`Tel: 571-272-7822
`
`
`
`
`
`
`Paper 50
`Entered: January 12, 2015
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`
`TOYOTA MOTOR CORPORATION,
`Petitioner,
`
`v.
`
`AMERICAN VEHICULAR SCIENCES LLC,
`Patent Owner.
`
`
`
`Case IPR2013-00424
`Patent 5,845,000
`
`
`
`
`Before JAMESON LEE, TREVOR M. JEFFERSON,
`and LYNNE E. PETTIGREW, Administrative Patent Judges.
`
`
`JEFFERSON, Administrative Patent Judge.
`
`
`
`
`
`
`
`FINAL WRITTEN DECISION
`35 U.S.C. § 318(a) and 37 C.F.R. § 42.73
`
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`I.
`
`INTRODUCTION
`
`Toyota Motor Corporation (“Toyota” or “Petitioner”) filed a petition
`
`requesting an inter partes review of claims 10, 11, 16, 17, 19, 20, and 23 of
`
`U.S. Patent No. 5,845,000 (Ex. 1001, “the ’000 patent”). Paper 2 (“Pet.”).
`
`On January 14, 2014, we instituted an inter partes review of claims 10, 11,
`
`16, 17, 19, 20, and 23 on three grounds of unpatentability. Paper 16 (“Dec.
`
`on Inst.”). American Vehicular Sciences (“AVS” or “Patent Owner”) filed a
`
`Patent Owner Response (Paper 29, “PO Resp.”) and Petitioner filed a Reply
`
`To Patent Owner’s Response (Paper 34, “Reply”).
`
`Patent Owner did not file a motion to amend the claims.
`
`A consolidated oral hearing for IPR2013-00419 and IPR2013-00424,
`
`both involving the same Petitioner and the same Patent Owner, was held on
`
`August 18, 2014. A transcript of the joint hearing was entered in the record.
`
`Paper 49 (“Tr.”).
`
`We have jurisdiction under 35 U.S.C. § 6(c). This final written
`
`decision is issued pursuant to 35 U.S.C. § 318(a) and 37 C.F.R. § 42.73.
`
`For the reasons that follow, we determine that Petitioner has not
`
`shown by a preponderance of the evidence that claims 10, 11, 16, 17, 19, 20,
`
`and 23 of the ’000 patent are unpatentable.
`
`A. Related Proceedings
`
`Petitioner and Patent Owner notify us that the ’000 patent has been
`
`asserted by AVS in the following district court cases: (1) American
`
`Vehicular Sciences LLC v. Toyota Motor Corp., Civil Action No. 6:12-CV-
`
`406 (E.D. Tex.) (filed June 25, 2012); (2) American Vehicular Sciences LLC
`2
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`v. BMW Grp. A/K/A BMW AG, Civil Action No. 6:12-CV-413 (E.D. Tex.)
`
`(filed June 25, 2012); and (3) American Vehicular Sciences LLC v.
`
`Mercedes-Benz U.S. Intl., Inc., Civil Action No. 6:13-CV-308 (E.D. Tex.)
`
`(filed April 3, 2013). Pet. 1; Paper 23, 2–3.
`
`B. The ’000 Patent
`
`The ’000 patent is directed to a vehicle interior monitoring system that
`
`monitors, identifies, and locates occupants and other objects in the passenger
`
`compartment of a vehicle and objects outside of the vehicle. Ex. 1001,
`
`Abstract: 1–4. Objects are illuminated with electromagnetic radiation, and a
`
`lens is used to focus the illuminated images onto the arrays of a charge
`
`coupled device (CCD). Id. at Abstract: 1–9, 7:26–40. Computational means
`
`using trained pattern recognition analyzes the signals received at the CCD to
`
`classify, identify, or locate the contents of external objects, which, in turn,
`
`are used to affect the operation of other vehicular systems. Id. at Abstract:
`
`10–12. The ’000 patent discloses that a vehicle computation system uses a
`
`“trainable or a trained pattern recognition system” which relies on pattern
`
`recognition to process signals and to “identify” an object exterior to the
`
`vehicle or an object within the vehicle’s interior. Id. at 3:21–44.
`
`Figures 7 and 7A, reproduced below, illustrate portions of the sensor
`
`system that use transmitters, receivers, circuitry, and processors to perform
`
`pattern recognition of external objects in anticipation of a side-impact
`
`collision:
`
`3
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`
`
`Figure 7, with Figure 7A inset, depicts vehicle 720 approaching the
`
`side of another vehicle 710 and shows transmitter 730 and receivers 734 and
`
`736. Ex. 1001, 9:48–52, 18:28–40. Figure 7A provides a detailed view of
`
`the electronics that drive transmitter 730 and circuitry 744 containing neural
`
`computer 745 to process signals returned from the receivers using pattern
`
`recognition. Id. at 18:33–40.
`
`Figure 8 also illustrates an exterior monitoring system and is
`
`reproduced below:
`
`4
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`
`
`Figure 8 depicts a system for detecting the headlights or taillights of
`
`other vehicles used in conjunction with an automatic headlight dimming
`
`system. Ex. 1001, 9:54–58. CCD array in Figure 8 is designed to be
`
`sensitive to visible light and does not use a separate source of illumination as
`
`depicted in Figure 7. Id.
`
`The Summary of the Invention discusses an invention related to
`
`detection of objects in the interior of the vehicle and objects external to the
`
`vehicle. Id. at 7:25–30. Specifically, external objects are illuminated with
`
`“electromagnetic, and specifically infrared, radiation,” and lenses are used to
`
`focus images onto one or more CCD arrays. Id. The disclosure further
`
`states that the invention provides (1) an “anticipatory sensor” located within
`
`the vehicle to “identify about-to-impact object[s] in the presence of snow
`
`and/or fog,” (2) “a smart headlight dimmer system” to sense and identify
`
`headlights and taillights and distinguish them from other reflective surfaces,
`
`and (3) blind spot detection. Id. at 8:37–53.
`
`5
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`C. Illustrative Claims
`
`We instituted inter partes review of independent claims 10, 16 and 23,
`
`and dependent claims 11, 17, 19, and 20. Independent claims 10, 16, and 23,
`
`provided below with disputed limitations in italics, are illustrative of the
`
`subject matter of the ’000 patent:
`
`In a motor vehicle having an interior and an exterior, a
`10.
`monitoring system for monitoring at least one object exterior to
`said vehicle comprising:
`a) transmitter means for transmitting electromagnetic
`waves to illuminate the at least one exterior object;
`b) reception means for receiving reflected
`electromagnetic illumination from the at least one exterior
`object;
`c) processor means coupled to said reception means for
`processing said received illumination and creating an electronic
`signal characteristic of said exterior object based thereon;
`d) categorization means coupled to said processor means
`for categorizing said electronic signal to identify said exterior
`object, said categorization means comprising trained pattern
`recognition means for processing said electronic signal based
`on said received illumination from said exterior object to
`provide an identification of said exterior object based thereon,
`said pattern recognition means being structured and arranged to
`apply a pattern recognition algorithm generated from data of
`possible exterior objects and patterns of received
`electromagnetic illumination from the possible exterior objects;
`and
`
`e) output means coupled to said categorization means for
`affecting another system in the vehicle in response to the
`identification of said exterior object.
`
`In a motor vehicle having an interior and an exterior, an
`16.
`automatic headlight dimming system comprising:
`
`6
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`a) reception means for receiving electromagnetic
`radiation from the exterior of the vehicle;
`b) processor means coupled to said reception means for
`processing the received radiation and creating an electronic
`signal characteristic of the received radiation;
`c) categorization means coupled to said processor means
`for categorizing said electronic signal to identify a source of the
`radiation, said categorization means comprising trained pattern
`recognition means for processing said electronic signal based
`on said received radiation to provide an identification of the
`source of the radiation based thereon, said pattern recognition
`means being structured and arranged to apply a pattern
`recognition algorithm generated from data of possible sources
`of radiation including lights of vehicles and patterns of received
`radiation from the possible sources; and
`d) output means coupled to said categorization means for
`dimming the headlights in said vehicle in response to the
`identification of the source of the radiation.
`
`
`23. A method for affecting a system in a vehicle based on an
`object exterior of the vehicle, comprising the steps of:
`a) transmitting electromagnetic waves to illuminate the
`exterior object;
`b) receiving reflected electromagnetic illumination from
`the object on an array;
`c) processing the received illumination and creating an
`electronic signal characteristic of the exterior object based
`thereon;
`d) processing the electronic signal based on the received
`illumination from the exterior object to identify the exterior
`object, said processing step comprising the steps of generating
`a pattern recognition algorithm from data of possible exterior
`objects and patterns of received electromagnetic illumination
`from the possible exterior objects, storing the algorithm within
`a pattern recognition system and applying the pattern
`
`7
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`recognition algorithm using the electronic signal as input to
`obtain the identification of the exterior object; and
`e) affecting the system in the vehicle in response to the
`identification of the exterior object.
`
`
`Ex. 1001, 21:35–61, 22:17–39, 23:19–24:2 (emphases added).
`
`D. The Asserted Grounds
`
`The asserted grounds of unpatentability in this inter partes review are
`
`as follows (Dec. on Inst. 45):
`
`Reference[s]
`Lemelson1
`Lemelson and Asayama2
`Lemelson and Yanagawa3
`
`Basis
`§ 102(e)
`§ 103(a)
`§ 103(a)
`
`Claims Challenged
`10, 11, 16, 17, 19, 20, and 23
`10, 11, 19, and 23
`16, 17, and 20
`
`II. DISCUSSION
`
`A. Claim Construction
`
`In the Decision on Institution, we applied the broadest reasonable
`
`claim interpretation and interpreted certain claim terms as follows:
`
`
`
`1 U.S. Patent No. 6,553,130, issued on April 22, 2003 (Ex. 1002,
`“Lemelson”) from a continuation application of U.S. Application No.
`08/105,304 filed on Aug. 11, 1993 (Ex. 1003, “the ’304 appl.”).
`2 U.S. Patent No. 5,214,408, issued on May 25, 1993 (Ex. 1004,
`“Asayama”).
`3 Japanese Unexamined Patent Application Publication No. S62-131837,
`June 15, 1987 (Ex. 1008, “Yanagawa Japanese”). Citations herein are to the
`English translation of Ex. 1008 (Ex. 1009, “Yanagawa”).
`
`8
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`Claim Term
`“pattern recognition
`algorithm”
`
`“trained pattern recognition
`means . . .”
`
`“identify” and
`“identification”
`“transmitter means for
`transmitting . . .”
`
`“reception means for
`receiving . . .”
`“processor means . . . for
`processing”
`“categorization means
`. . . for categorizing”
`“output means . . .”
`
`“dimming the headlights”
`
`“measurement means for
`measuring . . .”
`“wherein said categories
`further comprise radiation
`from taillights of a vehicle-
`in-front”
`
`
`
`Construction
`“an algorithm which processes a signal that
`is generated by an object, or is modified by
`interacting with an object, for determining
`to which one of a set of classes the object
`belongs”
`“a neural computer or microprocessor
`trained for pattern recognition, and
`equivalents thereof”
`“determining that the object belongs to a
`particular set or class”
`“infrared, radar, and pulsed GaAs laser
`systems” and “transmitters which emit
`visible light”
`
`“a CCD array and CCD transducer”
`
`recited processor provides sufficient
`structure
`“a neural computer, a microprocessor, and
`their equivalents”
`“electronic circuit or circuits capable of
`outputting a signal to another vehicle
`system”
`“decreasing the intensity or output of the
`headlight to a lower level of illumination”
`
`recited radar provides sufficient structure
`
`“categorizing radiation from taillights of a
`vehicle-in-front, which may include
`additional types of radiation”
`
`Dec. on Inst. 9–26. AVS does not contest these constructions for purposes
`
`of this proceeding, PO Resp. 9–12, and Toyota does not dispute these
`
`9
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`constructions in its Reply. We maintain these constructions for this Final
`
`Written Decision.
`
`B. Asserted Grounds of Unpatentability Based, in Part, on Lemelson
`
`The central and dispositive issue in the parties’ dispute as to whether
`
`the challenged claims are unpatentable based, in part, on Lemelson turns on
`
`whether or not Lemelson discloses the “generating the pattern recognition
`
`algorithm” limitations of independent claims 10 and 16 (“pattern recognition
`
`algorithm generated from . . .”) and independent claim 23 (“generating a
`
`pattern recognition algorithm from . . .”). PO Resp. 12–21; Reply 3–11.
`
`Although we construed “trained pattern recognition algorithm” in our
`
`Decision on Institution, we did not provide an express construction for the
`
`“generated from” language following that term in the claims. For this Final
`
`Written Decision, we construe the “generated from” limitation according to
`
`its broadest reasonable interpretation in light of the specification of the ’057
`
`patent. See 37 C.F.R. § 42.100(b).
`
`1. “a pattern recognition algorithm generated from . . .” (claims 10
`and 16) and “generating a pattern recognition algorithm from
`. . .” (claim 23)
`
`AVS contends that the claim limitations for generating the pattern
`
`recognition algorithm in claims 10, 16, and 23 require a specific type of
`
`training to generate the claimed algorithm. PO Resp. 7, 12. AVS relies on
`
`the Declaration of Professor Cris Koutsougeras, PhD (Ex. 2002) to support
`
`its contention that the ’000 patent discloses and claims a specific method for
`
`training the algorithm using (1) data of possible exterior objects or data of
`
`10
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`possible radiation sources, and (2) patterns of received waves from the
`
`possible sources. PO Resp. 7 (citing Ex. 2002 ¶¶ 19, 20, 53). AVS asserts
`
`that the type of training the ’000 patent discloses is the use of “real radar
`
`waves” or is based on real radar waves as the received radar waves from
`
`possible objects used to generate the algorithm. See PO Resp. 7 (citing Ex.
`
`2002 ¶¶ 19–20). AVS contrasts the use of real waves (or data) to train the
`
`pattern recognition system as recited in the claims with other methods of
`
`training, such as the use of simulated data (e.g., a computer simulation of
`
`radar waves). PO Resp. 7–8 (citing Ex. 2002 ¶¶ 49, 57–64).
`
`Toyota argues that the “generated from” language of independent
`
`claims 10, 16 and 23, is not limited to training with real data because the
`
`claims merely require that the data and patterns used to train the algorithm
`
`represent possible exterior objects and received electromagnetic illumination
`
`as recited in the independent claims. Reply 5. Toyota argues that because
`
`the claims refer to “data of” and “patterns of” and not “data from” and
`
`“patterns from,” the claim language encompasses training using simulated
`
`data and patterns that represent possible objects and received waves,
`
`respectively. Reply 5.
`
`As neither party asserts that these terms are defined in the
`
`specification, we refer to the terms’ ordinary and customary meaning as they
`
`would be understood by one of ordinary skill in the art in the context of the
`
`entire disclosure. In re Translogic Tech. Inc., 504 F.3d 1249, 1257 (Fed.
`
`Cir. 2007).
`
`11
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`With respect to the ’000 patent written description, the limited
`
`discussion of training a neural network describes that a large number of real
`
`possible objects is used to train such a network to detect objects in the
`
`interior of a vehicle. Ex. 1001, 16:61–17:2 (discussing the use of real
`
`interior objects to train a neural network). Thus, the sole example of training
`
`described in the specification uses real objects.
`
`We are not persuaded by Toyota’s arguments that the claim language
`
`reliance on the term “of” rather than “from” alters the interpretation of the
`
`claim. We determine that neither the specification nor the claim language in
`
`context supports such parsing.
`
`In context, we find that the plain language of the limitations at issue in
`
`claims 10 and 23 expressly states that two types of training inputs are
`
`required—both “data of possible exterior objects and patterns of received
`
`electromagnetic illumination from the possible exterior objects.” Similarly,
`
`claim 16, in context, requires two types of training inputs—both “data of
`
`possible sources of radiation including lights of vehicles and patterns of
`
`received radiation from the possible sources.”
`
`In view of the claim language and the description in the ’000 patent of
`
`a training session using signal patterns actually received from real objects,
`
`we see no reasonable basis for interpreting the “generating from . . .”
`
`limitations of claims 10 and 16 (“pattern recognition algorithm generated
`
`from . . .”) and claim 23 (“generating a pattern recognition algorithm
`
`from . . .”) to encompass training of a pattern recognition algorithm using
`
`simulated wave patterns. Therefore, the broadest reasonable construction of
`
`12
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`the claim language at issue requires a pattern recognition algorithm that has
`
`been generated using patterns of waves actually received from possible
`
`exterior objects.
`
`2. Lemelson
`
`Lemelson is directed to a vehicle computer system to monitor and
`
`analyze image information for external objects by identifying objects and the
`
`distance between a vehicle and external object or objects. Ex. 1002,
`
`Abstract, 1:10–16, 2:14–23, 2:39–3:39. Figure 1, reproduced below, shows
`
`a block diagram of the vehicle image analysis computer:
`
`
`
`Figure 1 shows computer control system 10 including microprocessor
`
`11 and image analyzing computer 19. Image analyzing computer 19
`
`employs neural networks and artificial intelligence along with fuzzy logic
`
`13
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`algorithms to identify objects exterior to the vehicle. Id. at 5:15–24, 5:30–
`
`45. The system employs camera 16 and laser scanners to generate image
`
`data which is analyzed by computer 19 to control various vehicle systems,
`
`including warning and display systems, braking systems, and headlight
`
`systems. Id. at 5:45–59, Fig. 1 (items 31, 32, 33, 41, and 42).
`
`Lemelson discloses using image analysis computer 19 in a hazard or
`
`external object avoidance system. Id. at 4:40–43. The imaging system
`
`detects objects and the distance between the vehicle and exterior object, and
`
`affects the operation of other vehicle systems. Id. at 6:9–20.
`
`Figure 2, showing image analysis computer 19 of Figure 1 in further
`
`detail, is reproduced below:
`
`
`
`Figure 2 shows a computer architecture based on neural networks that
`
`use a parallel processing system with dedicated imaging proicessing
`
`hardware. Id. at 6:21–27. The imaging system uses video camera 16,
`
`described as a CCD array, but also may use image intensifying electron gun
`
`14
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`and infrared imaging methods on the front, side, and rear of the vehicle to
`
`capture image data. Id. at 6:31–42.
`
`Lemelson further discloses that image analyzing computer 19 uses
`
`neural network processing that is trained to recognize roadway hazards. Id.
`
`at 8:1–4, 7:47–50. The neural network training in Lemelson “involves
`
`providing known inputs to the network resulting in desired output responses”
`
`and applies various learning algorithms. Id. at 8:5–8 (emphasis added).
`
`3. Anticipation by Lemelson—Claims 10, 11, 16, 17, 19, 20, and 23
`
`To establish anticipation under § 102(e), “all of the elements and
`
`limitations of the claim must be shown in a single prior reference, arranged
`
`as in the claim.” Karsten Mfg. Corp. v. Cleveland Golf Co., 242 F.3d 1376,
`
`1383 (Fed. Cir. 2001). “A claim is anticipated only if each and every
`
`element as set forth in the claim is found, either expressly or inherently
`
`described, in a single prior art reference.” Verdegaal Bros. v. Union Oil Co.
`
`of California, 814 F.2d 628, 631 (Fed. Cir. 1987). “Inherency, however, may
`
`not be established by probabilities or possibilities. The mere fact that a
`
`certain thing may result from a given set of circumstances is not sufficient.”
`
`In re Robertson, 169 F.3d 743, 745 (Fed. Cir. 1999) (citations omitted).
`
`Toyota contends that Lemelson discloses a neural computing network
`
`that uses training involving known inputs and that various learning
`
`algorithms may be applied to the neural computing network. Pet. 19 (citing
`
`Ex. 1002, 7:47–8:24). Toyota relies on the declaration testimony of
`
`Dr. Nikolaos Papanikolopoulos (Ex. 1013) to establish that Lemelson
`
`discloses use of neural networks that are trained to identify and,
`
`15
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`subsequently, differentiate between the types of radiation received as inputs
`
`and that such training uses known inputs. See Pet. 21–22 (citing Ex. 1013
`
`¶¶ 56–59, 63). The support for Toyota’s contention that Lemelson discloses
`
`the training of the neural network disclosed in Lemelson is the statement that
`
`“[t]raining involves providing known inputs to the network resulting in
`
`desired output responses.” Ex. 1002, 8:4–6; see Pet. 19, 25–26; Ex. 1013
`
`¶ 59 (quoting same).4 Toyota’s Petition states that “Lemelson explains how
`
`[a image analyzing computer] may be implemented as a ‘neural computing
`
`network’ that is ‘trained’ using ‘known inputs.’” Pet. 19 (citing Ex. 1002,
`
`7:47–8:24).
`
`AVS contends that Lemelson does not disclose, either expressly or
`
`inherently, the specific type of training of the pattern recognition recited in
`
`independent claims 10, 16, and 23. PO Resp. 12–13. Although AVS admits
`
`that Lemelson discloses a system for identifying objects exterior to a vehicle
`
`
`
`4 The pertinent part of Lemelson cited in Toyota’s claim chart for claim 10
`(and related claims 16 and 23) (Pet. 26, 29, 30) states:
`Neural networks used in the vehicle [] warning system are
`trained to recognize roadway hazards which the vehicle is
`approaching including automobiles, trucks, and pedestrians.
`Training involves providing known inputs to the network
`resulting
`in desired output responses. The weights are
`automatically adjusted based on error signal measurements until
`the desired outputs are generated. Various learning algorithms
`may be applied. Adaptive operation is also possible with on-
`line adjustment of network weights
`to meet
`imaging
`requirements.
`Ex. 1002, 8:1–10; see Pet. 26, 29, 30.
`
`16
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`and discloses using a neural network (a type of pattern recognition
`
`algorithm) to identify such objects, AVS contends that the claim language
`
`requires a specific type of training to generate the claimed algorithm, which
`
`Lemelson fails to disclose. PO Resp. 14. AVS argues that Lemelson fails to
`
`disclose “generating” the neural network (pattern recognition algorithm).
`
`Pet. 14.
`
`AVS further argues that because Lemelson could have involved
`
`generating the pattern recognition algorithm using completely “simulated
`
`data,” it does not disclose using “data from possible exterior objects and
`
`patterns of received waves (e.g., received electromagnetic illumination) from
`
`the possible exterior objects.” PO Resp. 17 (citing Ex. 2002 ¶¶ 60–64).
`
`AVS relies on the testimony of Prof. Koutsougeras to establish that
`
`“[s]imulated data is data that does not include any ‘patterns of
`
`electromagnetic illumination from the possible exterior objects’ or ‘patterns
`
`of received radiation from the possible sources’ of radiation.” Id. Such
`
`simulated data is generated by computers to simulate sensor readings for
`
`object detection. Id. Such simulated data or “made-up data,” AVS
`
`contends, would not constitute data from objects or patterns of waves from
`
`objects. Id. (quoting Ex. 2002 ¶ 58).
`
`AVS asserts that Toyota’s Petition and expert testimony rely only on
`
`the reference in Lemelson to “known inputs” to train the neural computer to
`
`disclose the specified algorithm generating limitations of independent claims
`
`10, 16, and 23. PO Resp. 14. AVS’s expert, Prof. Koutsougeras, testifies
`
`that the reference to “known inputs” in Lemelson relied upon by Toyota is
`
`17
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`silent as to the type of known inputs and could encompass the use of
`
`simulated data for generating a pattern recognition algorithm. PO Resp. 17–
`
`18 (citing Ex. 2002 ¶¶ 57–64).
`
`Similarly, AVS argues that Lemelson’s reference to training with
`
`“known inputs” does not expressly or inherently disclose “data of possible
`
`exterior objects and patterns of received electromagnetic illumination from
`
`the possible exterior objects” because actual objects may not have been used
`
`to provide the inputs. PO Resp. 20 (citing Ex. 2002 ¶¶ 65, 66, 68–70).
`
`In other words, because “known inputs” used for training in Lemelson
`
`could encompass simulated or real data, Lemelson does not anticipate the
`
`claimed training of the pattern recognition algorithm as recited in claims 10,
`
`16, and 23.
`
`4. Analysis
`
`In light of our determination above that “a pattern recognition
`
`algorithm generated from . . .” (claims 10 and 16) and “generating a pattern
`
`recognition algorithm from . . .” (claim 23) limitations require training using
`
`patterns of waves actually received from possible exterior objects and the
`
`parties’ contentions regarding Lemelson’s disclosure, we determine that
`
`Petitioner has not demonstrated by a preponderance of the evidence that
`
`Lemelson’s reference to training using “known inputs” satisfies the pattern
`
`recognition algorithm “generated from” limitations of claims 10, 16, and 23.
`
`Toyota’s Petition and expert testimony equates training with “known
`
`inputs” to the specified training in claims 10, 16, and 23, but fails to provide
`
`sufficient evidence to support a finding by a preponderance of the evidence
`
`18
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`that “known inputs” refers to training, either expressly or inherently, with
`
`actual data of possible exterior objects.
`
`We credit the testimony of Prof. Koutsougeras that one of ordinary
`
`skill in the art would have interpreted “known inputs” used for training in
`
`Lemelson as open with respect to the type of data—real or simulated—used
`
`to train the neural network. Ex. 2002 ¶¶ 57–64. This understanding is
`
`supported by Toyota’s counsel, who was asked “does the term ‘known
`
`inputs’ [in Lemelson] refer to just real sensor data or is it understood as
`
`both,” and answered that “one of ordinary skill in the art would have
`
`understood [known inputs] as real sensor data, but it is not to the exclusion
`
`of simulated data.” Tr. 27:25–28:6 (emphasis added). In addition, Toyota’s
`
`expert acknowledges that the use of simulated data was a possibility to train
`
`pattern recognition algorithms. See Ex. 2003, Deposition Transcript of Dr.
`
`Papanikolopoulos, 102:5–14 (stating that “in this particular domain, you go
`
`to simulated data, or if you don’t have access to real data, to real images” for
`
`training pattern recognition systems to detect automobiles); see also Ex.
`
`2003, 104:9–23. Thus, the “known inputs” reference in Lemelson is equally
`
`applicable to simulated or real data.
`
`We find Prof. Koutsougeras’s testimony credible that “known inputs”
`
`as referenced in Lemelson could include real or simulated data for training
`
`the neural computer. Ex. 2002 ¶¶ 57–64; see Deposition of Prof.
`
`Koutsougeras, Ex. 1019 at 132:24–138:5, 157:12–159:14, 163:18–164:7.
`
`We disagree with Toyota’s argument that Dr. Koutsougeras’s testimony
`
`should be given little weight because he has limited experience with pattern
`
`19
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`recognition in vehicles. Reply 10–11. To the contrary, AVS’s expert,
`
`Dr. Koutsougeras, testified that his dissertation was in neural networks,
`
`particularly methods of training neural networks. Ex. 1019, Koutsougeras
`
`Deposition 20:19–21:22. We are not persuaded by Toyota’s argument that
`
`experience in training neural networks specifically for vehicle exterior
`
`monitoring application is necessary to support Dr. Koutsougeras’s testimony
`
`regarding an ordinarily skilled artisan’s understanding of the training using
`
`“known inputs” in Lemelson at the time of patenting.
`
`Toyota’s Reply introduces several arguments and supporting
`
`declaration evidence that were not present in the filed Petition. Specifically,
`
`Toyota contends (1) that a person of ordinary skill in the art would have
`
`understood that training a neural network to identify exterior objects or
`
`sources of radiation in Lemelson would have been done with real data and
`
`not with simulated or partial data (Reply 8 (citing Reply Declaration of
`
`Nikolas Papanikolopoulos, Ph.D., Ex. 1020 ¶¶ 10–24)); and (2) the
`
`“generated from” limitation is not a limitation for purposes of patentability
`
`because it is a product-by-process claim that merely specifies the method of
`
`creating an algorithm and does not structurally limit the claim in any way.
`
`Reply 3–4.5
`
`
`
`5 Citing SmithKline Beecham Corp. v. Apotex Corp., 439 F.3d 1312, 1317,
`1319 (Fed. Cir. 2006); In re Warmerdam, 33 F.3d 1354, 1360–61 & n. 6
`(Fed. Cir. 1994); Greenliant Sys., Inc. v. Xicor LLC, 692 F.3d 1261, 1268
`(Fed. Cir. 2012).
`
`20
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`A Reply affords the Petitioner an opportunity to refute arguments and
`
`evidence advanced by the Patent Owner, not an opportunity to cure
`
`deficiencies in its Petition. 37 C.F.R. § 42.23(b); Rules of Practice for Trials
`
`Before the Patent Trial and Appeal Board and Judicial Review of Patent
`
`Trial and Appeal Board Decisions; Final Rule, 77 Fed. Reg. 48,612, 48,620
`
`(Aug. 14, 2012) (“Section 42.23 provides that oppositions and replies must
`
`comply with the content requirements for a motion and that a reply may only
`
`respond to arguments raised in the corresponding opposition. Oppositions
`
`and replies may rely upon appropriate evidence to support the positions
`
`asserted. Reply evidence, however, must be responsive and not merely new
`
`evidence that could have been presented earlier to support the movant’s
`
`motion.”). Replies that raise new issues or belatedly present evidence will
`
`not be considered. 77 Fed. Reg. at 48,767 (stating that “[w]hile replies can
`
`help crystalize issues for decision, a reply that raises a new issue or belatedly
`
`presents evidence will not be considered and may be returned”).
`
`With respect to Toyota’s evidence in support of its argument that a
`
`person of ordinary skill in the art would have interpreted “known inputs” in
`
`Lemelson as referring to actual or real data, Toyota cannot rely belatedly on
`
`this evidence in its Reply and Reply Declaration of Nikolaos
`
`Papaniokolopoulos, PhD (Ex. 1020) to make up for the deficiencies in its
`
`Petition. See, e.g., 37 C.F.R. § 42.23(b) (noting that “[a]ll arguments for the
`
`relief requested in a motion must be made in the motion,” and that a “reply
`
`may only respond to arguments raised in the corresponding opposition or
`
`patent owner response”).
`
`21
`
`
`
`IPR2013-00424
`Patent 5,845,000
`
`
`Even if timely, Petitioner has not shown by a preponderance of the
`
`evidence that one of ordinary skill in the art would have understood that
`
`training a neural network to identify exterior objects or sources of radiation
`
`in Lemelson would have been done with real data and not with simulated or
`
`partial data. Reply 8. Toyota’s belated expert testimony indicates only that
`
`one of ordinary skill in the art may have preferred real over simulated or
`
`partial data for various applications, but does not explain how the reference
`
`to known inputs in Lemelson in context would expressly disclose to one of
`
`ordinary skill in the art such a preference. See Ex. 1020 ¶¶ 10–20. A
`
`preference for real data over simulated or partially simulated data does not
`
`show by a preponderance of the evidence that Lemelson discloses the use of
`
`real data or actual received waves from possible objects to train the neural
`
`computer.
`
`In addition, Petitioner’s untimely citation to portions of Lemelson that
`
`discuss “adaptive operation” and “