UNITED STATES PATENT AND TRADEMARK OFFICE
`
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`____________
`
`WESTERNGECO L.L.C.,
`
`Petitioner,
`
`v.
`
`PGS GEOPHYSICAL AS,
`
`Patent Owner.
`
`____________
`
`Case IPR2015-00309
`
`U.S. Patent No. 6,906,981
`
`____________
`
`
`PETITIONER’S REPLY BRIEF
`
`Pursuant to 37 CFR §§ 42.23 and 42.24(c)(1), WesternGeco L.L.C. (“WG”
`
`
`
`or “Petitioner”) hereby submits the following Reply in support of its Petition for
`
`Inter Partes Review (“IPR”).
`
`
`
`
`
`
`
`
`
`

`
`
`
`
`I.
`
`II.
`
`Case IPR2015-00309
`
`TABLE OF CONTENTS
`
`INTRODUCTION ........................................................................................... 1
`
`DE KOK ANTICIPATES CLAIMS 1, 2, 7, AND 10-21 ............................... 2
`
`A. De Kok Discloses “Enabl[ing] Separate Identification of Seismic
`Events” .................................................................................................. 2
`
`B.
`
`C.
`
`De Kok Teaches “Index[ing]” “the Times of Firing the First and
`Second Source” ..................................................................................... 4
`
`De Kok Discloses That the Second Energy Source is Towed “At a
`Selected Distance From the First Seismic Energy Source” .................. 5
`
`D. De Kok Teaches “Determining Trace to Trace Coherent Components”
` ............................................................................................................... 6
`
`E.
`
`F.
`
`De Kok Teaches “Time Aligning” the Recorded Signals ..................... 7
`
`De Kok Teaches the Use of CMP Gathers to Determine Shot to Shot
`Coherence .............................................................................................. 8
`
`III. BEASLEY AND EDINGTON RENDER OBVIOUS CLAIMS 1-6 AND
`16-22 ................................................................................................................ 9
`
`A. Dr. Lynn Admits Beasley and Edington Disclose All Elements of
`Claim 1 ................................................................................................ 10
`
`B.
`
`POSA Would Have Been Motivated to Combine Beasley and
`Edington .............................................................................................. 10
`
`1.
`
`2.
`
`3.
`
`Beasley Can be Used With Time Delay Encoding ................... 14
`
`Beasley and Edington’s Encoding and Decoding Methods Are
`Compatible ................................................................................ 16
`
`Beasley and Edington Are Not Incompatible ........................... 18
`
`C.
`
`D.
`
`Beasley and Edington Render Obvious Quasi-Random and Random
`Time Delays ........................................................................................ 20
`
`Beasley and Edington Render Obvious Time Delays That Vary in
`Steps of About 100 Milliseconds ........................................................ 21
`
`
`
`
`
`ii
`
`

`
`
`
`Case IPR2015-00309
`
`E.
`
`Beasley and Edington Render Obvious the Use of CMP Gathers to
`Determine Shot to Shot Coherence ..................................................... 22
`
`IV. PGS FAILS TO ESTABLISH A NEXUS SUPPORTING ITS ALLEGED
`SECONDARY CONSIDERATIONS OF NONOBVIOUSNESS ............... 23
`
`
`
`
`
`
`
`
`
`
`
`
`
`iii
`
`

`
`
`I.
`
`INTRODUCTION
`By Patent Owner PGS Geophysical AS’s (“PGS’s”) own admission, U.S.
`
`Case IPR2015-00309
`
`Patent No. 6,906,981 (“the ’981 patent” or “Vaage”), filed in 2002, does nothing
`
`more than use “conventional marine seismic survey” equipment to exploit a natural
`
`phenomenon that was recognized in marine seismic surveys by 1987 and used in
`
`land seismic surveys by at least 1989. Although PGS masquerades this as
`
`invention, it is not.
`
`PGS alleges the “invention” of the ’981 patent was that if multiple marine
`
`sources overlapped—but had a variable time delay between them, i.e., were
`
`“asynchronous”—then their traces could be separated using conventional common-
`
`midpoint (“CMP”) gathers. But, this was well known to those skilled in the art.
`
`For example, De Kok teaches using CMP gathers to separate variably time-delayed
`
`sources for marine seismic surveys. De Kok is therefore anticipatory.
`
`Time delayed sources were also known in land seismic surveys, e.g.,
`
`Edington, and a POSA had a strong motivation for combining that land-based
`
`technique into a marine-based system that used CMP gathers, e.g., Beasley. That
`
`variable time delays could work to separate sources in marine systems was not in
`
`doubt: PGS’s own expert, Dr. Walt Lynn, acknowledges that this principle was
`
`well known in a paper he authored over a decade before the ’981 patent’s earliest
`
`priority date. (Ex. 1021.) Beasley itself also taught time delays between its
`
`
`
`
`
`1
`
`

`
`
`sources, and that those sources could be asynchronous. The ’981 patent’s trivial
`
`Case IPR2015-00309
`
`“improvement” of pulling these known pieces together to yield predictable results
`
`is obvious.
`
`The challenged claims of the ’981 patent are both anticipated by De Kok
`
`(claims 1, 2, 7, and 10-21) and rendered obvious by Beasley in view of Edington
`
`(claims 1-6 and 16-22).
`
`II.
`
`DE KOK ANTICIPATES CLAIMS 1, 2, 7, AND 10-21
`PGS’s Response sets forth six reasons that De Kok does not anticipate
`
`claims 1, 2, 7, and 10–21—all of which are easily dismissed. PGS’s arguments are
`
`either based upon an unduly narrow reading of the challenged claims, a trivial
`
`distinction that is nonetheless taught by De Kok, or both.
`
`A. De Kok Discloses “Enabl[ing] Separate Identification of Seismic
`Events”
`Claim 1 recites “separate identification of seismic events originating from
`
`the first source and . . . from the second source.” (Ex. 1001 at 11:19-22.) De Kok
`
`expressly teaches separating “recorded seismic energy” “into source recordings
`
`responsive to individual seismic sources,” thereby discriminating between the first
`
`and second source. (Ex. 1003 at 2:47-50.) De Kok accomplishes this separation
`
`by manipulating the timing—and hence phases—of the two sources so that the first
`
`source reinforces and the second source cancels out, or vice versa. PGS’s expert,
`
`Dr. Lynn, admitted that this separation is shown, for example, in De Kok’s figure
`
`
`
`
`
`2
`
`

`
`
`5. (Ex. 1022 at 131:25-132:6 (“Q. So Figure 5B shows that De Kok enhances
`
`Case IPR2015-00309
`
`source one and suppresses source two? A. Correct. Q. And Figure 5C shows that
`
`De Kok enhances source two and suppresses source one? A. Correct.”); see also
`
`id. at 136:17-137:2.) The recited “separate identification” is taught in De Kok.
`
`PGS attempts to distinguish De Kok as summing signals from a given
`
`source, rather than extracting individual shots from each source firing. (POR at
`
`23.) However, this improperly reads the word “individual” into claim 1. Claim 1
`
`does not require the separate identification of “individual” seismic events (i.e.,
`
`each source firing), only a discrimination between the first and second sources.
`
`(Ex. 1001 at 11:19-22.)
`
`Moreover, the ’981 patent teaches, as PGS acknowledges, using multiple
`
`firings of each source, i.e., CMP trace gathers, to separate signals corresponding to
`
`each source, just like De Kok. (POR at 22; Ex. 1001 at 8:43-56.) The ’981 patent
`
`explains that these grouped recordings, and not individual shots, are used to
`
`separate the two sources’ signals. (Ex. 1001 at 8:51-56 (“CMP gathers with
`
`respect to source A will have a high coherence for events which correspond to the
`
`firing of source A” and “will have a very low coherence for source B originating
`
`signals.”).) PGS’s unduly narrow claim construction reads the ’981 patent’s
`
`teaching out of its own claims.
`
`De Kok takes “recorded seismic energy” and separates it “into source
`
`
`
`
`
`3
`
`

`
`
`recordings responsive to individual seismic sources,” which is all claim 1 requires.
`
`Case IPR2015-00309
`
`(Ex. 1003 at 2:47-50.) In fact, De Kok teaches the same CMP approach as the
`
`’981 patent as part of this separation. (Ex. 1003 at 4:50-55 (“The separation of
`
`individual source contributions . . . is achieved during processing, preferably in the
`
`common mid-point (CMP) domain…”), 6:41-48 (“[T]he CMP gather may be the
`
`preferred domain to execute the source discrimination.”).) Not only does De Kok
`
`teach this limitation of claim 1, it teaches the same preferred embodiment as the
`
`’981 patent.
`
`B. De Kok Teaches “Index[ing]” “the Times of Firing the First and
`Second Source”
`Both Dr. Ikelle and Dr. Lynn explained that the seismic data for each trace,
`
`including firing times, is recorded in a trace header. (Ex. 2002 at 182:15-19; Ex.
`
`1022 at 82:7-13.) Each trace header “[has] associated with it the firing time, the
`
`location, and . . . other parameters” of the survey. (Ex. 1022 at 82:14-22.)
`
`Typically, these trace headers are recorded and saved in the SEG Y format. (Ex.
`
`2002 at 182:15-19; Ex. 1022 at 82:7-13; see also Ex. 1003 at 4:17-20 (noting that
`
`“SEG standard format information” includes “time of day identifiers”).)
`
`PGS argues that there is no explicit discussion of the industry standard in De
`
`Kok, and therefore no “indexing” of the source firings. This argument borders on
`
`specious. De Kok inherently “indexes” or records the firing times of the first and
`
`second sources; otherwise, De Kok would not be able to process the collected
`
`
`
`
`
`4
`
`

`
`
`seismic data. (Ex. 2002 at 182:15-19; Ex. 1022 at 135:17-.137:2.) De Kok does
`
`Case IPR2015-00309
`
`not specify how it records firing times, but it need not, as recording the firing times
`
`was necessarily embodied in industry standards long before De Kok was
`
`published. (Ex. 2002 at 182:15-19; Ex. 1024.)
`
`Moreover, the separation technique taught in De Kok requires that the firing
`
`times be indexed. De Kok teaches using time delays that are “determined or
`
`chosen for the acquisition program” (Ex. 1003 at 6:23-30), which means that the
`
`time delays must be known and predetermined and saved in the programming.
`
`(Ex. 2002 at 159:18-21.) While PGS argues that “select[ing]” or “know[ing]” the
`
`time delay does not mean that De Kok’s time delays are “indexed” (POR at 24-25),
`
`“indexed” simply means that “you have to save . . . your time delay,” i.e., that the
`
`times are recorded. (Ex. 2002, 182:13-15.) And, as Dr. Lynn explained, De Kok
`
`“has to control [the] timings in the field” in order for the summations and source
`
`separations to work. (Ex. 1022 at 135:17-137:2.) Thus, De Kok teaches
`
`“index[ing]” “the times of firing the first and second source” as claimed.
`
`C. De Kok Discloses That the Second Energy Source is Towed “At a
`Selected Distance From the First Seismic Energy Source”
`De Kok teaches two sources that are separated by a selected distance. (Ex.
`
`1003 at Fig. 4 (showing, e.g., sources 203 and 205), id. at 5:36-40 (discussing “the
`
`configuration of [] the two sources,” that some sources are “relatively close to each
`
`other” and that others are spaced “preceding streamers” and “at the back of
`
`
`
`
`
`5
`
`

`
`
`streamers”).) PGS argues that this is not a “selected” distance as claimed because
`
`Case IPR2015-00309
`
`no specific dimensions are recited. (POR at 29-30.) But the ’981 patent does not
`
`claim selecting a specific distance between the sources, and no specific dimensions
`
`are taught anywhere within the ’981 patent. Dr. Lynn admitted that “there is no
`
`limitation in Vaage about selecting specific locations for [seismic sources]” and
`
`that the seismic sources could be within any range of each other as long as that
`
`range is “reasonable” and produces “useful information.” (Ex. 1022 at 112:4-16.)
`
`Thus, there is no distinction between De Kok and the ’981 patent in this regard.
`
`D. De Kok Teaches “Determining Trace to Trace Coherent
`Components”
`The ’981 patent explains that determining trace to trace coherence
`
`“separat[es] the components of the signals which are caused by [one source] from
`
`the random noise.” (Ex. 1001 at 8:13-18.) As Dr. Lynn explained, random noise
`
`is removed with CMPs. (Ex. 1022 at 57:7-12 (“Q. You suppress the random noise
`
`by doing a CMP stack? A. Correct. Q. And that’s one of the main purposes of
`
`doing a CMP stack? A. Yes.”); id. at 76:16-19 (“Q. So with the CMP gather
`
`example, could you visually identify the trace-to-trace coherency? A. Yeah.”) De
`
`Kok’s separation is achieved during processing “preferably in the common mid-
`
`point (CMP) domain,” which results in “coherent data,” thus anticipating the
`
`challenged claims. (Ex. 1003 at 4:47-55; Ex. 2002 at 281:20-23.)
`
`More generally, removing random noise such as that from another source or
`
`
`
`
`
`6
`
`

`
`
`a ship propeller is an inherent part of seismic surveying, has been done since the
`
`Case IPR2015-00309
`
`1960s, and is necessarily done in De Kok. (Ex. 2002 at 202:22-25.) Particularly
`
`where the goal is to “separat[e] individual source contributions into source
`
`records” (as in De Kok), random noise must be removed during processing. (Ex.
`
`1003 at 4:47-55.) Dr. Lynn likewise acknowledged that “[a]ny summation of
`
`traces,” such as that done in De Kok, “is going to suppress the level of random
`
`noise.” (Ex. 1022 at 133:8-11.) Thus, both experts agree that De Kok’s CMP
`
`gathers necessarily remove random noise, i.e., “determin[e] trace to trace . . .
`
`coherent components” as recited in claims 12-15 and 17-20.
`
`E. De Kok Teaches “Time Aligning” the Recorded Signals
`De Kok’s encoding method uses a variety of positive and negative time
`
`delays between the activation of the seismic sources and then uses “polarity
`
`decoding to enhance and separate energy” from one seismic source shot series
`
`from a second or third seismic source shot series.1 (Ex. 1003 at 6:41-65.) This
`
`process is the time aligning step of the challenged claims.
`
`De Kok further discloses separating sources “in the common mid-point
`
`domain” which, as known to POSA, necessarily includes normal move-out
`
`
`1 As De Kok’s polarities are a direct function of the programmed time delays, De
`
`Kok’s polarity decoding includes time aligning. (Ex. 2002 at 18:3-7; 187:19-
`
`189:5.)
`
`
`
`
`
`7
`
`

`
`
`correction (“NMO”) to time align seismic signals prior to summing them. (Ex.
`
`Case IPR2015-00309
`
`1022 at 46:9-14.) During his deposition, Dr. Lynn explained that “any time you
`
`would do a CMP stack,” a NMO correction would be done prior to generating the
`
`CMP stack. (Ex. 1022 at 44:21-24; 45:5-8; 68:24-69:3 (“Q. And that gets back to
`
`any time you would do a CMP stack, you would do an NMO time shift first? A.
`
`Correct.”).) Thus, by disclosing separating sources in the CMP domain, De Kok
`
`necessarily discloses “time aligning” the recorded signals as recited in claims 14,
`
`15, 19, and 20, either directly or through inherency.
`
`F. De Kok Teaches the Use of CMP Gathers to Determine Shot to
`Shot Coherence
`“Shot to shot coherence” refers to differentiating signals from two sources.
`
`(Ex. 1001 at 8:18-25 (discussing “generat[ing] coherent traces from shot to shot” to
`
`“represent seismic signals resulting only from source A” and that “noise from
`
`source B . . . will be substantially absent”).) Dr. Lynn confirmed that shot to shot
`
`coherence involves emphasizing signals from one source while suppressing those
`
`from the second source. (Ex. 1022 at 70:18-71:6.)
`
`De Kok expressly discloses this “separation of individual source
`
`contributions.” (Ex. 1003 at 4:50-55.) Moreover, Dr. Lynn confirmed that this
`
`separation is taught in De Kok. (Ex. 1022 at 136:22-137:2 (“Q. In theory. But
`
`that’s what De Kok is at least teaching, is that if you do this method, you will be
`
`able to isolate source one or isolate source two? A. Yes.”); id. at 131:25-132:6
`
`
`
`
`
`8
`
`

`
`
`(“Q. So Figure 5B shows that De Kok enhances source one and suppresses source
`
`Case IPR2015-00309
`
`two? A. Correct. Q. And Figure 5C shows that De Kok enhances source two and
`
`suppresses source one? A. Correct.”)
`
`And, despite PGS’s protestation, De Kok uses the same CMP gathers as
`
`Vaage in determining this coherency. (Compare Ex. 1022 at 75:2-4 (“Q. The
`
`CMP gather in Vaage is what lets you determine the shot-to-shot coherency? A.
`
`That’s how he approaches it, yes.”) with Ex. 1003 at 4:50-55 (teaching that the
`
`“separation of individual source contributions” is preferably done “in the common
`
`mid-point (CMP) domain”) (emphasis added) Nothing else is required to
`
`anticipate the challenged claims.
`
`III. BEASLEY AND EDINGTON RENDER OBVIOUS CLAIMS 1-6 AND
`16-22
`Beasley and Edington form a tight combination. Beasley discloses using
`
`multiple marine sources with a programmable delay between them and using CMP
`
`gathers as part of separating the two sources. Edington discloses using variable
`
`time delays to separate two land sources. These patents form a tight combination
`
`because POSA looks to land-based seismic survey techniques for marine-based
`
`applications and has done so over the history of this industry. (See e.g., Ex. 1002
`
`at ¶ 29; Ex. 1009 at 161; Ex. 1010 at 477-78.) For PGS to argue otherwise,
`
`effectively puts blinders on POSA. Even Dr. Lynn conceded that “these
`
`techniques for suppressing an asynchronous second source” apply to both land and
`
`
`
`
`
`9
`
`

`
`
`marine and that the “natural phenomenon [] that if you have an asynchronous
`
`Case IPR2015-00309
`
`second source, it will be suppressed in a CMP stack” would “apply both for land
`
`and marine.” (Ex. 1022 at 94:5-14, 95:16-23.)
`
`A. Dr. Lynn Admits Beasley and Edington Disclose All Elements of
`Claim 1
`This case is about motivation to combine only. Dr. Lynn admitted during
`
`
`
`his deposition that Beasley by itself discloses all but one element of claim 1—“a
`
`time interval between firing the first source and the second source varied between
`
`successive ones of the firing sequences.” (Ex. 1022 at 162:6-167:25 (walking
`
`through elements of claim 1).)2 Dr. Lynn admitted, however, that Edington teaches
`
`this “missing” element. (Ex. 1022 at 143:20-25 (“Q. The time delay between the
`
`two sources varies from shot to shot. A. Okay, I’m with you. Q. That’s something
`
`that Edington teaches? A. Right.”) Thus, this case is merely about motivation, for
`
`which there is plenty, and this Board need look no further to cancel the challenged
`
`claims.
`
`B.
`
`POSA Would Have Been Motivated to Combine Beasley and
`Edington
`PGS argues that POSA would not have been motivated to take the variable
`
`time-delay of Edington and attempt to use it with Beasley because the prior related
`
`
`2 As discussed below, Beasley also teaches the “missing” asynchronicity in its
`
`priority application that is incorporated by reference. (Ex. 1023 at p. 25, cl. 2.)
`
`
`
`
`
`10
`
`

`
`
`to land surveys and the latter to marine. (POR at 38.) In doing so, PGS ignores the
`
`Case IPR2015-00309
`
`many well-recognized similarities between land and marine seismic acquisition
`
`techniques—including the use of impulsive sources, source encoding, NMO, and
`
`CMP gathers—and also ignores that many in the industry borrowed land
`
`technology for use in marine surveys and vice versa. (See e.g., Ex. 1002 at ¶¶ 29,
`
`30, 34; Ex. 1009 at 161; Ex. 1010 at 477-78; Ex. 1021 at 1502 (“Although our data
`
`examples are from the marine environment, the conclusions are also applicable to
`
`problems of interference in land surveys.”); Ex. 1022 at 93:9-12; 94:23-95:7)
`
`Notably, the marine patents at issue here, De Kok, Beasley, and even the ’981
`
`patent, all cite to Edington (and other land references) as relevant prior art.
`
`Multiple inventors in the relevant time frame considered Edington a close enough
`
`technology to disclose it under the applicable duty of candor.3 Not only would
`
`POSA involved in marine simultaneous surveys look to land references in the
`
`abstract, the record shows three inventors who actually did.
`
`Moreover, the marine prior art’s understanding of Edington is generalized to
`
`any delay-separation and not limited to land contexts. (See, e.g., Ex. 1003 at 2:6-
`
`
`3 Even the marine Herkenhoff patent, on which PGS and Dr. Lynn rely, cites
`
`Edington generally as “a method of time delay source coding.” (Ex. 2015 at 1:47-
`
`50.)
`
`
`
`
`
`11
`
`

`
`
`12). Edington’s teaching is described in the prior art in almost identical language
`
`Case IPR2015-00309
`
`as Dr. Lynn’s description of the ’981 patent:
`
`“A method disclosed by U.S. Pat.
`No. 4,953,657 to Edington
`discloses a suite of time delay
`differences between sources. To
`enhance the signal from a
`particular source, the
`corresponding signals are aligned
`and stacked. The contributions
`from the other source(s) are not
`aligned and do not stack to full
`strength.” (Ex. 1003 at 5:-101
`
`“Q. And in general, the Vaage patent uses a
`series of time delays between the sources in
`firing sequences? A. In general, that’s a
`correct statement…Q. To enhance the signal
`from a particular source, the Vaage patent
`teaches that the corresponding signals are
`aligned and stacked like you show in that
`figure on the left? A. Correct. Q. And if
`you do that, the Vaage patent teaches that the
`contributions from the other source are not
`aligned so they will not stack to full strength?
`A. Correct.” (Ex. 1022 at 67:5-67:20)
`
`The time delays in Edington are used for the same purpose as in the claimed
`
`’981 invention, and POSA would have expected success importing Edington’s time
`
`delays into the marine environment because it was known from a 1987 paper by
`
`Dr. Lynn that the phenomenon of time-delay separation applied as equally in the
`
`marine context as in the land context. (Ex. 1021; Ex. 1022 at 23:14-22, 60:6-10
`
`(“Q. So when the two sources were asynchronous with each other, you would be
`
`able to make the second source incoherent in the CMP gather? A. Correct.”); id.
`
`at 62 (“Q. And what you discovered in 1987 was the phenomenon that using
`
`regular equipment and using regular CMP techniques, you could suppress the
`
`
`
`
`
`12
`
`

`
`
`second source if it was asynchronous to the first? A. That’s correct.”).) And
`
`Case IPR2015-00309
`
`Beasley itself taught POSA that its system could be used with such asynchronous
`
`sources—the priority 08/829,485 application that was explicitly incorporated by
`
`reference into Beasley (Ex. 1004 at 1:5-14) discloses two timed intervals for two
`
`sources and “time-shifting the second [source] to occur asynchronously relatively
`
`to the . . . first.” (Ex. 1023 at p. 25, cl. 2.)4
`
`The record is clear that POSA would have had a reasonable expectation of
`
`success in using Edington’s asynchronous timings with Beasley’s time-delay
`
`sources (and asynchronous timings) to exploit the same phenomenon taught by
`
`Edington for land seismic, and observed by Dr. Lynn in 1987 for marine seismic,
`
`to achieve the shared goal of Edington and Beasley of separating the two seismic
`
`sources using CMPs.
`
`4 In his deposition, Dr. Lynn confirmed that he based his opinion on the inability to
`
`use Beasley with asynchronous sources. (Ex. 1022 at 172:9-17.) Not only is this
`
`predicate wrong in light of Beasley’s explicit disclosure of asynchronous sources
`
`discussed above, but Dr, Lynn also opined that it inherently anticipates the ’981
`
`invention. (Id. at 114:22-116:4 (“So if the sources are dithered or
`
`asynchronous . . . in a CMP domain, the energy—reflected energy coming back
`
`from source A will be coherent and the energy being reflected back from source B
`
`will be incoherent.”).)
`
`
`
`
`
`13
`
`

`
`
`
`Case IPR2015-00309
`
`1.
`Beasley Can be Used With Time Delay Encoding
`Beasley’s figure 7 discloses “decoders” 31 and 33 for separating the record
`
`signal into two sources. (Ex. 1004, Fig. 7.) Beasley’s text also teaches that “any
`
`desired type of coding” may be employed. (Ex. 1004 at 7:54-56.) The Edington
`
`patent, titled “Time Delay Source Coding,” would seem to fit this bill. (Ex. 1006.)
`
`More broadly, Beasley teaches it can be used with sources “that can be
`
`discriminated from each other due to some identifying characteristic, parameter,
`
`signature, or feature.” (Id. at 10:4-10.) Edington’s time-delay encoding fits
`
`squarely within this framework and Edington’s “determinable time delay” (Ex.
`
`1006 at 2:4-8) provides the very identifier sought by Beasley. POSA would
`
`therefore understand that Beasley could be used with time delay encoding, as
`
`taught by Edington. (Ex. 1002 at ¶ 241; Ex. 1004 at 7:55-56.)
`
`PGS objected to the combination of Beasley and Edington by focusing on a
`
`singular phrase–“any desired type of coding”–and arguing that it refers to source
`
`signature encoding only. (POR at 44-47.) Although Dr. Lynn stated that this
`
`sentence in Beasley excludes time delay encoding, he nevertheless admitted that
`
`“time delay encoding” as disclosed in the ’981 patent is “a type of source signature
`
`encoding for the marine environment” and that Edington’s time delay source
`
`coding is “a type of source signature encoding.” (Ex. 1022 at 18:7-14; 185:2-3;
`
`
`
`
`
`14
`
`

`
`
`148:18-23.) Even if this phrase is construed as PGS suggests, which is incorrect,
`
`Case IPR2015-00309
`
`the rest of Beasley’s teachings cannot be ignored.
`
`Importantly, POSA would have looked to Edington’s time delay encoding as
`
`a type of source signature encoding to use with Beasley because Beasley teaches
`
`time-delayed source firing. Although PGS argues that Beasley teaches only
`
`encoding with concurrent source firing (i.e., with no time delay,) that is
`
`demonstrably false. (POR at 42-44.) Figure 7 of Beasley shows coding/decoding
`
`with a time delay between the two sources. (Ex. 1004 at Fig. 7, 7:64-8:27
`
`(“[S]ource SL is first activated at time T0 . . . source ST is activated at time t0 after
`
`a time shift through delay line 29 . . . If the recorded reflected acoustic wavefields
`
`were encoded, of course optional decoders 31 and 33 . . . would be inserted….”);
`
`see also Ex. 1022 at 102:13-103:10 (discussing figures 7 and 8) ; id. at 180:13-21;
`
`id. at 191:24-192:4 (“Q. In the timing diagram shown in Figure 8, it looks like the
`
`two sources are going to overlap in the recordings. Is that fair? A. It looks like
`
`it.”).) More generally, Beasley teaches coded sources that are fired
`
`“simultaneously or nearly simultaneous,” just like the sources in Edington or in
`
`the ’981 patent itself. (Ex. 1004 at 8:46-47.) And perhaps most importantly,
`
`Beasley’s incorporated-by-reference priority document expressly teaches
`
`asynchronous time-delayed source firing. (Ex. 1023 at p. 25, cl. 2.)
`
`
`
`
`
`15
`
`

`
`
`
`Case IPR2015-00309
`
`PGS’s reliance on Beasley’s single use of the word “concurrently” does not
`
`change anything. As an initial matter, PGS’s reference to “exactly concurrently” is
`
`found nowhere within Beasley—it is from a 2007 reference that is irrelevant to this
`
`proceeding. Additionally, PGS overlooks the fact that “concurrently” has a
`
`specific meaning in the marine surveying context, instead basing its argument on
`
`the colloquial meaning of “concurrently.” “Concurrently,” however, like
`
`“simultaneous,” has a specific meaning in seismic surveying, and encompasses
`
`timing that is “near concurrently” or “near simultaneous.” (Ex. 2002 at 218:8 –
`
`219:2; 220:23-221:4.) Thus, POSA would have understood that Edington’s time
`
`delay encoding could be used to activate Beasley’s multiple sources
`
`“concurrently,” and that coding could be used with time-delayed sources as well.
`
`See, e.g., Phillips v. AWH Corp., 415 F.3d 1303, 1313 (Fed. Cir. 2005) (“[T]he
`
`ordinary and customary meaning of a claim term is the meaning that the term
`
`would have to a person of ordinary skill in the art.”) (emphasis added).
`
`2.
`
`Beasley and Edington’s Encoding and Decoding Methods
`Are Compatible
`PGS argues that WesternGeco somehow failed to consider whether
`
`Edington’s encoding and decoding methods would be compatible with Beasley.
`
`This is simply untrue. Rather, Dr. Ikelle opined that “the use of time delays is
`
`independent of the environment of the survey” and POSA would be motivated to
`
`combine Beasley and Edington because they “both address multishooting,
`
`
`
`
`
`16
`
`

`
`
`encoding, and decoding.” (Ex. 1002 at ¶¶ 241-43.)
`
`Case IPR2015-00309
`
`Both Edington and Beasley time-align traces for their decoding. (Ex. 1006
`
`at 5:60-64 (“The signals shown in FIG. 4 are then time shifted as shown in FIG. 5
`
`so that the signals 38 are aligned…”); Ex. 1004 at 4:16-40 (“To separate the
`
`sources’ data, the record is updated with one source’s geometry information (e.g.,
`
`x, y location coordinates and time of day identifier) . . . optionally sorted to order
`
`by known common mid-point (CMP) sorting methods . . . then re-done with the
`
`attachment of the other source’s geometry.”); Ex. 1022 at 68:24-69:3; 56:7-17;
`
`144:2-18.) PGS’s argument that Edington and Beasley are an encoding/decoding
`
`mismatch is a red herring and conspicuously avoids the real question—whether it
`
`would be within the skill level of POSA to use Edington’s random time variations
`
`with Beasley’s seismic surveying system to achieve the result predicted by the
`
`1987 Lynn paper. This answer is yes. (Ex. 1002 at ¶¶ 239-44.)
`
`PGS’s arguments ignore the fact that Beasley itself discloses the use of time
`
`delays (Ex. 1004 at 8:4-11; Figs. 7 (element 29), 8 (timing diagram))—and the
`
`incorporated priority application explicitly discloses that those delays are
`
`asynchronous. (Ex. 1023 at p. 25, cl. 2.) Dr. Lynn also acknowledged that CMP
`
`stacking can be used with asynchronous sources to render one source coherent and
`
`the other incoherent. (Ex. 1022 at 114:22-116:4.) CMP stacking allows for the
`
`separation of two asynchronous sources after they have been time aligned using
`
`
`
`
`
`17
`
`

`
`
`NMO (Ex. 1022 at 60:6-10; Ex. 1021 at 1506-07), and both Dr. Ikelle and Dr.
`
`Case IPR2015-00309
`
`Lynn agree that CMP gathers can be used in either the marine or land context.
`
`(Ex. 1022 at 94:23-95:4; Ex. 1002 at ¶ 30.)
`
`And in any event, Edington and Beasley both decode by summing traces that
`
`correspond to the same seismic event. As Dr. Lynn explained, Edington can do
`
`this in the land context by repeating the same source-receiver locations. (Ex. 1022
`
`at 145:12-20 (“Q. And because the source and the receiver are in the same
`
`location, you can assume that the traces are reflecting the same seismic event? A.
`
`Correct.”)) But because “[i]n the marine context you generally don’t repeat the
`
`same source and receiver locations” (id. at 145:21-24), instead “you can use a
`
`CMP stack to enhance the signal” (id. at 146:23-147:2; see also id. at 56:7-17
`
`(explaining that “CMP traces . . . essentially relate to the same seismic event”).)
`
`This CMP stacking is taught in Beasley, and one of skill in the art would have
`
`expected it to separate Edington’s asynchronous sources in light of the known
`
`phenomenon discussed in, e.g., Dr. Lynn’s 1987 paper.
`
`3.
`Beasley and Edington Are Not Incompatible
`PGS’s argument that Beasley and Edington are incompatible is based on
`
`PGS’s creation of an unsolvable smearing problem that simply does not exist in
`
`practice. Specifically, PGS argues that using Edington’s decoding technique with
`
`Beasley’s encoding technique would result in the loss of spatial resolution, which
`
`
`
`
`
`18
`
`

`
`
`PGS alleges Beasley seeks to avoid. (POR at 48.) As an initial matter, and as
`
`Case IPR2015-00309
`
`discussed above, there is no reason that POSA would have to utilize Edington’s
`
`specific decoding method in the proposed combination with Beasley because
`
`POSA would understand that Edington’s decoding method is analogous to CMP
`
`stacking, which was commonly used in marine seismic surveys and explicitly
`
`disclosed in Beasley. (Ex. 1022 at 146:9-147:2.) As discussed above, POSA
`
`would have known that Beasley’s CMP teaching would separate Edington’s
`
`asynchronous sources as evidenced by, e.g., Dr. Lynn’s 1987 paper showing such a
`
`result. (Ex. 1022 at 62:19-24 (“And what you discovered in 1987 was the
`
`phenomenon that using regular equipment and using regular CMP techniques, you
`
`could suppress the second source if it was asynchronous to the first? A. That’s
`
`correct.”).)
`
`PGS alleges that spatial smearing occurs when shot records or traces are
`
`considered as a single s