`Paper No.
`Filed: June 10, 2019
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`CATERPILLAR INC.
`Petitioner
`
`v.
`
`WIRTGEN AMERICA, INC.,
`Patent Owner
`
`
`Case No. IPR2018-01091
`U.S. Patent No. 8,308,395
`
`
`
`
`
`
`
`
`
`
`PETITIONER’S REPLY
`
`
`
`

`

`I.
`II.
`
`TABLE OF CONTENTS
`
`Introduction ...................................................................................................... 1
`Brabec is Analogous Art to the ’395 Patent .................................................... 1
`A.
`Prior art cited in the Petition establishes that Brabec is
`analogous. .............................................................................................. 1
`Prior art cited during the ’395 patent prosecution establishes
`that Brabec is analogous. ...................................................................... 2
`Brabec teaches that its sensor switching technology is
`applicable to a variety of construction machines. ................................. 3
`D. Wirtgen conjures up a self-serving description of “application-
`related” reasons for sensor switching as the basis to exclude
`Brabec as prior art. ................................................................................ 3
`III. All Evidence Points to Davis Controlling Its Milling Drum Using Two
`of Its Three Sensors ......................................................................................... 4
`A.
`The evidence is uncontroverted that Davis needs only two of its
`three sensors to control its milling drum. .............................................. 4
`Before being faced with this IPR, Wirtgen admitted to the PTO
`and EPO that Davis’s controller uses only two sensors for
`control. ................................................................................................... 5
`C. Davis does not disclose using all three sensors for controlling
`its milling drum. .................................................................................... 6
`IV. Wirtgen Fails to Rebut the Ample Motivation Provided by Petitioner
`to Combine the Teachings of Davis and Brabec ............................................. 7
`The Davis-Brabec Combination Teaches Switching Control Based
`Upon a Second Selected Subset During Milling Operation .......................... 11
`A. Wirtgen’s construction of “selected subset” is inconsistent with
`the intrinsic evidence. .......................................................................... 11
`The correct construction of “selected subset” is “a subset that is
`chosen.” ............................................................................................... 13
`
`B.
`
`C.
`
`V.
`
`B.
`
`B.
`
`i
`
`

`

`C.
`
`The Davis-Brabec combination teaches the “selected subset”
`feature under either construction. ........................................................ 14
`VI. Brabec’s Description of a Manual Switchover System Constitutes
`Prior Art ......................................................................................................... 15
`VII. The Davis-Brabec Combination Discloses Every Element of
`Independent Claims 1 and 11 ........................................................................ 16
`VIII. The Davis-Brabec Combination Discloses Every Element of Claims 5
`and 15 ............................................................................................................. 19
`A. Davis discloses pre-selecting and pre-setting ..................................... 19
`B.
`Petitioner relies upon the manual switchover system described
`in Brabec for claims 5 and 15, not a combination of the manual
`and automatic system .......................................................................... 20
`IX. The Board Should Invalidate Claims 3, 6-8, 10, 13, 16, 17, 19, 20, 22,
`24, 26, and 27 ................................................................................................. 22
`The Board Was Misled by Wirtgen Concerning the Complete
`Teachings of Krieg, Which When Combined With Davis Invalidate
`the Claims Challenged Under Ground 2 ....................................................... 22
`XI. Wirtgen Has Now Admitted Dr. Bevly is Qualified to Render His
`Opinions ......................................................................................................... 25
`XII. Conclusion ..................................................................................................... 26
`
`
`X.
`
`
`
`
`
`ii
`
`

`

`I.
`
`Introduction
`The ’395 patent invented nothing new. Prior art already disclosed milling
`
`machines and control systems for controlling milling drum orientation. Switching
`
`sensors used to control work implement orientation without interrupting machine
`
`operations was also well known. The challenged claims simply applied known
`
`sensor-switching technology to a known milling machine.
`
`Petitioner will debunk each assertion made by Wirtgen in its failed attempt
`
`to overcome Ground 1.
`
`Petitioner will further show that Wirtgen’s attempt to overcome Ground 2 is
`
`based entirely on its misrepresentation to the Board that Krieg only discloses a
`
`motor grader switching sensors between passes. This Board must reject Wirtgen’s
`
`inaccurate representations of Krieg.
`
`II. Brabec is Analogous Art to the ’395 Patent
`Prior art cited in the Petition establishes that Brabec is analogous.
`A.
`The Petition cited prior art disclosing the well-known practice of
`
`implementing similar tool control systems on different types of road construction
`
`machines. Petition, 28; Ex-1002, ¶105 (citing Exs. 1012, 1013). For example,
`
`Ex-1012 states:
`
`The G 176 plus controller is used for the automatic
`leveling of a tool…. Most of the time the unit is used for
`the leveling on asphalt pavers …. The G 176 plus is also
`
`
`
`1
`
`

`

`designed for milling machines, concrete pavers and
`other construction equipment.
`
`Ex-1012, 11 (emphasis added). Similarly, Ex-1013 discloses a tool control system
`
`for a milling machine, stating “the use of the laser-regulating means in accordance
`
`with the present invention is possible with each construction machine which
`
`changes the profile of a surface, i.e. which removes or applies layers, to obtain a
`
`desired profile and, in the process, references to a reference surface.” Ex-1013, 9:9-
`
`15. Moreover, grade and slope control systems having similar features and
`
`capabilities were routinely implemented on cold planers, motor graders, pavers,
`
`dirt trimmers, and reclaimers. Ex-1046, ¶¶35-36. The prevalence of grade and
`
`slope control systems similar to Brabec on different road construction machines
`
`demonstrates that Brabec is analogous prior art.
`
`B.
`
`Prior art cited during the ’395 patent prosecution establishes that
`Brabec is analogous.
`Wirtgen did not raise its non-analogous art argument during prosecution
`
`when the Examiner applied the concrete screed of Hohmann to reject the claims.
`
`Ex-1008, 62-65; Petition, 7; Ex-1002, ¶87. Wirtgen did not challenge Hohmann as
`
`non-analogous art, arguing instead that the Hohmann system did not disclose
`
`sensor switchover “without disrupting the milled surface.” Ex-1008, 64. The
`
`concrete screed of Hohmann, which Wirtgen accepted as analogous art, is the same
`
`type equipment as the exemplary concrete screed of Brabec. The screed
`
`2
`
`

`

`technology was properly considered analogous art during prosecution, just as it
`
`should be considered in this IPR.
`
`Further, Wirtgen itself cited prior art unrelated to milling machines as being
`
`material prior art in an Information Disclosure Statement. Ex-1008, 119.
`
`Specifically, Wirtgen cited Johnson, which discloses a cutting tool for cutting
`
`depressions along shoulders or center lines of roads. Ex-1044, ¶¶2, 3. Johnson does
`
`not disclose a road milling machine. Thus, Wirtgen itself recognized that the ’395
`
`patent’s field of art encompasses more than road milling machines.
`
`C. Brabec teaches that its sensor switching technology is applicable
`to a variety of construction machines.
`Wirtgen misleadingly portrays Brabec as limited to concrete screeds.
`
`POR, 32. But, Brabec describes a concrete screed merely as “one possible
`
`platform.” Ex-1005, ¶15. Brabec teaches that its disclosure applies to a variety of
`
`construction machines having grading implements for earthmoving (id., ¶¶1-4).
`
`D. Wirtgen conjures up a self-serving description of “application-
`related” reasons for sensor switching as the basis to exclude
`Brabec as prior art.
`The ’395 patent does not describe what constitutes application-related
`
`reasons for switching sensors. Dr. Wilhelm conceded this (Ex-1048, 60:16-61:3)
`
`and further that he relied upon extrinsic information provided by Wirtgen in
`
`deciding what constituted “application related” reasons (id., 58:7-25). Curiously,
`
`Dr. Wilhelm did not talk with the ’395 inventors so he cannot possibly speak to
`
`3
`
`

`

`what the inventors thought or opine that Brabec does not address a problem facing
`
`the ’395 patent inventors. Id., 29:16-30:1.
`
`Contrary to Wirtgen’s position, sensors frequently fail on milling machines
`
`(Ex-1046, ¶¶34, 37-43) and, therefore, sensor failure is an application-related
`
`reason. Ex-2006, 99:20-100:2. Further, Caterpillar has provided evidence in the
`
`form of the Netka declaration that laser sensors on milling machines face the same
`
`problem (i.e. blockage of the laser beam) faced by Brabec. Ex-1046, ¶43.
`
`Therefore, Brabec is pertinent to a problem addressed by the ’395 patent.
`
`III. All Evidence Points to Davis Controlling Its Milling Drum Using Two of
`Its Three Sensors
`A. The evidence is uncontroverted that Davis needs only two of its
`three sensors to control its milling drum.
`Caterpillar presented unrefuted evidence that the milling drum of Davis
`
`could be controlled using only two sensors. The cited evidence includes Swisher,
`
`the admissions in the ’395 patent (Ex-1001, 1:31-33), and supporting testimony
`
`from Dr. Bevly. Petition, 26-27; Ex.1002, ¶¶107-113. Dr. Bevly further explained
`
`in his deposition:
`
`It has to deal with terminology known as degrees of
`freedom. So I have two degrees that I’m trying to control;
`that’s a depth and an angle. I have two actuators to achieve
`that, which suggests that I need two measurements to
`monitor the two degrees of freedom.
`
`4
`
`

`

`Ex-2006, 176:10-15; see also id., 183:6-18 (noting that Swisher, Brabec, and Krieg
`
`all disclose controlling the work implement depth and slope using only two
`
`sensors.)
`
`Dr. Wilhelm agreed with Dr. Bevly’s geometric explanation of why only
`
`two sensors were needed and agreed further that the milling drum of Davis could
`
`be controlled with 2 sensors. Ex-1048, 67:7-17, 68:8-12.
`
`B.
`
`Before being faced with this IPR, Wirtgen admitted to the PTO
`and EPO that Davis’s controller uses only two sensors for control.
`Wirtgen characterized Davis as using only two of its three sensors for
`
`milling drum control twice, once to the USPTO and once to the EPO. Petition, 26-
`
`27. Wirtgen told the USPTO “the only thing that the EP reference [i.e. Davis]
`
`teaches is that before starting the milling operation, the operator can chose to
`
`control right and left side milling depth and slope by selecting to control based on
`
`sensors for any two of those three parameters.” Ex-1008, ¶104 (emphasis
`
`added). Likewise, Wirtgen told the EPO: “An exchange of the sensors during the
`
`milling operation is not provided according to [EP-Davis]. Two sensors have to
`
`be selected prior to the milling operation.” Ex-1011, 20 (emphasis added).
`
`Wirtgen’s and Dr. Wilhelm’s attempts to walk away from these admissions
`
`based on some alleged “conflation” excuse ignores Wirtgen’s unambiguous
`
`statements, made not once but twice. Curiously, Dr. Wilhelm was not even aware
`
`of Wirtgen’s representations to the EPO. Ex-1048, 81:19-25, 82:2-3.
`
`5
`
`

`

`C. Davis does not disclose using all three sensors for controlling its
`milling drum.
`As Dr. Bevly correctly recognized (Ex-2006, 153:20-154:1), Davis suggests
`
`that its controller may not receive all 3 signals. Davis discloses that “a difference
`
`signal δS, δD and δα is sent out by the first data processing system 16, or 17, or
`
`19, respectively.” Ex-1004, ¶42 (emphasis added). Davis explains: “As soon as the
`
`scarification work is started, … the first data processing systems send out a
`
`difference signal δS, δD or δα” and that these signals are received by the PLC.
`
`Id., ¶49 (emphasis added). By specifically using the disjunctive “or,” Davis
`
`suggests that its PLC may not always receive all three signals. In fact, Davis
`
`discloses that “if the readings substantially correspond to the preset values, no
`
`difference signal will be sent out by the first data processing systems.” Id., ¶50.
`
`Wirtgen’s argument relies on the false premise that Davis must use all three
`
`sensors for controlling milling depth and slope because its controller allegedly
`
`receives all three signals. POR, 41, 43. However, it was well-known in the art that
`
`a controller could receive many signals but use only two to control the depth and
`
`slope of a work tool. For example, Brabec discloses a controller that receives
`
`signals from three sensors, uses two to control the work implement depth and
`
`slope, and uses the third only when one of the other two sensor fails. Ex-1005, ¶35.
`
`This Brabec teaching alone refutes Wirtgen’s argument.
`
`6
`
`

`

`Moreover, even assuming arguendo, that Davis’s PLC receives all three
`
`difference signals, there is still no evidence that the PLC determines control signals
`
`21, 22 based on all three difference signals. Dr. Wilhelm agrees, stating “Davis
`
`does not specifically disclose how the control signals to the hydraulic actuators are
`
`generated.” Ex-2004, ¶101. But then he dismisses this, saying “the control details
`
`are not critical to understand the solution offered by Davis.” Id. Dr. Wilhelm’s
`
`reasoning stretches credulity because those control details would in fact disclose
`
`whether Davis uses all three sensor signals.
`
`Although Davis does not explicitly disclose how it generates its two control
`
`signals, a POSITA would understand that Davis’s controller need use only two of
`
`the three difference signals to generate the control signals. In re Preda, 401 F.2d
`
`825, 826 (CCPA 1968) (stating “it is proper to take into account not only specific
`
`teachings of the reference but also the inferences which one skilled in the art would
`
`reasonably be expected to draw therefrom.”)
`
`IV. Wirtgen Fails to Rebut the Ample Motivation Provided by Petitioner to
`Combine the Teachings of Davis and Brabec
`First Wirtgen argues that Davis and Brabec cannot be combined because
`
`Davis does not include an unused sensor. POR, 45. But, as explained above, a
`
`POSITA would have understood Davis as needing only two of its three sensors,
`
`leaving one unused sensor.
`
`7
`
`

`

`Next, Wirtgen argues that modifying Davis with Brabec “would
`
`impermissibly change its principle of operation,” Id.; Ex-2004, ¶133. Neither
`
`Wirtgen nor its expert explain why this would occur or, frankly, what this even
`
`means.
`
`Wirtgen also argues that Brabec’s reasons for switching sensors are
`
`inapplicable to Davis because loss of sensor signals was not a problem for Davis’s
`
`scarifier. POR, 46-47 (citing Ex-2008, ¶30). Wirtgen’s only evidence comes from
`
`its employee, Jan Schmidt. POR, 47. But, Mr. Schmidt provides no underlying data
`
`to support his questionable conclusion that sensor failure is not an issue. Indeed,
`
`Mr. Schmidt offers internally inconsistent testimony on sensor failure. After
`
`asserting that sensor failure was not a concern in road milling, Mr. Schmidt then
`
`proceeds to describe sensor failures that do occur and how these failures are dealt
`
`with. Ex-2008, ¶30. Moreover, Mr. Schmidt’s conclusion defies common sense
`
`since milling machines operate in harsh environments that would naturally lead to
`
`sensor failure issues.
`
`Caterpillar has provided Mr. Netka’s declaration testifying that sensors
`
`routinely fail on milling machines due to wear and tear or damage caused by debris
`
`or obstructions on the job site. Ex-1046, ¶¶34, 37-43. Prior art also confirms Mr.
`
`Netka’s declaration testimony that grade and slope sensors were prone to
`
`malfunction or failure. For example, Carew discloses that malfunctioning grade
`
`8
`
`

`

`and/or slope sensors on cold planers could affect the grade and slope of the
`
`machined road surface. Ex-1041, 1:30-33; 3:7-8.
`
`Wirtgen next argues that “a POSA would not have considered [loss of sensor
`
`signals] as one of the ‘application-related reasons,’” and falsely asserts that Dr.
`
`Bevly conceded this point. POR, 47. Dr. Bevly, however, testified exactly the
`
`opposite. Ex-2006, 99:22-100:4 (“Q: What are those application-related reasons?
`
`A. It could be a variety of things. It could be a sensor failure, a sensor outage,
`
`which would be kind of a temporary failure.”)
`
`Citing only its own employee’s testimony, Wirtgen next argues that even if a
`
`sensor did fail on a milling machine “it would be undesirable to automatically
`
`switch the sensors … because it would risk adversely affecting the work product.”
`
`POR, 48; see also id., 35-36. The ’395 patent itself refutes Wirtgen’s newly-minted
`
`focus on manual switchover and its new-found argument that automatic switchover
`
`is inapplicable. Of 27 claims in the ’395 patent, only two (claims 5 and 15) recite
`
`“a manually operable switchover device.” It is strange that 25 of the 27 claims
`
`cover an automatic switchover system that Wirtgen now says would not work.
`
`Moreover, Wirtgen provides no evidence as to why automatic switchover would be
`
`undesirable. Nowhere does Mr. Schmidt provide any evidence that militates
`
`against automatic switchover of sensors. See Ex-2006, ¶30. Dr. Wilhelm merely
`
`9
`
`

`

`repeats Wirtgen’s attorney argument and provides no independent reasons or
`
`evidence against automatic switchover. Ex-2004, ¶136.
`
`Wirtgen also argues that an operator would want to stop the milling
`
`operation because” allegedly “there are too many conditions (e.g., curves, curbs,
`
`ramps, soft shoulders, etc.) where specific sensors are relied upon for proper road
`
`milling.” POR, 48. Again, Wirtgen provides only its employee Mr. Schmidt’s
`
`contradictory testimony devoid of any underlying facts. Id.
`
`Wirtgen’s position is refuted by what actually occurs during road
`
`construction projects. As Mr. Netka, who has worked extensively with milling
`
`contractors and has over 5000 hours of milling machine operating experience,
`
`explains: “Cold planers are expensive machines and as a result milling contractors
`
`want to maximize the utilization of these machines. Therefore, a milling contractor
`
`assigned to complete milling of a particular stretch of road is highly motivated to
`
`finish the job so that the machine may be redeployed to another job at the same or
`
`different location.” Ex-1046, ¶45. Mr. Netka further explains that in the event of a
`
`sensor failure, milling contractors typically will choose to continue the milling
`
`operations using an alternate sensor to maximize usage of their machines.
`
`Id., 44-52.
`
`Wirtgen also argues that the loss of sensor signal in Brabec occurs due to
`
`blockage of a laser beam inside a building and that “this concern is not present in
`
`10
`
`

`

`Davis’s operating environment.” POR, 49. This is false. As Mr. Netka explained,
`
`laser sensors of the type used by Brabec can and do fail due to laser beam blockage
`
`even on milling machines that operate outdoors. Ex-1046, ¶43.
`
`Finally, citing a single response in Dr. Bevly’s deposition, Wirtgen argues
`
`that he relied on hindsight. A single deposition response, however, does not
`
`demonstrate hindsight. See Icon-IP Pty Ltd. v. Specialized Bicycle Components,
`
`Inc., 87 F. Supp.3d 928, 943 (N.D. Cal. Mar. 31, 2005). Hindsight occurs when a
`
`POSITA finds components of an invention in various prior art references and on
`
`that basis alone declares the invention obvious. Ruiz v. A.B. Chance Co., 357, F3d
`
`1270, 1275 (Fed. Cir. 2004). That is not what Dr. Bevly did here. Rather, he
`
`identified the teachings of Davis and Brabec, and explained why a POSITA would
`
`have combined those teachings. Ex-1002, ¶¶39-58, 97-261.
`
`V. The Davis-Brabec Combination Teaches Switching Control Based Upon
`a Second Selected Subset During Milling Operation
`Petitioner will now explain (1) why Wirtgen’s construction of “selected
`
`subset” is incorrect, (2) what the correct construction is, and (3) why the prior art
`
`teaches this “selected subset” feature under either construction.
`
`A. Wirtgen’s construction of “selected subset” is inconsistent with
`the intrinsic evidence.
`Wirtgen asserts that “selected subset” means “a subset of the plurality of
`
`selectable sensors purposefully chosen for preference or special quality.” POR, 16.
`
`11
`
`

`

`Neither the phrase “selected subset” nor any description of “purposefully choosing
`
`sensors for preference or special quality” appears anywhere in the specification or
`
`prosecution history. Ex-1047, ¶¶52-66. Even the single sentence cited by Wirtgen
`
`as support for its construction does not mention “selected subset,” instead reciting
`
`“switching over from the current sensor to the pre-selected sensor.” Ex-1001,
`
`2:27-30.
`
`The term “pre-selected” does not support Wirtgen’s construction. The prefix
`
`“pre” means “before.” Ex-1043, 3. Thus, a pre-selected sensor means a sensor
`
`chosen before some event. It does not mean a sensor purposefully chosen for
`
`preference or special quality.
`
`Wirtgen’s expert could not point to any disclosure in the ’395 specification
`
`to support its construction. Ex-1048, 65:15-22. Indeed, Wirtgen’s expert admitted
`
`that he was relying upon a single dictionary definition that was given to him by
`
`counsel. Id., 64:6-11, 66:9-12. A cherry-picked dictionary definition that has no
`
`support in the intrinsic evidence cannot be the legally correct claim construction.
`
`Innogenetics, N.V. v. Abbott Labs., 512 F.3d 1363, 1371 (Fed. Cir. 2008) (holding
`
`that reliance on a single dictionary was legal error where the dictionary definition
`
`cited by the accused infringer was one of several possible meanings.)
`
`12
`
`

`

`B.
`
`The correct construction of “selected subset” is “a subset that is
`chosen.”
`The Board should interpret “selected subset” to mean “a subset that is
`
`chosen.” The ’395 patent supports this construction. For example, claims 1 and 11
`
`recite “switch over … from control based on a first selected subset … to control
`
`based on upon a second selected subset.” (emphasis added). Thus, the ’395
`
`patent’s claim language indicates that selected subset means a subset that is chosen
`
`for control.
`
`The ’395 specification discloses “a selectable sensor (B) that is to be
`
`exchanged for the sensor (A, C) currently in use.” Ex-1001, 4:30-31. The ’395
`
`patent does not disclose that the first selected subset (A, C) is selected for
`
`preference or special quality, only that the subset is “currently in use.” Thus, the
`
`first subset (A, C) is chosen for controlling depth and slope of a work implement.
`
`Likewise, subset (A, B) or (B, C) constitute the second selected subset that is
`
`chosen for control simply by exchanging one sensor A or C with sensor B. Thus,
`
`the ’395 specification supports Petitioner’s construction of “selected subset” as a
`
`“subset that is chosen” for control.
`
`Consistent with the specification, a standard dictionary defines “select” as
`
`“to choose or pick out.” Ex-1043, 4. It is appropriate to refer to a dictionary to
`
`confirm a construction consistent with the specification. Praxair, Inc. v. ATMI,
`
`Inc., 543 F.3d 1306, 1325, (Fed. Cir. 2008) (holding that dictionary definitions
`
`13
`
`

`

`may appropriately be relied upon “so long as the dictionary definition does not
`
`contradict any definition found in or ascertained by a reading of the patent
`
`documents.”) Petitioner’s proposed construction is legally correct under a district
`
`court standard. Here, where the standard is broader under BRI, there is no doubt
`
`that Petitioner’s proposed construction is correct.
`
`C. The Davis-Brabec combination teaches the “selected subset”
`feature under either construction.
`Using Petitioner’s correct construction, Brabec discloses sensors 20 and 22
`
`as the first selected subset, and sensors 20 and 42 as the second selected subset.
`
`Each “selected subset” is chosen to control the work implement.
`
`Using Wirtgen’s incorrect construction, sensors 20 and 22 are selected as the
`
`first selected subset. This first subset may be laser sensors or may be selected from
`
`a group of other sensors including sonic distance or contacting sensors. Ex-1005,
`
`¶¶25, 41. When switching to the second selected subset, Brabec teaches there may
`
`be multiple alternative sensors 42 from which to choose. Id., 41. After initially
`
`disputing these teachings of Brabec, Dr. Wilhelm conceded that Brabec does teach
`
`selecting its sensor subsets from a larger group of available sensors. Ex-1048,
`
`83:6-84:8, 85:10-17.
`
`14
`
`

`

`VI. Brabec’s Description of a Manual Switchover System Constitutes Prior
`Art
`Wirtgen argues that Brabec does not disclose manual sensor switching
`
`because its background section merely describes the general state of the art.
`
`POR, 56. This argument ignores longstanding precedent that “[a] reference must be
`
`considered for everything it teaches by way of technology and is not limited to the
`
`particular invention it is describing and attempting to protect.” Belden Inc. v. Berk-
`
`Tek LLC, 805 F.3d 1064, 1076 (Fed. Cir. 2015); In re Heck, 699 F.2d 1331, 1333
`
`(Fed. Cir. 1983) (“The use of patents as references is not limited to what the
`
`patentees describe as their own inventions or to the problems with which they are
`
`concerned. They are part of the literature of the art, relevant for all they contain.”)
`
`In re Transaction Holdings, 484 Fed.Appx. 469, 471-472 (Fed. Cir. 2012)
`
`(nonprecedential) (affirming Board’s conclusion of obviousness based on
`
`disclosure in the background section of one reference and the disclosed
`
`embodiments of a second reference); In re Yufa, 452 Fed.Appx. 998, 1001 (Fed.
`
`Cir. 2012) (nonprecedential) (affirming Board’s conclusion of obviousness based
`
`on a combination of the disclosure in the background and the disclosed
`
`embodiments of a single prior art reference).
`
`As Dr. Bevly explained, a POSITA would understand Brabec’s background
`
`as disclosing a known manual switchover system. Petition, 48; Ex-1002, ¶¶172174.
`
`Wirtgen agrees, characterizing the descriptions of manual switchover in Brabec as
`
`15
`
`

`

`“previously-known.” POR, 57. Caterpillar’s Petition also explained that a POSITA
`
`would understand from Brabec’s description that an operator of the manual
`
`switchover system would engage buttons, switches, or levers to perform the
`
`disclosed functions. Petition, 47-48. Thus, the Petition explained how Brabec
`
`discloses manual sensor switching and a manual switchover device. The manual
`
`system described by Brabec also teaches the features of switching sensors during
`
`machine operation without altering the work implement performance. Ex-1002,
`
`¶¶57, 238; Ex-1047, ¶¶75-76, 100-101. This disclosure in Brabec of a manual
`
`switchover system constitutes prior art regardless of whether it is the invention
`
`claimed. KSR Intl. Co. v. Teleflex Inc., 550 U.S. 398, 420-21 (2007) (holding that a
`
`reference may be read for all that it teaches, including uses beyond its primary
`
`purpose).
`
`The Board should reject Dr. Wilhelm’s opinions on Brabec because he
`
`failed to address ¶¶8-9 of Brabec that describe the manual system. Dr. Wilhelm
`
`disregarded this portion of Brabec because he didn’t believe it was Brabec’s
`
`invention. Ex-1048, 69:4-7, 76:6-14. Dr. Wilhelm’s approach constitutes legal
`
`error.
`
`VII. The Davis-Brabec Combination Discloses Every Element of Independent
`Claims 1 and 11
`Wirtgen argues that Caterpillar’s Petition failed to show that Davis and
`
`Brabec disclose the first and second selected subsets. Wirtgen also argues that the
`
`16
`
`

`

`Davis-Brabec combination does not disclose selection of the second subset during
`
`milling operations. The Board should reject these arguments for the following
`
`reasons.
`
`Contrary to Wirtgen’s assertions, Caterpillar’s Petition discussed both the
`
`first and second selected subsets. The phrase “selected subset” of sensors means a
`
`subset of sensors that is chosen for control. Caterpillar explained that in normal
`
`operation, Brabec’s controller uses left and right sensors (20, 22) to control work
`
`implement orientation. Petition, 41. Dr. Bevly testified that a POSITA would have
`
`understood “sensors 20 and 22 correspond to a first subset of sensors” chosen for
`
`controlling work implement orientation. Id., citing Ex-1002, ¶151. Likewise,
`
`Caterpillar explained that when sensor 22 is unavailable, Brabec’s controller
`
`exchanges alternate sensor 42 with sensor 22 to control implement orientation.
`
`Petition, 42. Caterpillar cited Dr. Bevly’s testimony that a POSITA would have
`
`understood “sensors 20 and 42 correspond to a second subset of sensors” chosen
`
`for control. Id., citing Ex-1002, ¶152.
`
`Wirtgen next argues that Brabec does not disclose a “first selected subset”
`
`because the primary sensors are “always used as the starting sensors.” POR, 52.
`
`This is not true. Brabec teaches that a variety of sensors can be selected as the
`
`starting sensors 20, 22. One embodiment selects laser sensors, but other
`
`17
`
`

`

`alternatives are described in Brabec. Ex-1005 at ¶¶25, 41. Dr. Wilhelm agrees.
`
`Ex-1048, 83:24-84:8
`
`Moreover, Wirtgen does not identify where the claims prohibit selection of
`
`the same two sensors every time. As Dr. Bevly explained in his deposition, the first
`
`subset of sensors could be selected a priori, for example, by “designers and
`
`builders of the machine.” Ex-2006, 134:2-3, 135:5-11.
`
`Wirtgen also argues that Brabec does not disclose the second selected subset
`
`because “Brabec’s system was forced, by the chance unavailability of the other
`
`primary sensor, into using the remaining functional sensor.” POR, 53. As discussed
`
`earlier, this is not true. Brabec teaches multiple alternative sensors 42 from which a
`
`selection can be made. Ex-1005, ¶41. Again Dr. Wilhelm agrees. Ex-1048,
`
`85:10-17.
`
`Wirtgen further incorrectly argues that the combination of Davis and Brabec
`
`does not disclose selection of a second selected subset during milling operation.
`
`POR, 54. But claims 1 and 11 merely require switching control from a first
`
`selected subset to a second selected subset during milling operations.
`
`Caterpillar’s Petition explained how the controller of Davis’s scarifier modified
`
`with Brabec’s teachings would switch control to the second subset of sensors
`
`during milling operations. Petition, 43-45.
`
`18
`
`

`

`Wirtgen also incorrectly argues that Brabec does not disclose selecting the
`
`second selected subset during milling operations because Brabec’s controller
`
`selects a control mode based on a predetermined priority. POR, 54-55. Brabec’s
`
`operator sets “a selection priority among the alternative sensors” and “controller 28
`
`selects the appropriate mode based upon a predetermined priority set by the
`
`operator.” Ex-1005, ¶39-40. Although Brabec discloses automatically switching to
`
`the alternate sensor, nothing in Brabec prohibits the operator from setting or
`
`altering selection priorities during machine operations but prior to the switchover.
`
`Ex-2006, 206:5-14. So long as the operator sets the priority before the switchover,
`
`Brabec’s controller would still select a sensor based on a predetermined priority.
`
`Thus the term “predetermined” does not limit Brabec to setting the priority before
`
`commencing machine operations.
`
`VIII. The Davis-Brabec Combination Discloses Every Element of Claims 5
`and 15
`A. Davis discloses pre-selecting and pre-setting
`Wirtgen asserts that Davis does not disclose pre-selecting or pre-setting.
`
`POR, 55-56. This is incorrect. Claims 5 and 15 recite pre-selection and pre-setting
`
`“prior to effecting the switchover.” This claim language does not prohibit pre-
`
`selection and pre-setting before beginning milling operations. As Caterpillar
`
`demonstrated, Davis discloses manually pre-selecting two sensors and also that
`
`pre-setting target values in two of Davis’s sensors acts to pre-set the target value in
`
`19
`
`

`

`its third sensor. Petition, 47. Thus, Caterpillar demonstrated that