`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` M. TRIANTAFYLLOU
` UNITED STATES PATENT AND TRADEMARK OFFICE
` BEFORE THE PATENT TRIAL AND APPEAL BOARD
` ____________________________
` PETROLEUM GEO-SERVICES INC.
` Petitioner
` v.
` WESTERNGECO LLC
` Patent Owner
` _____________________________
`Case No. IPR2014-01475, -01476, -01477, -91478
` Patent No. 7,162,520 B2
` Patent No. 7,162,967 B2
` Patent No. 7,080,607
` _____________________________
`
` DEPOSITION OF DR. MICHAEL TRIANTAFYLLOU
` Volume 2
` Alexandria, Virginia
` August 28, 2015
`
` Reported by: Mary Ann Payonk
` Job No. 96926
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`PGS Exhibit 1118, pg. 1
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 357
`
`Page 358
`
` M. TRIANTAFYLLOU
`
` August 28, 2015
` 8:37 a.m.
`
` Deposition of DR. MICHAEL TRIANTAFYLLOU,
`Ph.D., Volume 2, held at the offices of Oblon,
`McClelland, Maier & Neustadt, 1940 Duke Street,
`Alexandria, Virginia, pursuant to Notice before
`Mary Ann Payonk, Nationally Certified Realtime
`Reporter and Notary Public of the District of
`Columbia, Commonwealth of Virginia, States of
`Maryland and New York, CA-CSR No. 13431.
`
`Page 359
`
` M. TRIANTAFYLLOU
`MICHAEL TRIANTAFYLLOU,
` recalled as a witness, having been duly
` sworn, was admonished of his former
` oath, examined and testified as follows:
` EXAMINATION (Cont'd.)
`BY MS. BERNIKER:
` Q. Doctor, you understand that you're
`still under oath from yesterday?
` A. Thank you.
` MR. KIKLIS: Can we just make a
` note for the record that we have a new
` attendee, Mr. Suarez.
` MS. BERNIKER: Mr. Suarez is here
` today and I think Mr. Swafford is not
` here today, both from Williams &
` Connolly.
`BY MS. BERNIKER:
` Q. All right. Doctor, at any point
`since we began the deposition, have you
`communicated with the attorneys for WesternGeco
`about the deposition?
` A. No.
` Q. Okay. I would like to talk for a
`minute about hydrophone noise. Can you explain
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
` M. TRIANTAFYLLOU
`APPEARANCES:
`ON BEHALF OF PETITIONER:
` THOMAS FLETCHER, ESQUIRE
` JESSAMYN BERNIKER, ESQUIRE
` CHRISTOPHER SUAREZ, ESQUIRE
` WILLIAMS & CONNOLLY
` 725 12th Street, N.W.
` Washington, D.C. 20005
`
`ON BEHALF OF PATENT OWNER:
` MICHAEL KIKLIS, ESQUIRE
` CHRISTOPHER RICCIUTI, ESQUIRE
` KATHERINE CAPPAERT, ESQUIRE
` OBLON McCLELLAND MAIER & NEUSTADT
` 1940 Duke Street
` Alexandria, VA 22314
`
`ALSO PRESENT:
` Kevin Hart, Petroleum Geo-Services,
` Inc. (By phone)
`
`Page 360
`
` M. TRIANTAFYLLOU
`to me why hydrophone noise is a concern in
`marine seismic surveying?
` A. Okay. So let's start with the
`principle of hydrophones. The only way of
`communicating under water is through sound,
`because the ocean is opaque to electromagnetic
`waves. So, for example, there's no GPS below
`let's say 5 centimeters below the ocean
`surface, which causes a great problem to any --
`underwater, you have to use acoustics, there's
`no other means of communication at a distance.
` On the other hand, the acoustical
`waves, the water in general is an excellent
`conductor of acoustic waves. They can travel
`fast. They can travel far if the frequencies
`are low. They can travel short distances at
`high frequency, providing great accuracy, and
`so on and so forth.
` So what is important in acoustical
`waves is the frequency -- if you take whatever
`sound waves are analyzed, there will be
`components of high frequency, low frequency and
`so on and so forth. There will be a range of
`frequencies. So when you do seismic
`
`1
`
`2345
`
`6
`
`78
`
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`2
`
`PGS Exhibit 1118, pg. 2
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 361
`
` M. TRIANTAFYLLOU
`exploration, you need to travel far, signals
`have to penetrate the bottom of the ocean, be
`reflected by the layers underneath. So you
`work in a certain frequency range which allows
`you to do this.
` At the same time, for short
`communications, you want higher accuracy, so
`you go to higher frequencies. So you're trying
`immediately to divide your hydrodynamic
`acoustical signals by the purpose they have so
`you can use them for exploring the underground
`deposits with certain lower frequencies, and
`use higher frequencies to find positions and
`the like.
` But then you have sources of noise.
`And let's restrict ourselves to hydrodynamic
`noise as your question addressed. So the -- if
`you have something which is perfectly
`streamlined, streamlined means that it has a
`shape like we envision a fish or an airplane or
`a torpedo. It produces relatively small
`amounts of noise, except where there are
`protrusions, there are sharp cutoffs and the
`like. You put them at an angle of attack, you
`
`Page 363
`
` M. TRIANTAFYLLOU
`body, down to what's called the Kolmogorov
`scale, which is submillimeter, the size of
`those eddies in the turbulent flow.
` And so when you take a -- a power
`spectrum of that, when you find the frequency
`analysis, you have a very wide frequency range,
`and that noise interferes with the signals. In
`other words, when you measure what you are
`interested in, whether it is position of a body
`with high frequencies, so the short distance
`signals, or the long distance, lower frequency
`signals, these have a dynamic noise because it
`comes from turbulence, which is a chaotic
`phenomenon with many, many time scales,
`produces a wide band noise.
` So, for example, a foil that will
`stall, let's take the DigiFIN or the DigiBIRD.
`You create an angle of attack which is
`25 degrees because you want a big force. It
`will stall and then it will create -- if it's
`half a meter, it will create eddies from half a
`meter down to submillimeter, and all these are
`going to create a lot of hydrodynamic noise.
`And this noise will interfere with the accuracy
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 362
`
` M. TRIANTAFYLLOU
`turn them, and then the noise increases.
` So, for example, a streamer, if you
`tow it perfectly aligned, the only sources of
`noise would be any -- anything that is attached
`on it. That would be the primary source of
`noise. There is some noise anyway because of
`the boundary layer, which is characteristic of
`anything that is in the flow and moves. It's a
`layer adjacent to the body called the boundary
`layer.
` But the main source of sound comes
`from what is called flow separation, when the
`flow, instead of hiding the body, splits and
`generates eddies. That's a major topic for
`many people, including the hydrodynamic noise
`they cause.
` So what happens when that separation
`happens, you create turbulent flow. Turbulence
`is a vague term, but in essence, it means you
`have hundreds of scales, meaning small, large,
`depending on the size of the object that makes
`them, but they range in scale from, let's say
`if you have a body which is half a meter, they
`will go from half a meter, the size of the
`
`Page 364
`
` M. TRIANTAFYLLOU
`of anything that you do. So as a result, it is
`imperative that whatever works around
`hydrophones be of limited generation of such
`hydrodynamic noise.
` Q. I want to make sure I understood your
`answer. I'm obviously not going to go through
`every sentence you said, but is the
`hydrodynamic noise -- actually, let me take a
`step back.
` Is the hydrophone noise -- that's
`important during data acquisition; right?
` A. And it can be also during positioning
`of -- determining the position of the
`streamers, positioning the streamers, trying to
`determine the position of the streamers.
` Q. So it can be important when you're
`trying to determine the position of the
`streamers and it can be important when you're
`acquiring seismic data; is that right?
` A. Yes.
` Q. Okay. And why in your view is it
`important when you're determining the position
`of the streamers?
` A. Because again, if the frequencies
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`3
`
`PGS Exhibit 1118, pg. 3
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 365
`
` M. TRIANTAFYLLOU
`interfere with the frequencies of the
`positioning devices, then that increases the
`noise at the same frequencies. In other words,
`if I have a signal which consists of two
`frequencies different, so if the noise is high
`frequency and my signal is low frequency, I use
`a Kalman filter and just shoot it out. Nothing
`is left. But if the noise is at the same
`frequency as my signal, then it's very
`difficult or impossible sometimes to
`distinguish. Often it's impossible to
`distinguish the two, unless there's some
`special trick you can do to distinguish the two
`signals. So it poisons basically the signal,
`the measurements.
` Q. How long has it been known that
`hydrodynamic noise -- I'm sorry, that
`hydrodynamic noise is something that you'd try
`to avoid as much as possible near the
`hydrophones?
` A. It is generally known that you should
`not have noise around any acoustical device.
`The least amount of noise around acoustical
`devices. So it's not limited to streamers,
`
`Page 367
`
` M. TRIANTAFYLLOU
`to position streamer positioning devices along
`the length of a streamer, you'd want them to be
`small?
` MR. KIKLIS: Objection, form,
` misstates.
` A. It's not so much small that it's
`the -- the important item is a combination of
`things. It's the shape and the size, of
`course, because the noise will increase with
`size. But it will increase more -- much more
`with shape than with size. And also, those go
`hand in hand because when you design something
`which is big, it has to be strong, and so it
`further accentuates the problems.
` And also, the operating envelope,
`meaning if you put a small foil, small -- too
`small, so let's say, when it's sitting there,
`will make practically no noise, but the minute
`you start tilting it, it will start producing
`noise. So if you're continuously operating at
`high envelopes because it's so small that in
`order to -- to avoid it stalling, they will
`produce a lot of noise. So it's a balance of
`things. It's not a recipe that smaller,
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 366
`
` M. TRIANTAFYLLOU
`side scan sonars, which we use a lot. You go
`to extreme measures to make sure that the noise
`is not there.
` Q. Was that known as far back as the
`'80s?
` MR. KIKLIS: Objection, scope.
` A. It is a consideration for any
`acoustical device that you should not have
`noise of the same frequencies around it.
`BY MS. BERNIKER:
` Q. Yeah. And I'm just trying to get a
`sense for how long people have known that. Can
`you give me an estimate?
` A. The fact that hydrodynamic noise is
`important dates back to the '50s in some
`applications, you know.
` Q. And --
` A. And certainly the military know it
`for much longer than that.
` Q. Yeah, okay. Is that one of the
`reasons that -- I think you said in your expert
`report, but apologies if I'm not quite
`construing that right -- is that one of the
`reasons that you would want, if you were going
`
`Page 368
`
` M. TRIANTAFYLLOU
`better.
`BY MS. BERNIKER:
` Q. Is a foil, is that like a wing or is
`that different?
` A. Foil, wing. We use them
`interchangeably.
` Q. Is that -- does that -- is it also
`the case then, given what you explained, that
`you would want to try to keep the number of
`protrusions like wings to a minimum?
` A. If you are talking protrusions in
`general, the wing actually is one of the least
`offensive of the devices because it's
`streamlined naturally. Other protrusions, like
`if you attach something that is bulky and put
`it on, then that will make quite a bit of noise
`continuously, which is even worse. But also
`foils, if -- you have to use them at large
`angles of attack, because they will stall.
` Q. Sounds like generally speaking
`protrusions are something that, to the extent
`possible, you'd avoid in order to avoid
`hydrophone noise?
` A. Yes.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`4
`
`PGS Exhibit 1118, pg. 4
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 369
`
` M. TRIANTAFYLLOU
` Q. So I want to direct your attention to
`the '520 patent specification. Do you have
`that patent in front of you?
` A. Which patent you are talking about?
` Q. The '520.
` A. I have the '520.
` Q. If you could look at column 4,
`please.
` A. Yes.
` Q. It says -- do you see in column 4
`around, starting at line 40, there's a
`sentence? Are you with me?
` A. Yes.
` Q. Okay. It says: "Because the
`movement of the seismic streamer 12 causes
`acoustic noise (both from seawater flow past
`the bird wing structures as well as cross
`current flow across the streamer skin itself),
`it is important that the streamer movements be
`restrained and kept to the minimum correction
`required to properly position the streamers."
` Do you see that?
` A. Yes.
` Q. Do you understand the '520 patent to
`
`Page 371
`
` M. TRIANTAFYLLOU
` Q. Okay. So he is taking noise into
`consideration when he tells you how to go about
`the method of his patent; right?
` A. He mentions, yes, the importance.
` Q. Right. And he says you should
`balance the noise that's created with -- you
`should balance the noise that's created by
`using the streamer positioning devices to
`correct the positioning of the streamers with
`the rest of what you're doing in figuring out
`how to go about your survey; right?
` MR. KIKLIS: Objection, form.
` A. This is -- he puts the considerations
`that they are there. And when you design a
`control system, you use what is called the
`penalty on what you want to control, and the
`penalty on what you want -- on the control
`force you are going to put. Because if you
`design a control system and you don't put
`restrictions on how much force you are going to
`put, the answer will be infinite control. If
`you tell your controller do whatever you want,
`the answer is infinite. If you tell your son
`or daughter to go and buy a car and say go and
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 370
`
` M. TRIANTAFYLLOU
`say that you should be keeping use of the
`streamer positioning devices to a minimum to
`the extent possible?
` A. You should -- you should -- what it
`says, it's the -- that the streamers should be
`kept to -- the streamers should not move
`laterally.
` Q. To the extent you can avoid it?
` A. Yes. So the -- there are some
`priorities. So it says it is important that
`the streamer movements be restrained. Okay.
`Why? Because as I said, if you tow the
`streamer straight, it produces a certain amount
`of noise. But once it starts creating kinks or
`moves, then the noise goes way up, because you
`have cross-flow drag, which causes this --
`these effects. And at the same time, when you
`move the birds -- he says minimum correction
`required -- the birds when they move to angles
`of attack, they will increase the noise.
` So the first order of business is you
`should restrain the movement of the streamers,
`but also try to keep the correction required to
`a minimum.
`
`Page 372
`
` M. TRIANTAFYLLOU
`buy a car, they go and buy a Maserati.
` So you have to put some restrictions.
`So the instructions here are to a control
`person, and what they are told is that one of
`the considerations you have to put is take into
`account the noise so when you penalize, as it's
`called in control, keep in mind that the noise
`is an issue on this. Because the other issue
`would be power. But the electronics engineer
`may say we have plenty of power, we pass it
`through the wires, why don't we do a good job.
`We have powerful thrusters to really wiggle
`around.
` So this instruction here says ah,
`you're going to minimize the control base and
`considerations. Now, you have to -- what it
`is, of noise. And then what you are going to
`control is the motion of the streamers. Again,
`the motion of the streamers influences the
`hydrodynamic noise.
` So these are two considerations which
`he says have to enter into the table. There
`are other considerations too. Okay?
`BY MS. BERNIKER:
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`5
`
`PGS Exhibit 1118, pg. 5
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 373
`
` M. TRIANTAFYLLOU
` Q. Okay. What happens if you have a
`very strong -- so I'm talking about now what
`happens pursuant to the '520 method, let's say.
`Okay? So what do you understand he would
`suggest that you do if you have a very strong
`current that's essentially about to create a
`tangle between the streamers?
` MR. KIKLIS: Objection, form.
` A. Are we talking about which mode to
`use? I mean --
`BY MS. BERNIKER:
` Q. No, I just mean -- well, that's one
`question, I suppose. But what I was -- what I
`meant was -- okay, let me change it, because
`that probably wasn't a good question. I
`apologize.
` Let's say there's a very strong
`current that's about to create a tangle. And
`in order to avoid the tangle, the streamer
`positioning devices have to be used to such a
`degree that the noise is going to make the
`data -- it's going to kind of overpower the
`ability to collect data. Do you have an
`understanding of what the '520 patent says
`
`Page 375
`
` M. TRIANTAFYLLOU
`there will always be probability of something
`occurring, however high. There will never be a
`maximum.
` So some combinations will lead to --
`in cases where the thrusters may stall, and
`then the noise may be excessive. But when you
`have a system which coordinates everything and
`the like, it makes this probability much, much
`smaller than with another system. That's the
`only thing you can say. And I'm not saying it
`simply for the operation of the streamers.
`Sometimes we design systems in the probability
`of failure, meaning a ship may drown -- may
`sink and drown the people onboard, and still
`you calculate the probability of that
`happening, which is unthinkable to the average
`person that you consider. But that's
`unavoidable. Okay?
` So same here. So what you are
`talking about is one probability. Such systems
`make it much less, but it will always be that
`they may stall if the noise is excessive,
`because two or three of these streamers
`stalled, then they may have to repeat the
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 374
`
` M. TRIANTAFYLLOU
`is -- should be done in that circumstance?
` MR. KIKLIS: Objection, form,
` assumes facts.
` A. So this is again a hypothetical
`question, but we have to view this in the
`statistical world because that's how you design
`systems for the ocean. The classic pitch is
`the required class for the master's degree in
`ocean engineering because it teaches you that
`anything in the ocean has to be stochastic,
`because there are waves, there is turbulence in
`the ocean. So we design systems
`stochastically, meaning you will never expect
`to find a wave 3 meters high and 50 meters
`long. It will be a spectrum, a range of such
`things.
` So your question has to be answered
`in the same manner. In other words, we have a
`stochastic environment. The currents are
`variable. There may be waves too. So during
`those combination of random events, there may
`come times when things get exceeded, okay? So
`it will be a lie to say I will design a system
`that will never do that, because statistics say
`
`Page 376
`
` M. TRIANTAFYLLOU
`shoot -- the line, for example. Such events
`may happen, okay? But they get much, much less
`with the systems. If it happens every day, it
`means your system is under-designed or poorly
`designed. Okay?
`BY MS. BERNIKER:
` Q. So if I understand you correctly,
`there are times in the course of executing on a
`marine seismic survey when the concern about
`noise is trumped by other things and you're
`willing to accept the noise in order to, for
`example, avoid a tangle?
` A. It is an unlikely event because when
`you have such multiple streamers and the like,
`and distributed control, you don't think -- you
`don't let things develop as you would
`otherwise. You limit them.
` So we're talking about events which
`are rare, and there may be cases where indeed
`you say look, if we tangle the streamers, we're
`going to lose a few days replacing them and the
`like. If we lose the -- the line, we will lose
`three or four hours to turn around and do it
`again. So you balance what is more expedient.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`6
`
`PGS Exhibit 1118, pg. 6
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 377
`
` M. TRIANTAFYLLOU
` Q. Okay. If we could -- I'd like to
`take a look at paragraph 210 of your expert
`report. This is under the heading "Workman
`Does Not Disclose Model-Based Control."
` Do you see that, sir?
` A. Yes, I do.
` Q. And this is, if you turn back to page
`85, under the heading, "The '520 Patent";
`right?
` A. Yes.
` Q. So this is a discussion of the '520
`patent; right?
` A. Yes.
` Q. In connection with Workman and their
`relationship; right?
` A. Yes.
` Q. Okay. So you say in paragraph 210
`Workman does not disclose model-based control.
` A. Yes.
` Q. When you say model-based control,
`what do you mean by that?
` A. Behavior, behavior-based.
` Q. The behavior-predictive control that
`we discussed yesterday?
`
`Page 379
`
` M. TRIANTAFYLLOU
`Claim 18 because of certain language
`specifically in Claim 18, or is it because the
`specification has language suggesting that the
`invention has to require it in your view?
` A. He makes a point in many locations to
`explain why the behavior of streamers and the
`like, and then he puts it in explicit terms,
`that the inventive control system utilizes. So
`he doesn't say sometimes utilizes or anything.
` Q. You're referring to the column 4
`language?
` A. Yes.
` Q. Okay. And I understand that you
`think that the column 4 language means that
`that must be part of the inventive control
`system. I guess my question is more specific
`to Claim 18. Is your understanding that the
`model-based control is required in Claim 18, is
`that because of specific language of Claim 18
`or is that just because the language in column
`4 you think applies throughout the claims of
`Dr. Bittleston's patents?
` A. First of all, there is the column 4
`explicit statement where he describes what his
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 378
`
` M. TRIANTAFYLLOU
` A. Behavior-predictive model-based
`control.
` Q. What we discussed yesterday?
` A. Yes.
` Q. If you could take a look at claim 18
`of the '520 patent. Can you tell me which
`claim term you're reading to require
`model-based control in Claim 18?
` A. Are you talking about the Claim 18,
`"an apparatus comprising"?
` Q. Yes.
` A. Well, we had a long discussion
`yesterday that he puts -- on column 4, line 16,
`where he says, "To compensate for these
`localized current fluctuations, the inventive
`control system utilizes a distributed
`processing control architecture and
`behavior-predictive model-based control logic
`to properly control the streamer positioning
`devices."
` So his invention has to have this.
` Q. And is it -- so when you read
`Claim 18 do you think that this is -- the
`model-based control is a requirement of
`
`Page 380
`
` M. TRIANTAFYLLOU
`invention would like to have, or would require
`to have. And the second, we see that he says a
`control system configured to use a control
`mode -- 18B, Claim 18B where he says that the
`control system is configured to use a control
`mode from a feather angle mode and so on. And
`a mode requires following a specific
`predetermined curve, and all the discussion
`before has led to these modes because of the
`need to have the model-based control. So the
`language together with the statements leads to
`the model-based, behavior-predictive
`model-based.
` Q. So just to be clear, when you're
`pointing to the language in 18B, is it the
`language "a control system" that you think
`leads -- or supports the conclusion that the
`model-based control is required, or is it the
`language "control mode," or is it both?
` A. It's "control mode" in conjunction
`with what the patent specifications determine,
`specify.
` Q. Okay. I understand. And I think you
`said that you think that a mode requires
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`7
`
`PGS Exhibit 1118, pg. 7
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 381
`
` M. TRIANTAFYLLOU
`following a specific predetermined curve.
` A. That's what "mode" implies in my
`mind.
` Q. Okay. Do you know if that's the same
`construction that WesternGeco's proposing in
`this case?
` A. At this moment, all I recall is my
`statement in my constructions that I had looked
`at the WesternGeco's propositions at the time
`when I wrote my report, and I agreed with them.
`So I have to look back at the details to -- for
`the specific one and which one exactly -- what
`WesternGeco's specifications were to give you a
`definitive answer. You want me to look it up?
` Q. Sure. Let's look at that.
` A. So I'll go to the table of contents.
`Claim construction, 47, and we want
`specifically the feather angle mode?
` Q. No, just the word "mode" or "control
`mode." So you don't have a section in your
`claim construction section on what that term
`means; right? On specifically "mode" or
`"control mode"?
` A. I'm sorry, I thought you meant the
`
`Page 383
`
` M. TRIANTAFYLLOU
`it as a claim construction, no. So go ahead,
`please.
` Q. Okay. So now I just want to talk
`about how you -- kind of how you've used the
`word "control mode" as you've analyzed the
`Workman reference, because you used the term in
`analyzing whether the reference met that
`limitation; right?
` A. Right.
` Q. I'd like to understand what you
`understood the term to mean when you performed
`your analysis in the section that starts on
`page 90 about whether or not Workman discloses
`a control mode.
` A. Right.
` Q. So how did you understand that term
`when you --
` A. So it will be a loose definition
`because I don't have in front of me, you know,
`my analysis or text here to help me with it.
` Q. Oh, it starts on page 90 if you need
`that. I believe it starts on page 90, if you
`want to take a look.
` A. The discussion on modes?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 382
`
` M. TRIANTAFYLLOU
`feather angle mode, because that's what I had
`read.
` Q. Okay. I'm just talking about --
`okay, let's start over then to make sure we're
`in the same place, because I don't want to have
`anything confused.
` Just the concept of a control mode,
`not a specific control mode but just a control
`mode.
` A. I haven't made the construction of
`that.
` Q. You haven't? Okay. And you haven't
`been asked to opine on that question?
` A. Right now, I don't recall whether I
`have specifically seen a discussion of a mode,
`but we may have, but --
` Q. Okay.
` A. Not for a specification, but a
`discussion of what it means, perhaps. I'm not
`sure.
` Q. Okay. Well, you don't have a section
`in your claim construction section on it.
` A. Right. So my reaction is since I
`didn't make a note of it, I -- I didn't discuss
`
`Page 384
`
` M. TRIANTAFYLLOU
` Q. No, the discussion of your analysis
`of Workman.
` A. Right.
` Q. What did you say you didn't have in
`front of you?
` A. All I'm saying is I don't have in
`front of me the discussion -- any discussion of
`the definition of modes. Let me explain
`loosely. So a mode means you set in the -- to
`make it more specific to streamers, you set a
`configuration, an ideal configuration for the
`streamer, which you set and then you try to
`maintain. And of course you set it in such a
`way as to be in agreement with the
`requirements, low noise and so on and so forth.
`So you're not going to put a crooked
`configuration or anything like that,
`intentionally.
` So that's loosely the definition of a
`mode, something that you set, an ideal
`configuration and then try to maintain, which
`is by the way what ION is doing with its
`system, the ghost modes and so on and so forth.
`So they are, I assume, the same kind of loose
`
`TSG Reporting - Worldwide
`(877) 702-9580
`
`8
`
`PGS Exhibit 1118, pg. 8
`PGS v. WesternGeco (IPR2014-01475)
`
`
`
`Page 385
`
` M. TRIANTAFYLLOU
`definition.
` Q. So your loose definition as you've
`been applying it is an ideal configuration
`which you set and try to maintain?
` A. Yes.
` Q. Okay.
` A. And continuously maintain, yeah.
` Q. Continuously maintain?
` A. Maintain has the continuously in it.
` Q. Okay. And you I think have
`acknowledged that you can have -- there's some
`ability to have a manual input in connection
`with a mode. In other words, you can input the
`initial parameters manually; right?
` A. In any system which is operable by --
`especially by captains, you always have a
`manual override because no matter what, the
`captain will be responsible ultimately for the
`vessel so there has to be some manual override
`always.
` Q. But putting apart override, just in
`terms of kind of starting the execution of the
`mode -- well, let me take a step back.
` If you take a look at the '520 patent
`
`Page 387
`
` M. TRIANTAFYLLOU
` Q. You have to say what you want. And
`it's still a mode if the way you enter the
`parameter that you want is through manual
`entering of it; right?
` A. What do you mean, manually? So we --
`I understand that we -- we have
`misunderstandings here. What do you mean,
`manually?
` Q. I thought we meant manually in this
`context to include, for example, using a
`computer terminal to type it in.
` A. So let's say that it's human input,
`user input.
` Q. User input?
` A. Which is a better word than
`"manually."
` Q. The point is it's still a mode even
`if there is a parameter that the user has to
`input; right?
` A. A mode has several parametric
`specifications. The feather mode -- the
`feather angle mode, you have to set the feather
`angle. So it has one parameter which has to be
`specified, yes.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 386
`
` M. TRIANTAFYLLOU
`and you look at Claim 3, it says, "the method
`of Claim 2 comprising inputting the feather
`angle mode manually."
` Do you see that?
` A. Yes.
` Q.