8, 2015
`
`April 7, 2015 Hannemann
`Dep.
`’862 Patent
`
`Feb. 15, 1994
`
`Page 5 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`I.
`
`Introduction
`
`I, Gregory Davis, hereby declare as follows:
`
`1.
`
`I previously submitted a declaration on April 4, 2014 at the request of
`
`Ford Motor Company in the matter of Inter Partes Review of U.S. Patent No.
`
`7,104,347 (“the ’347 Patent”) to Severinsky et al.
`
`2. My previous declaration was based on my review of five prior art
`
`references referred to as Bumby I (Ex. 1103), Bumby II (Ex. 1104), Bumby III (Ex.
`
`1105), Bumby IV (Ex. 1106) and Bumby V (Ex. 1107).
`
`3.
`
`I provide this supplemental declaration in response to arguments
`
`presented by the Patent Owner.
`
`4.
`
`In addition to the exhibits I list in my initial declaration, I have also
`
`reviewed a thesis by Philip Masding titled “Some drive train control problems in
`
`hybrid i.c engine/battery electric vehicles” from Durham University. (Ex. 2104) It is
`
`my understanding that the Patent Owner provided Exhibit 2104 as part of its
`
`response. I will refer to this as the “Masding Thesis.”
`
`II. Motivation to Combine
`
`5.
`
`The Masding Thesis confirms my initial opinion that a person having
`
`ordinary skill in the art would have understood that Bumby I-V pertain to a hybrid
`
`project that was undertaken at the University of Durham in the late 1980’s. (see Ex.
`
`1108 at ¶¶175-233.)
`
`6.
`
`It is my opinion that the Masding Thesis builds upon the teachings of
`
`Page 6 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`the Bumby references (Bumby I-V) and even expressly references the work discussed
`
`in each Bumby publications.
`
`7.
`
`For example, the “Abstract” acknowledges that “All the fundamental
`
`systems needed to produce an operational vehicle have been developed and tested
`
`using a full sized experimental rig in the laboratory.” (Ex. 2104 [Masding Thesis] at 6.)
`
`8.
`
`Specifically, the Masding Thesis confirms that there existed two phases
`
`to the University of Durham hybrid vehicle project. The first phase pertained to
`
`evaluating control strategies for operating the hybrid vehicle. And the second phase
`
`involved developing a full scale parallel hybrid test rig to test the control strategies
`
`developed during the first phase.
`
`Such fundamental questions of how best to operate the vehicle were
`
`answered by the first phase of hybrid vehicle research carried out at
`
`Durham University. During this first phase a vehicle simulation package
`
`was developed which demonstrated the potential of the parallel hybrid
`
`and used an optimisation study to show how best to control it. Such
`
`control requires that full use is made of all the operating modes possible
`
`with the hybrid, but the simulation could not attempt to address the
`
`practical problems of achieving these modes nor how to switch between
`
`them. It was for this reason that phase two of the work, on the rig, was
`
`begun.
`
`(Ex. 2104 [Masding Thesis] at 235-236.)
`
`9.
`
`Phase one of the Durham Project was more fully explained under the
`
`Page 7 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`section titled “The Context of the Present Work” where the Masding Thesis provides
`
`summary of work previously accomplished and published in Bumby I, Bumby II, and
`
`Bumby III. (Ex. 2104 [Masding Thesis] at 35-49.)
`
`10.
`
`Specifically, this introductory section begins by summarizing the work
`
`accomplished and published by Bumby I:
`
`Having brought this survey of hybrid vehicle technology up to date with
`
`the discussion of the latest Volkswagen results, the relevance of the
`
`present work and the computer studies which lead up to it, can now be
`
`established. Computer studies of hybrid vehicles have been carried out at
`
`the University of Durham using a general purpose road vehicle
`
`simulation package called Janus [Bumby et al, 1985] [Bumby I]. This
`
`program, developed over a number of years in the School of
`
`Engineering and Applied Science, is capable of predicting the energy use
`
`of a variety of power train configurations.
`
`(Ex. 2104 [Masding Thesis] at 35.)
`
`11. The Masding Thesis then explains that Bumby II used the Janus
`
`simulator explained in Bumby I to investigate control strategies for hybrid vehicles
`
`that could provide maximum fuel economy.
`
`Once the simulated cycle is complete the user has at his disposal of
`
`breakdown of energy requirements on an individual component basis.
`
`Using a method similar to this Janus has been used to thoroughly
`
`investigate the economic potential of parallel i.e. engine/electric hybrids.
`
`[Bumby and Forster, 1987] [Bumby II].
`
`Page 8 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`(Ex. 2104 [Masding Thesis] at 36.)
`
`12. This introductory section also explains that Bumby III continued to
`
`investigate the overall energy savings of the control strategies being investigated.
`
`This petroleum substitution potential
`
`is also sensitive
`
`to
`
`the
`
`conventional vehicle technology used for comparison, although to a
`
`lesser extent than overall energy saving. Placing a precise figure on the
`
`percentage of petroleum which might be saved by the hybrid is
`
`complicated by the vehicle use pattern [Forster and Bumby, 1998]
`
`[Bumby III]… By assuming equal use of engine and motor at 90 km/h
`
`Bumby and Forster [Forster and Bumby, 1988] [Bumby III] calculated
`
`that a parallel hybrid could save 50% of petrol when compared with an
`
`advanced conventional vehicle featuring an efficient continuously
`
`variable transmission (CVT). Clearly such a vehicle represents a much
`
`more formidable target performance than vehicles considered in
`
`American studies.
`
`(Ex. 2104 [Masding Thesis] at 36-37.)
`
`13. This introductory section explains that the “sub-optimal control
`
`algorithm” developed in Bumby II had been selected to operate the vehicle. This
`
`control was developed to simplify the complex calculations required to fully optimize
`
`the vehicle control. Problems in coordinating controlling between all three of these
`
`components provided the motivation for Mr. Masding’s thesis.
`
`Carrying out the optimisation process in full involves quite complex
`
`calculation, particularly to determine losses in the prime movers.
`
`Consequently although it would be ideal to include it in an operational
`
`Page 9 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`vehicle there simply is too much work involved for real time
`
`computation. By considering the usage patterns for both the engine and
`
`the motor over an optimally controlled cycle however, it is possible to
`
`devise a sub-optimal control algorithm, based on a number of simple
`
`rules, which produces virtually the same economy as the fully optimal
`
`case [Bumby and Forster, 1987]. It is the practical component control
`
`problems raised by this sub-optimal control strategy which provided the
`
`motivation for the work described in this thesis.
`
`(Ex. 2104 [Masding Thesis] at 40.)
`
`14. The Masding Thesis recognized, however, that the component control
`
`problem uncovered was not unique to the sub-optimal control algorithm. Instead,
`
`such problems would occur with any vehicle that was trying to accurately control two
`
`power sources (i.e., electric motor and engine) as well as a transmission.
`
`Even if the precise form of sub-optimal strategy mentioned above were
`
`not employed, any other mode controller would still pose the same set
`
`of component control problems. As it responds to the driver's ever
`
`changing input the mode controller produces only three outputs; the
`
`torque demand for the motor, the torque demand for the engine and the
`
`required gear ratio. Component control then involves matching the
`
`torque output of engine and motor to their respective demands and
`
`controlling the transmission. If the theoretical economic potential of the
`
`vehicle is to be realised in practice these demands must be met as quickly
`
`and accurately as possible. A further complication arises in the case of
`
`the engine, in that when it is not needed it must be shut down and
`
`remain stationary. Once torque is again required from the engine a
`
`Page 10 of 51
`
`
`
`FORD 1140
`
`

`

`control subsystem is needed to start it and match its speed with the rest
`
`of the drive train so that torque is reapplied smoothly.
`
`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`(Ex. 2104 [Masding Thesis] at 42.)
`
`15. The Masding Thesis states that to analyze the control of actual vehicle
`
`components, the test bed was the next logical progression from the simulation studies
`
`performed in Bumby I-III. Again, the building and use of a test bed was recognized as
`
`phase two of the hybrid project.
`
`Microprocessor based control systems capable of meeting all of these
`
`control needs have been developed on a full-sized laboratory test bed
`
`and are described over subsequent chapters. Using a test bed represents
`
`a logical progression from the computer simulations previously carried
`
`out at Durham. Unlike an operational vehicle problems such as
`
`component packaging do not have to be tackled during the development
`
`of the rig. This leaves the way clear to test the fundamental feasability of
`
`the control strategies suggested by the simulations.
`
`(Ex. 2104 [Masding Thesis] at 42-43.)
`
`16. The Masding Thesis then confirms that the test bed designed during the
`
`second phase of the project was intended to be used with the sub-optimal control
`
`algorithm. However, the Masding Thesis confirms that this test bed could be used
`
`with other mode control strategies.
`
`Although this system is intended for use with the sub-optimal mode
`
`controller described earlier, its operation is independent of the mode
`
`Page 11 of 51
`
`
`
`FORD 1140
`
`

`

`control strategy and is a necessary part of any hybrid drive train
`
`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`controller.
`
`(Ex. 2104 [Masding Thesis] at 43.)
`
`17. Because the test rig is capable of operating mode control strategies other
`
`than the sub-optimal control algorithm, a simplified “speed based mode controller” is
`
`discussed as being used for testing and correcting the component control problem.
`
`The Masding Thesis is clear, however, that this speed based mode controller was
`
`simply for test purposes as it would not provide the efficient control needed or
`
`desired for a hybrid vehicle.
`
`Eventually the hybrid mode controller must carry out a complex
`
`efficiency oriented strategy. For test purposes however, a simpler
`
`speed based strategy was used to investigate the interaction between
`
`mode controller, sequencing
`
`logic and component control. This
`
`algorithm has the advantage that mode transitions occur predictably,
`
`which is very useful for demonstration purposes and means that data
`
`sampling can be easily arranged to cover mode transition points in detail.
`
`(Ex. 2102 [Masding Thesis] at 207-208, emphasis added.)
`
`18.
`
`Further, Bumby V likewise discloses that the speed based controller is
`
`intended to be used only for test purposes. Bumby V also clearly states that once
`
`testing is completed, the goal was to implement the more sophisticated sub-optimal
`
`control algorithm devised during Bumby II.
`
`Page 12 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`Results obtained in the previous section for the ECE15 cycle used a
`
`completely arbitrary control strategy to determine individual use of the
`
`engine and the motor and transmission shifting. This arbitrary strategy is
`
`intended purely to demonstrate that the fully integrated control system is
`
`capable of following the dictates of any more sophisticated control
`
`strategy such as those described in Bumby and Forster (1987).
`
`(Ex. 1107 at 19.)
`
`19.
`
`In providing a conclusion to the work completed during phase two of
`
`the project, the Masding Thesis reiterates that the speed based mode controller was
`
`for testing the component sequencing. The Masding Thesis again reiterates that the
`
`speed based controller did not schedule loading between the engine and motor in an
`
`efficient manner. A person having ordinary skill in the art would have understood that
`
`the Bumby references did not intend to implement the speed based controller instead
`
`of the sub-optimal control algorithm. Based on my review of the Masding Thesis and
`
`Bumby V, it is my opinion that the speed based controller was intended for
`
`demonstration purposes (i.e., testing) only to evaluate switching between all the hybrid
`
`vehicle components (e.g., motor, engine, transmission).
`
`Above component sequencing in the vehicle control hierarchy is the
`
`mode controller. This software must interpret the driver's pedal
`
`positions as a power demand and then schedule the load between the
`
`engine and motor. In chapter 7 a simple speed based mode controller
`
`was developed to carry this out. This strategy did not attempt to
`
`schedule loading in the most efficient way, but simply switched between
`
`Page 13 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`modes and gears at predetermined speeds. Despite not achieving good
`
`efficiency, the speed based strategy was useful for demonstrating that the
`
`rig drive train could switch quickly and smoothly between all of its
`
`important operating modes under driving cycle conditions.
`
`(Ex. 2104 [Masding Thesis] at 239-240.)
`
`20.
`
`It is also my opinion, which is confirmed by the Masding Thesis, that the
`
`efficiency based sub-optimal control algorithm was the intended controller to be
`
`implemented once the component control problems were corrected using the
`
`inefficient speed based controller.
`
`Once correct action of the component controllers and associated
`
`sequencing logic had been demonstrated with the speed based mode
`
`strategy, the logical extension is to introduce a mode control strategy
`
`aimed at maximizing vehicle efficiency. To do this the sub-optimal
`
`controller, devised in previous work at Durham, is most appropriate. At
`
`this point however the necessary software to implement such control has
`
`not been perfected, specifically problems have arisen in avoiding
`
`excessive numbers of gear shifts.
`
`(Ex. 2104 [Masding Thesis] at 240.)
`
`21. The discussion throughout the Masding Thesis is also consistent with
`
`the teachings of Bumby IV and V. In fact, both Bumby IV is referenced and then
`
`more fully explained in chapter 2 of the Masding Thesis.
`
`Page 14 of 51
`
`
`
`FORD 1140
`
`

`

`Bumby, J.R., Masding, P.W., 1988, 'A Test Facility for a Hybrid i.e
`
`Engine/ Battery Electric Road Vehicle Drive Train'. Trans. Inst. Meas. and
`
`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`Control, Vol. 10, No. 2, pp 87-97.
`
`(Ex. 2104 [Masding Thesis] at 245.)
`
`CHAPTER 2
`
`THE LABORATORY TEST SYSTEM
`
`The laboratory test system is the hardware realisation of the parallel
`
`hybrid drive shown in figure 1.2. The component layout in the rig
`
`system is shown by figure 2.1. Overall the rig consists of three
`
`subsystems; firstly the flywheel and dynamometer representing vehicle
`
`inertia and road load, secondly the hybrid drive train itself and finally the
`
`computers and signal monitoring equipment [Bumby and Masding,
`
`1988]. Each of these subsystems are now described in detail.
`
`(Ex. 2104 at 50.)
`
`22. Bumby V is also referenced as a paper that was yet to be published and
`
`much of the paper is discussed in chapter 5 of the Masding Thesis.
`
`CHAPTER 5
`
`CONTROLLER DESIGN
`
`An additional system is developed in chapter 7 to control the speed of
`
`the flywheel in response to the demands of a test driving cycle. Unlike
`
`the systems mentioned above however, this control loop is not strictly
`
`part of the vehicle control system since it replaces a driver. Of the four
`
`remaining systems, the control of motor speed on no load is unique, in
`
`Page 15 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`that no formal design method was used. Instead a simple method of
`
`tuning the control parameters on line was adopted because of the
`
`peculiar operating conditions of this system, within the gear changing
`
`process. In the case of the other three systems a single design technique
`
`was adopted and proved successful in all cases [Masding and Bumby,
`
`1988 (d)].
`
`(Ex. 2104 [Masding Thesis] at 136.)
`
`23.
`
` It is my opinion that nowhere throughout the Bumby IV, Bumby V or
`
`the Masding Thesis is there any teaching that the sub-optimal controller is inoperable.
`
`Instead, the Masding Thesis confirms my initial opinion that a real-world hybrid
`
`project was going through the various phases of development. Phase one was
`
`investigating and determining an implementable control strategy for a hybrid vehicle
`
`as discussed in Bumby I-III. While the optimal control provided the most efficient
`
`results, it was determined that real-world vehicles did not have the processing power
`
`for implementing this strategy. (Ex. 1104 at 7.) As such, a more simplistic sub-optimal
`
`control algorithm was developed that provided nearly the same efficiency results.
`
`24. Having decided that the sub-optimal control algorithm was the best
`
`control strategy for actual implementation, phase two of the hybrid project
`
`commenced. Phase two involved designing and building a laboratory rig for testing
`
`the sub-optimal control algorithm. It was during this phase of actual real-world
`
`implementation that component control problems were uncovered. Such problems
`
`are not uncommon during the development of project of this complexity. Also, it was
`
`Page 16 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`the recognizing of these problems that led to Philip Masding’s work. However,
`
`throughout the Masding Thesis it is continually stated that the sub-optimal control
`
`algorithm was operable and would be implemented onto the test rig and ultimately the
`
`control strategy would be incorporated into the “Lucas Chloride” hybrid vehicle.
`
`Whatever improvements are made to the rig based system, ultimate
`
`vindication of the hybrid control concepts that it uses can only be made
`
`by testing them in an operational vehicle. As a result the most important
`
`new phase of work should be to develop a compact version of the
`
`control system and incorporate it in the Lucas Chloride hybrid vehicle
`
`currently in the possession of Durham University.
`
`(Ex. 2104 [Masding Thesis] at 243.)
`
`25.
`
`It is my understanding that Mr. Hannemann has also argued that the
`
`sub-optimal algorithm is inoperable because a “fundamental flaw [exists] in the ‘sub-
`
`optimal control algorithm is an ‘excessive numbers of gear shifts.” (Ex. 2102 at 35.) I
`
`disagree. Mr. Hannemann appears to be taking snippets out of context from the full
`
`Masding Thesis in presenting his argument. For instance, the “excessive numbers of
`
`gear shifts” that he cites is found in Sections 8.1 & 8.2 (chapter 8) of the Masding
`
`Thesis. (Ex. 2104 at 240-241.) The concluding paragraph of section 8.1, however,
`
`clearly states that corrective action is being taken and that the sub-optimal control
`
`algorithm would be implemented. It is clear reading the entire paragraph in context
`
`that the Masding Thesis does not teach the sub-optimal control algorithm is
`
`Page 17 of 51
`
`
`
`FORD 1140
`
`

`

`inoperable.
`
`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`Once correct action of the component controllers and associated
`
`sequencing logic had been demonstrated with the speed based mode
`
`strategy, the logical extension is to introduce a mode control strategy
`
`aimed at maximizing vehicle efficiency. To do this the sub-optimal
`
`controller, devised in previous work at Durham, is most appropriate. At
`
`this point however the necessary software to implement such control has
`
`not been perfected, specifically problems have arisen in avoiding
`
`excessive numbers of gear shifts.
`
`(Ex. 2104 [Masding Thesis] at 240.)
`
`26.
`
`Section 8.2 also does not teach the sub-optimal control algorithm is
`
`inoperable. The introductory sentence to this section clearly states that the sub-
`
`optimal control algorithm would be used once the component control and associated
`
`sequencing logic is correct. In fact, as emphasized below, the Masding Thesis
`
`confirms that simulation studies done during phase one of the project (i.e., Bumby II
`
`and III) did not produce uncover any problems changing gears. It was not until phase
`
`two when the test rig was built that gear shifting problems were noticed. This is not
`
`uncommon when engineering a system that goes from computer simulation studies to
`
`real world physical implementation. Such problems do not teach a person having
`
`ordinary skill that something is inoperable. Instead, as the Masding Thesis states this
`
`problem occurred for both the speed based controller and sub-optimal controller. The
`
`Masding Thesis also identifies that gear shifting problems were experienced by other
`
`Page 18 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`researchers working on “optimal gear change strategies.” The Masding Thesis then
`
`continues to discuss that “steps” were being implemented to correct the problem
`
`including “hysteresis bands” and time shift delays. A person having ordinary skill in
`
`the art would therefore understand that when dealing with a complex, efficiency based
`
`hybrid vehicle (like the one being researched at Durham), problems would be
`
`presented. Engineers are taught and trained to solve problems. Simply because the
`
`engineers working on the hybrid vehicle uncovered a problem and worked to solve it
`
`does not teach the system as a whole is inoperable.
`
`8.2 Proposals for Future Work
`
`In the light of the above remarks, perhaps the most immediate challenge
`
`to be faced on the rig is to perfect the use of the sub-optimal mode
`
`controller, by tackling the gear shifting problem Even without
`
`shortcomings in the software, changing gear on grounds of efficiency
`
`leads to considerably more shifts than occur in normal driving. This is
`
`clearly illustrated by the results obtained by Bumby and Forster [Bumby
`
`and Forster, 1987] for a simulated ECE15 cycle. In these simulated
`
`results there is no problem in changing gear whenever efficiency
`
`considerations dictate, since shift times were assumed to be
`
`negligible and carried no penalty. On the rig however above average
`
`use of the gears, coupled with the rather poor shift times associated with
`
`the pneumatic system, would probably lead to unacceptable driveability
`
`on the road. Results from the speed based controller clearly showed
`
`considerable lag behind the cycle demand during a gear shift, although
`
`this is exaggerated by the low flywheel inertia. This phenomenon
`
`Page 19 of 51
`
`
`
`FORD 1140
`
`

`

`Case No.: IPR2014-00579
`Attorney Docket No.: FPGP0101IPR3
`
`
`naturally causes the computer based driver to demand high torque
`
`immediately after the gear shift in order to catch up with the cycle. Such
`
`action has further unfortunate consequences when the sub-optimal
`
`controller is used since it bases its gear changing strategy on torque
`
`demand as well as speed. Preliminary trials have already indicated that
`
`the peak in torque demand after one gear shift may immediately cause
`
`another. Similar problems to those outlined above have been
`
`encountered by other workers researching optimal gear change
`
`strategies. In work at Ford [Kuzak et al, 1987] transmission output
`
`speed and throttle position were used as indicators of engine efficiency
`
`and thus a shift strategy was based on them. Several steps were needed
`
`to reduce numbers of shifts from those produced with fully optimal
`
`control,
`
`the simplest defined hysteresis bands on
`
`the
`
`throttle
`
`angle/speed plane between the zones defined for each gear. Secondly a
`
`minimum time between shifts was imposed to improve driveability.
`
`Such a limitation could easily be applied to the sub-optimal mode
`
`controller, and it is encouraging to note that the Ford [Kuzak et al,
`
`1987] result showed that by imposing a 5 second limit between shifts,
`
`fuel economy in city driving suffered by only 3% but shift frequency was
`
`reduced by 50%.
`
`(Ex. 2104 [Masding Thesis] at 240-241, Emphasis added.)
`
`27.
`
`Further, it is my understanding that Mr. Hannemann relied on the result
`
`for a “three-cylinder engine” for opining a person having ordinary skill in the art
`
`wouldn’t use the sub-optimal control algorithm (Ex. 2102 at ¶¶69-76.) I disagree. For
`
`instance, when the “base configuration” conventional vehicle is compared to the
`
`Page 20 of 51
`
`
`
`FORD 1140
`
`

`

`Case N