`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Paper No.
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`ZTE (USA), Inc.
`
`Petitioner
`
`v.
`
`Cywee Group Ltd.
`
`(record) Patent Owner
`
`IPR2019-00525
`Patent No. 8,552,978
`
`PETITION FOR INTER PARTES REVIEW
`UNDER 35 U.S.C. §§311-319 AND 37 C.F.R. §42.100 ET. SEQ
`
`1
`
`
`
`
`TABLE OF EXHIBITS ............................................................................................. 4
`
`TABLE OF CONTENTS
`
`NOTICE OF LEAD AND BACKUP COUNSEL ..................................................... 5
`
`NOTICE OF THE REAL-PARTIES-IN-INTEREST ............................................... 5
`
`NOTICE OF RELATED MATTERS ........................................................................ 5
`
`NOTICE OF SERVICE INFORMATION ................................................................ 6
`
`GROUNDS FOR STANDING .................................................................................. 6
`
`STATEMENT OF PRECISE RELIEF REQUESTED .............................................. 7
`
`THRESHOLD REQUIREMENT FOR INTER PARTES REVIEW ........................ 7
`
`I.
`
`INTRODUCTION ........................................................................................... 7
`
`A.
`
`Prosecution History and Issued Claims .............................................. 11
`
`II.
`
`CLAIM CONSTRUCTION .......................................................................... 14
`
`A.
`
`B.
`
`Claim 10—“spatial reference frame” and similar terms ..................... 14
`
`Claim 10—“rotation output” ............................................................... 18
`
`GROUNDS .............................................................................................................. 18
`
`Ground 1. Claims 10 and 12 are obvious over Zhang in view of
`Bachmann. .......................................................................................... 18
`
` Overview of the Combination ............................................................. 19
`
` Rationale for the Combination ............................................................ 30
`
` Ability to Implement and Reasonable Expectation of Success ............ 37
`
` Graham Factors ..................................................................................... 39
`
` Claim Mapping ...................................................................................... 39
`2
`
`
`
`
`Ground 2. Claims 10 and 12 are unpatentable over Liberty in view
`of Bachmann. ...................................................................................... 60
`
` Overview of the Combination ............................................................. 61
`
` Rationale for the Combination ............................................................. 68
`
` Ability to Implement and Reasonable Expectation of Success ............ 74
`
`Difference Between the Combination and Prior Discussion of
`Liberty .................................................................................................. 76
`
` Graham Factors .................................................................................... 76
`
` Claim Mapping ..................................................................................... 77
`
`CERTIFICATE OF SERVICE ................................................................................ 84
`
`CERTIFICATE OF WORD COUNT ...................................................................... 85
`
`
`
`
`
`
`
`
`3
`
`
`
`TABLE OF EXHIBITS
`
`Exhibit No.
`1001
`1002
`1003
`1004
`1005
`1006
`1007
`
`1008
`
`1009
`1010
`
`1011
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Description
`U.S. Pat. No. 8,552,978 (“the ’978 patent”).
`Declaration of Professor Majid Sarrafzadeh.
`C.V. of Professor Majid Sarrafzadeh.
`U.S. Pat. No. 7,089,148 (“Bachmann”).
`U.S. Pat. App. Pub. 2004/0095317 (“Zhang”).
`U.S. Pat. 7,158,118 (“Liberty”).
`Return of Service for Cywee Group Ltd. v. Google, Inc., Case No.
`1-18-cv-00571, (D. Del.).
`Return of Service for Cywee Group Ltd. v. Huawei Technologies
`Co., Inc. et al., Case No. 2-17-cv-00495, (E.D. Tex.).
`File History of U.S. Pat. App. 13/176,771.
`
`Joint Claim Construction and Prehearing Statement in Cywee
`Group Ltd. v. Samsung Electronics Co. Ltd. et al., Case No. 2-17-
`cv-00140, (E.D. Tex.).
`Exhibit E (Claim chart with of U.S. Pat. No. 8,552,978) to
`CyWee’s Complaint in Cywee Group Ltd. v. Google, Inc., Case
`No. 1-18-cv-00571, (D. Del.)
`
`1012-1049 Reserved
`
`1050
`
`1051
`
`Comparison between the Current Petition and Petition in IPR2018-
`01257
`
`Claim Construction Order, Cywee Group Ltd. v. Motorola Mobility
`LLC, 17-cv-00780, Jan. 4, 2019
`
`1052
`
`Declaration of Andrea Shoffstall
`
`4
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Petitioner ZTE (USA), Inc. (“Petitioner” or “ZTE”) respectfully requests
`
`inter partes review under 35 U.S.C. §311 of claims 10 and 12 of U.S. Pat. No.
`
`8,552,978 (“the ’978 patent”).
`
`NOTICE OF LEAD AND BACKUP COUNSEL
`
`Lead Counsel: James R. Sobieraj (Reg. No. 30,805); Tel: 312.321.4200
`
`Backup Counsel: Jon H. Beaupré (Reg. No. 54,729); Tel: 734.302.6000
`
`Backup Counsel: Yuezhong Feng (Reg. No. 58,657); Tel: 312.321.4200
`
`Backup Counsel: Andrea L. Shoffstall (Reg. No. 75,426); Tel: 312.321.4200
`
`Address of lead counsel: Brinks Gilson & Lione, NBC Tower, Suite 3600, 455
`
`N. Cityfront Plaza Dr., Chicago, Illinois 60611-5599.
`
`NOTICE OF THE REAL-PARTIES-IN-INTEREST
`The real-parties-in-interest for this petition are ZTE Corporation and ZTE
`
`(USA), Inc.
`
`NOTICE OF RELATED MATTERS
`The ’978 patent is asserted in the following matters:
`
`• Cywee Group Ltd. v. Google, Inc., Case No. 1-18-cv-00571, (D. Del.);
`
`• Cywee Group Ltd. v. ZTE Corporation et al., Case No. 3-17-cv-02130,
`
`(S.D. Cal.);
`
`• Cywee Group Ltd. v. HTC Corporation et al., Case No. 2-17-cv-00932,
`
`(W.D. Wash.);
`
`
`
`5
`
`
`
`• Cywee Group Ltd. v. Motorola Mobility LLC, Case No. 1-17-cv-00780,
`
`(D. Del.);
`
`• Cywee Group Ltd. v. Huawei Technologies Co., Inc. et al., Case No. 2-
`
`17-cv-00495, (E.D. Tex.);
`
`• Cywee Group Ltd. v. LG Electronics, Inc. et al., Case No. 3-17-cv-
`
`01102, (S.D. Cal.);
`
`• Cywee Group Ltd. v. Samsung Electronics Co. Ltd. et al., Case No. 2-
`
`17-cv-00140, (E.D. Tex.);
`
`• Cywee Group Ltd. v. Apple Inc., Case No. 4-14-cv-01853, (N.D. Cal.);
`
`• Google LLC v. Cywee Group Ltd., IPR2018-01257; and
`
`• Samsung Electronics Co., Ltd. et al v. CyWee Group Ltd., IPR2019-
`
`00534 (filed concurrently with a motion for joinder to IPR2018-01257).
`
`NOTICE OF SERVICE INFORMATION
`Please address all correspondence to the lead counsel at the addresses shown
`
`above.
`
`Petitioner
`
`consents
`
`to
`
`electronic
`
`service
`
`by
`
`
`at:
`
`ZTE_CyweeIPRs@brinksgilson.com;
`
`jsobieraj@brinksgilson.com,
`
`jbeaupre@brinksgilson.com,
`
`yfeng@brinksgilson.com,
`
`and
`
`ashoffstall@brinksgilson.com.
`
`GROUNDS FOR STANDING
`Petitioner hereby certifies that the patent for which review is sought is
`
`
`
`6
`
`
`
`available for inter partes review, and that the Petitioner is not barred or estopped
`
`from requesting an inter partes review on the grounds identified in the petition. This
`
`Petition is being submitted concurrently with a Motion for Joinder and Petitioner
`
`requests institution and joinder with Google LLC v. Cywee Group Ltd., IPR2018-
`
`01257 (“the Google IPR”), which the Board instituted on December 11, 2018. The
`
`district court suit against Google LLC was served on April 19, 2018 (Ex. 1007), while
`
`the suit against certain Huawei entities was served on June 14, 2017 (Ex. 1008).
`
`STATEMENT OF PRECISE RELIEF REQUESTED
`Petitioner respectfully requests that claims 10 and 12 of the ’978 patent be
`
`canceled based on the following grounds:
`
`Ground 1: Claims 10 and 12 are obvious over Zhang and Bachmann.
`
`Ground 2: Claims 10 and 12 are obvious over Liberty and Bachmann.
`
`THRESHOLD REQUIREMENT FOR INTER PARTES REVIEW
`This petition presents “a reasonable likelihood that the Petitioner would
`
`prevail with respect to at least one of the claims challenged in the petition”, 35 U.S.C.
`
`§314(a), as shown in the Grounds explained below.
`
`I.
`
`INTRODUCTION
`The present petition is supported by the declaration of Prof. Majid Sarrafzadeh
`
`(Ex. 1002). Professor Sarrafzadeh holds the title of Distinguished Professor of
`
`Computer Science & Electrical Engineering at the University of California, Los
`
`Angeles. Professor Sarrafzadeh’s CV is included as Exhibit 1003.
`7
`
`
`
`
`The ’978 patent relates to 3D pointing devices. (Ex. 1001, Title). The ’978
`
`patent describes the function of a 3D pointing device as “detecting motions of the
`
`device and translating the detected motions to a cursor display such as a cursor
`
`pointing on the screen…of a 2D display device….” (Ex. 1001, 1:31-33)(Ex. 1002,
`
`¶26). For example, a 3D pointing device could be a kind of computer mouse that
`
`detects movements and rotations of the mouse in three dimensions, allowing the
`
`movements and rotations to be translated into actions on a computer. (Ex. 1001,
`
`1:52-61)(Ex. 1002, ¶26). An example of such a device 110 (and a corresponding
`
`display 120) is shown in Fig. 1 of the ’978 patent, reproduced below:
`
`
`To keep track of the motions and rotations of a 3D pointing device, the ’978
`
`patent proposes using three kinds of sensors: rotation sensors (for detecting the
`
`angular velocity of rotation), accelerometers (for detecting axial accelerations), and
`8
`
`
`
`
`magnetometers (for detecting the local magnetic field). (Ex. 1001, Fig. 4)(Ex. 1002,
`
`¶27). These sensors are mounted in or on the 3D pointing device, and provide
`
`information on the movements and rotations of the device. (Ex. 1002, ¶27).
`
`The ’978 patent also purports to provide methods of using data output from
`
`the rotation sensors, accelerometers and magnetometers to calculate the orientation
`
`of the 3D pointing device. (Ex. 1001, 4:15-57)(Ex. 1002, ¶28). The “orientation”
`
`of the device (also called the “attitude” or “tilt” of the device) is the direction of the
`
`device, e.g. the angles between the device and the axes of any given coordinate
`
`system.1 (Ex. 1001, 1:62-64)(Ex. 1002, ¶28). For example, Fig. 2 of the ’978 patent
`
`shows the same device 110 in a different “orientation”, having been rotated about
`
`the x-axis by 90 degrees:
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`1 Orientation may be expressed in a number of equivalent ways, such as with a
`
`quaternion. (Ex. 1002, ¶¶30-32).
`
`
`
`9
`
`
`
`(Ex. 1001, 2:11-14)(Ex. 1002, ¶28).
`
`While the ’978 patent acknowledges the existence of prior-art 3D pointers
`
`using sensors to detect and calculate orientation, the ’978 patent criticizes the
`
`specific devices mentioned as allegedly unable to calculate orientation accurately.
`
`(Ex. 1001, 2:41-3:52)(Ex. 1002, ¶33). The ’978 patent purports to provide a solution
`
`to the alleged deficiencies of the prior art, by using additional sensors and
`
`“compensating” the output of the sensors to improve the accuracy of the orientation
`
`calculation. (Ex. 1001, 1:22-27).
`
`To “compensate” the output of the sensors, the ’978 patent discloses a
`
`mathematical method using quaternions. (Ex. 1001, 16:5 et seq.)(Ex. 1002, ¶34).
`
`As explained by Professor Sarrafzadeh, a “quaternion” is a way to represent an
`
`orientation (rotation angles) using a four-valued vector. (Ex. 1002, ¶¶30-32).
`
`Quaternion math operations (such as multiplication) are defined differently than for
`
`standard vectors, and can sometimes be used for efficient calculation of rotations.
`
`(Ex. 1002, ¶¶30-32).
`
`
`
`
`
`
`
`
`
`
`
`10
`
`
`
`A basic sketch of the ’978 patent method can be seen in Fig. 7, which is
`
`reproduced at right. The method
`
`of Fig. 7 obtains measured angular
`
`velocities at step 715 (Ex. 1001,
`
`16:27-30) and measured axial
`
`accelerations in step 725 (Ex. 1001,
`
`16:60-64). The method then
`
`calculates a predicted set of axial
`
`accelerations at step 730. (Ex. 1001,
`
`17:2-9). By comparing the actual
`
`and predicted accelerations (step
`
`735),
`
`the method purports
`
`to
`
`improve the estimate of orientation
`
`(called the “updated state (3rd quaternion)” in box 735). (Ex. 1001, 18:25-55)(Ex.
`
`1002, ¶34).
`
`Prosecution History and Issued Claims
`A.
`This petition challenges independent claim 10 and dependent claim 12. As
`
`originally filed, claim 10 (then numbered claim 12), read as follows:
`
`“12. A method for compensating rotations of a 3D pointing device,
`comprising:
`generating an orientation output associated with an orientation of
`11
`
`
`
`
`
`the 3D pointing device associated with three coordinate axes of a global
`reference frame associated with Earth;
`generating a rotation output associated with a rotation of the 3D
`pointing device associated with three coordinate axes of a spatial
`reference frame associated with the 3D pointing device; and
`using the orientation output and the rotation output to generate a
`transformed output associated with a fixed reference frame associated
`with a display device.”
`
`(Ex. 1009, p. 044-045). The claim was thus directed to generating a rotation output
`
`(i.e. output of rotation sensors), calculating orientation output, and then somehow
`
`“using” orientation output and rotation output to generate a “transformed output”.
`
`The Examiner initially rejected all original claims for double patenting, and
`
`as anticipated or obvious over U.S. Pat. Pub 2009/0262074 to Nasiri. (Ex. 1009, pp.
`
`071-089). The applicants responded by requesting an interview. (Ex. 1009, pp. 060-
`
`066). In the interview request, the applicants’ representative argued that Nasiri did
`
`not teach using a global reference frame associated with Earth, and that Nasiri “only
`
`briefly talks about ‘magnetometers’” (Ex. 1009, pp. 060-066).
`
`The applicants then submitted an amendment. (Ex. 1009, pp. 040-055). In
`
`the amendment, the applicants modified claim 12 to add several limitations. First,
`
`the applicants added language to claim 12 requiring generating signal sets associated
`
`with accelerometers and magnetometers. (Ex. 1009, pp. 044-045). Second, the
`
`applicants specified that the “orientation output” must be “based on the first signal
`12
`
`
`
`
`
`set, the second signal set and the rotation output or based on the first signal set and
`
`the second signal set”. (Id.). Third, the applicants specified that “the orientation
`
`output and the rotation output is generated by a nine-axis motion sensor module”,
`
`and that a “resultant deviation including a plurality of deviation angles” must be
`
`obtained “using” a “plurality of measured magnetisms Mx, My, Mz and a plurality
`
`of predicted magnetism Mx', My' and Mz' for the second signal set.” (Id.).
`
`Following the amendment, the Examiner allowed the claims without
`
`further comment. (Ex. 1009, pp. 024-027). As issued, independent claim 10
`
`reads as follows:
`
`“10. A method for compensating rotations of a 3D pointing device,
`comprising:
`generating an orientation output associated with an orientation of
`the 3D pointing device associated with three coordinate axes of a global
`reference frame associated with Earth;
`generatinq [sic] a first signal set comprising axial accelerations
`associated with movements and rotations of the 3D pointing device in
`the spatial reference frame;
`generating a second signal set associated with Earth's magnetism;
`generating the orientation output based on the first signal set, the
`second signal set and the rotation output or based on the first signal set
`and the second signal set;
`generating a rotation output associated with a rotation of the 3D
`pointing device associated with three coordinate axes of a spatial
`reference frame associated with the 3D pointing device; and
`13
`
`
`
`
`
`using the orientation output and the rotation output to generate a
`transformed output associated with a fixed reference frame associated
`with a display device, wherein the orientation output and the rotation
`output is generated by a nine-axis motion sensor module;
`obtaining one or more resultant deviation including a plurality of
`deviation angles using a plurality of measured magnetisms Mx, My, Mz
`and a plurality of predicted magnetism Mx', My' and Mz' for the second
`signal set.”
`
`II. CLAIM CONSTRUCTION
`“A claim in an unexpired patent shall be given its broadest reasonable
`
`construction in light of the specification of the patent in which it appears”. 37 C.F.R.
`
`§42.100(b); Cuozzo Speed Techs., LLC v. Lee, 195 L. Ed. 2d 423 (2016). For this
`
`proceeding, claim terms are presumed to take on their broadest reasonable ordinary
`
`meaning, which is explained in certain instances below. The constructions below
`
`are for the purpose of this petition only, and Petitioner reserves the right to change
`
`these constructions as appropriate in future proceedings. Petitioner also does not
`
`concede, by seeking this petition, that the challenged claims are of definite scope or
`
`properly described under 35 U.S.C. §112. Petitioner hereby discloses a recent claim
`
`construction order in Cywee v. Motorola, 17-cv-00780 (Ex. 1051), which does not
`
`construe any claim terms construed in this Petition.
`
`A. Claim 10—“spatial reference frame” and similar terms
`Claim 10 uses the phrases “spatial reference frame” and “spatial reference
`
`
`
`14
`
`
`
`frame associated with the 3D pointing device”. These phrases should be interpreted
`
`to mean “a reference frame associated with the 3D pointing device, which always
`
`has its origin at the same point in the device and in which the axes are always fixed
`
`with respect to the device”. (Ex. 1002, ¶37).
`
`The ’978 patent states as follows concerning the spatial reference frame:
`
`“There are two reference frames, such as the spatial pointer
`reference frame and the display frame, associated with the pointing
`device 110 and the display device 120, respectively. The first
`reference frame or spatial pointer reference frame associated with the
`pointing device 110 is defined by the coordinate axes XP, YP and ZP
`as shown in FIG. 1.”
`
`(Ex. 1001, 1:39-1:45)(Emphasis added)(Ex. 1002, ¶40). Thus, the “spatial pointer
`
`reference frame” is shown by the coordinate axes XP, YP and ZP in Fig. 1. Figure 1
`
`
`
`15
`
`
`
`is reproduced here:
`
`
`
`(Ex. 1002, ¶41). As can be seen from Fig. 1, the spatial pointer reference frame is a
`
`reference frame associated with the 3D pointing device, which has its origin at a
`
`point in the device. (Ex. 1002, ¶¶41-45).
`
`Furthermore, as shown in Fig. 2, when the device is rotated, the axes XP, YP
`
`and ZP rotate with the device. (Ex. 1002, ¶¶45-46). Figure 2 is reproduced below,
`
`and shows a 90-degree roll of the device, with correspondingly rotated axes YP and
`
`ZP:
`
`
`
`16
`
`
`
`
`
`(Ex. 1002, ¶¶42-43). For that reason, in the spatial pointer reference frame, the
`
`origin and axes of the frame stay fixed with respect to the device. (Ex. 1002, ¶¶42-
`
`46). Note that the ’978 patent treats each of the phrases “spatial reference frame”
`
`and “spatial pointer reference frame” as referring to a device-centered frame of
`
`reference analogous to the XP, YP and ZP axes. (Ex. 1001, 9:19-20, 1:39-47, 3:6-
`
`7)(Ex. 1002, ¶39-42). Because “spatial reference frame” already refers to a frame
`
`with its origin in the device, the longer phrase “spatial reference frame associated
`
`with the 3D pointing device” has the same meaning, as CyWee concedes. (Ex. 1002,
`
`¶¶38-41; Ex. 1010, p. 2).
`
`Thus, the phrases “spatial reference frame” and “spatial reference frame
`
`associated with the 3D pointing device” should both be interpreted to mean “a
`
`reference frame associated with the 3D pointing device, which always has its origin
`
`
`
`17
`
`
`
`at the same point in the device and in which the axes are always fixed with respect
`
`to the device” (Ex. 1002, ¶¶37-47). Cywee agreed to these constructions during a
`
`co-pending litigation. (Ex. 1010, p. 2).
`
`B. Claim 10—“rotation output”
`Claim 10 uses the phrase “rotation output”. In the specification, the ’978
`
`patent makes clear that the rotation output is the output of a rotation sensor (a sensor
`
`that detects rotation). For example, the ’978 patent states:
`
`“The rotation sensor generates a rotation output associated with a
`rotation of the 3D pointing device associated with three coordinate
`axes of a spatial reference frame associated with the 3D pointing
`device.”
`
`(Ex. 1001, 7:61-64)(Emphasis added)(Ex. 1002, ¶50).
`
`Thus, “rotation output” should be interpreted in accordance with the
`
`specification as “output of a rotation sensor”. (Ex. 1002, ¶¶48-52).
`
`GROUNDS
`Ground 1. Claims 10 and 12 are obvious over Zhang in view of Bachmann.
`Claims 10 and 12 are unpatentable as obvious over U.S. Pat. App. Pub.
`
`2004/0095317 (“Zhang”)(Ex. 1005), in view of U.S. Pat. No. 7,089,148
`
`(“Bachmann”)(Ex. 1004).
`
`Zhang was published on May 20, 2004, and is thus prior art under pre-AIA
`
`35 U.S.C. §102(b). Bachmann issued on August 8, 2006, and is thus also prior art
`18
`
`
`
`
`
`under pre-AIA 35 U.S.C. §102(b). Zhang and Bachmann are analogous art, because
`
`they are in the same field and reasonably related to the problems facing the named
`
`inventors, as shown by the discussion below.
`
`Neither Zhang nor Bachmann are listed as prior art of record on the face of
`
`the ’978 patent.
`
`Overview of the Combination
`Claim 10 is directed to a method for compensating rotations of a 3D pointing
`
`device. The combination of Zhang and Bachmann, broadly speaking, uses Zhang’s
`
`3D pointing device together with Bachmann’s extra sensors and method for
`
`compensating rotations.
`
`Zhang teaches a (sic) “a handheld pointing device” that is used for a
`
`“computer pointing control system”. (Ex. 1005, Abstract)(Ex. 1002, ¶53). Such
`
`a computer pointing control system is shown, for example, in Fig. 2 of Zhang
`
`(reproduced below), where the handheld device (a 3D pointer) has reference numeral
`
`100:
`
`
`
`19
`
`
`
`Inside Zhang’s device 100, there are several sensors that detect the orientation
`
`
`
`of the device. Zhang explains:
`
`“A universal pointing control system for televisions and computer
`displays is disclosed. The system is comprised of a remote handheld
`device, a display control unit and a command delivery unit. The
`remote handheld device includes a set of orientation sensors that
`detect the device’s current orientation.”
`
`(Ex. 1005, ¶0008)(Emphasis added)(Ex. 1002, ¶54).
`
`
`
`
`
`20
`
`
`
`Zhang discloses that the device 100 has several different orientation sensors.
`
`The orientation sensors are arranged on a circuit board in the housing of the device,
`
`as shown in Fig. 3 of Zhang,
`
`reproduced at right.
`
`In Fig. 3,
`
`numeral 160 is the circuit board,
`
`while numerals 120 and 130 are
`
`sensors.
`
`(Ex. 1005, ¶0025)(Ex.
`
`1002, ¶¶55-56). Numeral 120 is “a
`
`two-axis magnetic field sensor 120
`
`[that] is used to detect the device’s
`
`orientation relative to the direction of the earth’s magnetic field 25.” (Ex. 1005,
`
`¶0026)(Ex. 1002, ¶¶57-58). Numeral 130 is an “accelerometer sensor 130 [that]
`
`contains two orthogonally arranged acceleration detectors.” (Ex. 1005, ¶0027)(Ex.
`
`1002, ¶¶57-58). Numeral 110 is a microcontroller for performing calculations. (Ex.
`
`1005, ¶0025)(Ex. 1002, ¶59).
`
`A system diagram of Zhang’s device 100 is shown in Fig. 5, reproduced below
`
`at right. (Ex. 1005, ¶0029)(Ex. 1002, ¶59). In Fig. 5, the two sets of two sensors
`
`(magnetometers 120 and accelerometers 130) are shown on the left
`
`
`
`21
`
`
`
`side (the Petitioner has placed a red-
`
`dashed box around the numerals 120
`
`and 130). These sensors output signals
`
`to circuits 111-112, 121-124 and 131-
`
`134. (Ex. 1005, ¶0029)(Ex. 1002,
`
`¶60). These circuits condition the
`
`sensor output, convert it to digital
`
`format, and pass the digital data to the
`
`microcontroller (MCU) 110. (Ex.
`
`1005, ¶0029)(Ex. 1002, ¶60). The
`
`MCU 110 determines the device’s
`
`orientation, including azimuth and
`
`inclination angles (yaw and pitch). (Ex. 1005, ¶0029)(Ex. 1002, ¶60). These angles
`
`are shown in Figs. 4(a) and 4(b), reproduced below.
`
`
`
`
`
`22
`
`
`
`(Ex. 1002, ¶58). After Zhang’s device calculates its own orientation, Zhang’s
`
`system translates those angles into a display command (e.g. moving a cursor), by
`
`translating the angles into screen coordinates. (Ex. 1005, ¶¶0024, 0030)(Ex. 1002,
`
`
`
`¶60).
`
`Zhang’s primary embodiment has a four-axis sensor module (compared to the
`
`“nine-axis sensor module” required by claim 10). Zhang explains that more sensors
`
`can be used, and that different kinds of sensors can be used. For example, Zhang
`
`
`
`23
`
`
`
`states that gyro sensors (angular rate sensors that measure “rotation output”) could
`
`be used:
`
`“The orientation sensors’ mechanisms are shown in FIGS. 4a and 4b.
`The orientation sensor demonstrated in FIG. 4a is a magnetic field
`sensor, whereas the one in FIG. 4b is an accelerometer sensor.
`However, the orientation detection may not be limited to these
`types of sensors. Other sensors, for example, a gyro sensor, can also
`be used in the pointing control system.”
`
`(Ex. 1005, ¶0026)(Emphasis added)(Ex. 1002, ¶61). Zhang also mentions that
`
`accelerometers, magnetometers and gyro (angular rate) sensors can be used in
`
`combination. (Ex. 1005, ¶¶0006, 0026, claim 2)(Ex. 1002, ¶62).
`
`Bachmann, in turn, provides an example of a nine-axis sensor system that
`
`combines accelerometers, magnetometers and angular rate detectors (e.g.
`
`gyroscopes), as suggested by Zhang. Bachmann, for example, states:
`
`the magnetometers and
`embodiment,
`sensor
`another
`“In
`accelerometers are supplemented with angular rate detectors
`configured to detect the angular velocity of the sensor (comprising so-
`called Magnetic, Angular Rate, Gravity (MARG) sensors). Each
`MARG sensor contains angular rate detectors, accelerometers, and
`magnetometers.”
`
`(Ex. 1004, 7:34-41)(Emphasis added)(Ex. 1002, ¶64). In Bachmann’s system, each
`
`type of sensor is a three-axis sensor, making the entire system (3 sensor types x 3
`
`
`
`24
`
`
`
`axes per type) a nine-axis system. (Ex. 1002, ¶65).
`
`Bachmann teaches combining sensor data using an attitude estimation filter to
`
`produce an estimate of the orientation of a tracked object. Bachmann explains:
`
`“[T]he filter inputs are from a three-axis accelerometer (h1 h2 h3) 31,
`a three-axis magnetometer (b1 b2 b3) 32, and a three-axis angular
`rate sensor (p, q, r) 33. Its output is a quaternion representation of the
`orientation of the tracked object q̂ 39.”
`
`(Ex. 1004, 10:10-14)(Emphasis added)(Ex. 1002, ¶66). Bachmann thus takes the
`
`output of accelerometer, magnetometer and angular rate sensors, and uses these
`
`sensor outputs to calculate an orientation of a tracked device. (Ex. 1002, ¶67).
`
`To calculate the orientation from sensor inputs, Bachman uses a filter.
`
`Bachmann’s filter mirrors
`
`the claimed calculations of
`
`the ’978 patent. (Ex. 1002,
`
`¶68). A control diagram of
`
`Bachmann’s filter process is
`
`shown in Fig. 3, reproduced
`
`at
`
`right, where
`
`the
`
`Petitioner has drawn a red-
`
`dashed box around the
`
`output, q̂ , in the lower right.
`
`
`25
`
`
`
`(Ex. 1004, Fig. 3)(Ex. 1002, ¶68). The output q̂ is a quaternion representing the
`
`orientation of the tracked object in space. (Ex. 1004, 10:10-14)(Ex. 1002, ¶68).
`
`Bachmann’s filter as shown in Fig. 3 receives inputs from three sets of sensors
`
`(accelerometers, magnetometers and angular-rate sensors) marked 31, 32 and 33, on
`
`the left side of Fig. 3. These sensors are shown in red-dashed boxes, below:
`
`
`
`(Ex. 1002, ¶69).
`
`The output of the angular rate sensors (33) is used to calculate the orientation
`
`of the device q̂ . The calculation is shown in the boxes along the red-dashed line that
`
`has been added to the lower portion of Fig. 3, below:
`
`
`
`26
`
`
`
`
`
`(Ex. 1002, ¶70). In the figure, the output of the angular rate sensors (33) is a set of
`
`measured angular rates of rotation (p, q, r) about three axes. (Ex. 1004, 10:10-
`
`14)(Ex. 1002, ¶70). These rates are converted, in box 37, to a rate quaternion q̇ . (Ex.
`
`1004, 10:15-36)(Ex. 1002, ¶70). To the rate quaternion q̇ is added a correction factor
`
`q̇ ε (which will be explained below), to yield a corrected rate quaternion
`
`.
`
`(Ex. 1004, 10:15-65)(Ex. 1002, ¶¶70-74). The corrected rate quaternion
`
`is
`
`then integrated in box 42 and normalized to a unit length in box 43, to yield the
`
`orientation quaternion at the output, q̂ . (Ex. 1004, 10:15-65)(Ex. 1002, ¶¶70-74).
`
`Bachmann’s filter shown in Fig. 3 takes advantage of extra sensor
`
`
`
`27
`
`
`
`measurements from the accelerometers and magnetometers via the previously-
`
`mentioned correction factor, q̇ ε. Bachmann calculates this correction factor q̇ ε in
`
`steps 34-41 of Fig. 3. There, Bachmann first obtains actual sensor measurements
`
`from the accelerometers2 (31) and magnetometers (32), forming a six-valued
`
`measurement vector (h1 h2 h3 b1 b2 b3), as shown in box 34. (Ex. 1004, 10:10-14,
`
`3:13-17, 8:47-51)(Ex. 1002, ¶72). These six measurement values include three
`
`measurements of acceleration along the X, Y and Z axes of the sensors, and three
`
`measurements for magnetism, also along the X, Y and Z axes of the sensors. (Id.).
`
`The six actual measurements are then compared to six predicted measurements
`
`found in the vector y̅ (q̂ ), by subtracting the predicted measurements y̅ (q̂ ) from the
`
`actual measurements (h1 h2 h3 b1 b2 b3). (Ex. 1004, 8:63-9:18, 17:12-22)(Ex. 1002,
`
`¶72). This forms a six-valued error vector ε̅(q̂ ), numbered 36. (Ex. 1004, 17:12-
`
`22, 9:9-14)(Ex. 1002, ¶72).
`
`The six-valued error vector ε̅(q̂ ) is essentially a measure of how actual
`
`accelerometer and magnetometer measurements differ from what the filter predicts
`
`those measurements should be based on the angular rate sensor output. (Ex. 1004,
`
`17:12-22, 9:9-14)(Ex. 1002, ¶73). The difference ε̅(q̂ ) is utilized in boxes 38 and 41.
`
`
`2 The accelerometer measurements are first low-pass filtered to remove sudden
`
`accelerations. (Ex. 1004, 8:12-20)(Ex. 1002, ¶125).
`
`
`
`28
`
`
`
`There, the filter selects a correction factor q̇ ε that will minimize ε̅(q̂ ). (Ex. 1004, 9:9-
`
`35)(Ex. 1002, ¶73). That is, the filter will choose a correction factor q̇ ε that, when
`
`added to q̇ , will minimize the difference between the actual measurements (h1 h2 h3
`
`b1 b2 b3) and the predicted measurements for those same values. (Ex. 1004, 9:9-
`
`35)(Ex. 1002, ¶73). This has the effect of compensating the orientation output of
`
`the filter, q̂ . (Ex. 1002, ¶73).
`
`The combination proposes using the 3D pointer of Zhang (modified to include
`
`additional sensors), together with Bachmann’s filter process to calculate a device
`
`orientation. Once a device orientation has been calculated, it can be converted to the
`
`coordinate system of a display device, as disclosed in both Zhang and Bachmann.
`
`The combination can be illustrated with the Figure below, created by the Petitioner,
`
`showing the relevant modifications to the Zhang pointer (adding sensors and using
`
`Bachmann’s filter calculations):
`
`
`
`29
`
`
`
`(Ex. 1002, ¶93).
`
`Rationale for the Combination
`It would have been obvious to a person of skill in the relevant timeframe3 to
`
`use Zhang’s 3D pointer with Bachmann’s sensors and filter calculations. As
`
`
`
`
`3 The first provisional application in the chain of applications leading to the ’978
`
`patent was filed on January 6, 2010. Petitioner disagrees that this is the proper
`
`priority date, and notes that subsequent applications were continuations-in-part
`
`applications but for purposes of this Petition assumes that this date applies.
`
`
`
`30
`
`
`
`discussed above in the Overview section beginning on page 23, Zhang expressly
`
`states that additional sensors can be used, and in particular, that “gyro” (angular rate)
`
`sensors can be used. (Ex. 1005, ¶¶0006, 0025, 0026, claim 2)(Ex. 1002, ¶¶61-63,
`
`94). Thus, it would have been obvious to add sensors to Zhang, including the angular
`
`rate sensors of Bachmann, based on Zhang’s express suggestion. Furthermore, a
`
`person of ordinary skill would have understood that additional sensors, and
`
`additional types of sensors, would have yielded at least two benefits. (Ex. 1002,
`
`¶95). First, additional sensors (in particular sensor axes) an