Ex. PGS 1020
`(EXCERPTED)
`
`
`
`
`
`
`(EXCERPTED)
`
`
`
`
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 1 of 390
`
`1209
`
`UNITED STATES DISTRICT COURT
`SOUTHERN DISTRICT OF TEXAS
`HOUSTON DIVISION
`
`09-CV-1827
`Houston, Texas
`
`7:39 a.m.
`July 27, 2012
`
`*
`*
`
`**
`
`***
`
`WESTERNGECO LLC
`VS.
`ION GEOPHYSICAL
`CORPORATION, FUGRO
`GEOTEAM, INC., ET AL
`
`JURY TRIAL
`Volume 5
`
`BEFORE THE HONORABLE KEITH P. ELLISON
`UNITED STATES DISTRICT JUDGE
`
`APPEARANCES:
`FOR THE PLAINTIFF:
`Lee L. Kaplan
`SMYSER, KAPLAN & VESELKA, LLP
`700 Louisiana, Suite 2300
`Houston, Texas 77002
`713.221.2300
`
`Gregg F. LoCascio
`KIRKLAND & ELLIS LLP
`655 Fifteenth Street Northwest
`Washington, DC 20005
`202.879.5290
`
`Sarah Tsou
`Timothy K. Gilman
`KIRKLAND & ELLIS LLP
`Citigroup Center
`153 East 53rd Street
`New York, New York 10022
`212.446.6435
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 68 of 390
`Direct-Triantafyllou
`
`1276
`
`and where you're going with it. In every day language
`usually we say, now is -- the beginning is now, and future
`is ahead. But if you defined the beginning to be the
`depression, the future was rise forward from the
`depression. Okay?
`If we can go to the next slide. Can you give us an
`Q.
`example of some types of software that do this prediction
`to determine based on information you know, what a system
`will or what the mathematics at now or some point after
`the measurement?
`A.
`So there are a number of methodologies. Some of them
`you're going to hear. That's why I put some of them down.
`These are mathematical methods to do exactly what we spoke
`before, take something from a time and move it forward.
`So there's a system called an observer because it observes
`the system and looks forward.
`There's an ultimate observer, which is
`called the Kalman filter because Professor Rudy Kalman
`developed it.
`We've heard that term I think a little bit yesterday,
`Q.
`something called a Kalman filter?
`A.
`Yes.
`And that's this optimal observer?
`Q.
`A.
`Yes. It's an optimal observer.
`And it's named after a scientist?
`Q.
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`09:00:27
`
`09:00:44
`
`09:01:01
`
`09:01:14
`
`09:01:24
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 71 of 390
`Direct-Triantafyllou
`
`1279
`
`We have no idea.
`
`So if you have the Kalman filter on your
`car, what would that Kalman filter do? Would get your
`speedometer to see how fast you go, and which direction
`you go, west, north and south, and will say, based on your
`speed you have to be where this green place is. Makes a
`very simple calculation, velocity 30 miles per hour, times
`half an hour, so much in which direction.
`So it finds -- it gives you an estimate.
`It makes a prediction. Why is it a prediction? Why do
`you call it prediction? You don't have any data to
`corroborate. All you do is you take your own speed. You
`say, hey, I must be there.
`So what happens? The GPS comes in and the
`GPS is the yellow dot. The GPS is very accurate and the
`Kalman filter says hey, last week this guy changed the
`tires, so the speedometer doesn't work very well. The GPS
`is what I will believe. So, yes, I'm here. You got lucky
`this time.
`
`In the other example on the right, the GPS
`comes and shows you you are there. There's no way you're
`on the highway. You know that that's not the case.
`What I predict this green spot is good
`enough for me. So you feel confidence at least of where
`you are. That's what a predictor does for you. Okay? It
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`09:04:34
`
`09:04:52
`
`09:05:06
`
`09:05:26
`
`09:05:43
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 75 of 390
`Direct-Triantafyllou
`
`1283
`
`buy a television they give you the manual. They have such
`detailed manuals, which are very well written, both by
`Q-FIN and as we see from DigiFIN, they write very nice
`manuals and explain how things work.
`So in this case they say explicitly, "Due
`to low sample rate of the position observations," meaning
`the GPS is another example, doesn't come very often. "The
`software runs also a position predictor. The Kalman
`filter predicts where you are."
`So looking at the sentence that you've highlighted,
`Q.
`the low sample rate of the position observations, is that
`explaining that you don't always get an actual measurement
`of where the bird is?
`A.
`Exactly. It takes several seconds before the new
`observation comes. In the meanwhile, as we know those
`waves don't wait for you. They keep traveling. So you do
`this prediction to see where everything is to keep tab.
`And based on your analysis, was the WesternGeco
`Q.
`Q-Marine system covered by its own patent?
`A.
`Yes. So it contains those elements.
`If we can now look at the next slide, did you take a
`Q.
`look at ION's Q systems?
`A.
`Yes. I looked at the manuals of the DigiFIN and the
`related technology to the DigiFIN. And also, I went to
`the Websites of ION and looked at the product. And also I
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`09:10:42
`
`09:10:57
`
`09:11:12
`
`09:11:28
`
`09:11:51
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 137 of 390
`Direct-Triantafyllou/By Mr. LoCascio
`
`1345
`
`C and D. The Claim 15 is just A, B and C.
`Is there is this substantive difference or
`is there they just broke the paragraphs at different
`places?
`A.
`Well, it's a method, so it explains one more item
`that is needed to exercise.
`And did you look for all of those?
`Q.
`A.
`Yes.
`If we could furnish to your next slide. Can you
`Q.
`explain how ION system uses predictor software?
`A.
`So ORCA, which is the intelligence officer, so to
`speak, of the DigiFIN overall system, uses a specific
`methodology. It's called a Kalman filter. It's probably
`one of the most widely used for the -- for various.
`Is this the Kalman filter you talked about earlier?
`Q.
`A.
`It's the Kalman filter I talked earlier, and the one
`that you can put on your car to give you better
`predictions of where you are.
`And so, they use a Kalman filter to
`predict the position and velocity of each node and uses
`the measurements and predictions together to update the
`position and velocity. In other words, to tell you where
`you are now. Okay?
`So let's take the scenario just quickly
`again. You take measurements. The hydrophones, they tell
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`10:33:20
`
`10:33:32
`
`10:33:57
`
`10:34:10
`
`10:34:25
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 138 of 390
`Direct-Triantafyllou/By Mr. LoCascio
`
`1346
`
`you where you are. But the next time they're going to
`tell you when 10 seconds from now. It's a long time. Why
`10 seconds? It takes time to process the signal and the
`like.
`
`In that 10 seconds, you don't have any
`information what's going on. All you know is where you
`were. So you take your system and you say, okay, I'll do
`do a prediction. You'll use the equations of the system.
`It's not just taking a wild guess. And it tells you,
`You're going to be here at now. That's where you are.
`That's a prediction.
`Then the signals come in, and the Kalman
`filter has an ingenious algorithm to decide what is noise
`because some of these measurements are out. It's like a
`GPS that gives you crazy readings. But it's not doing
`just that. It combines that. So that doesn't throw out
`the measurements. It does a careful job. So it's a good
`algorithms to combine prediction and actual noisy
`measurements. There's no clear measurement in such field.
`They're all noisy.
`If we go to the next page, does the patent itself
`Q.
`talk about using predictor software to estimate this
`actual location of the bird?
`A.
`So if we look at the patent, which is in the upper
`left, it's one of the pages of the patent. It says you do
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`10:34:40
`
`10:34:56
`
`10:35:11
`
`10:35:33
`
`10:35:48
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 142 of 390
`Direct-Triantafyllou/By Mr. LoCascio
`
`1350
`
`prediction and filtering works in the accused ION system?
`A.
`Yes, indeed.
`MR. LOCASCIO: Turn to 69, Dave.
`THE WITNESS: If we turn to 69.
`BY MR. LOCASCIO:
`What did you say in there?
`Q.
`A.
`We see the Kalman filter and an arrow that says "raw"
`next to it. "Raw" means the data as they come. We
`haven't filtered and we haven't done anything to them.
`That's what "raw" means.
`And this produces an output, which is
`going to be used next. So the Kalman filter, as it says
`in the bottom, the Kalman filter part runs all the time,
`grabbing data and using that to update its position and
`velocity and streamer shape and arrow.
`And so, is the Kalman filter used by ION used to
`Q.
`estimate the current position of the birds?
`A.
`Yes, it is.
`In the next slide, are there other flow charts
`Q.
`identifying this Kalman filtering process in their
`operation document?
`A.
`Yes, it is. And here we see some of the prediction
`unit and the like, which is part of the Kalman filter. We
`don't need to go into any specific detail, but every
`Kalman filter has a predictor and an adjustment. The
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`10:39:20
`
`10:39:33
`
`10:39:54
`
`10:40:08
`
`10:40:25
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 143 of 390
`Direct-Triantafyllou/By Mr. LoCascio
`
`1351
`
`filter maintains its idea of time, meaning has a clock, so
`it can tell what time it is.
`When new data arise, the predicator is
`used to carry the state forward to time of observation.
`Time of observation is now, when the observation comes in.
`It cannot be the future in terms of the Kalman filter.
`Then the predictions of the data are
`thrown into the adjustment. So it's like a blender. You
`put them in and the algorithms comes up with an optimal
`solution. That's all there is to it.
`This passage here that you just read, do ION's own
`Q.
`documents talk about using the predictor to carry the
`state forward to time of observation, being present or
`actual time?
`A.
`Yes. Because in order to have observations, it can
`be at most now.
`We can't have the future observations?
`Q.
`A.
`The future will come later.
`If we look at the next slide, did ION's own employees
`Q.
`recognize, in your review of the deposition testimony?
`A.
`Yes. This was -- sometimes, you know, documents that
`were earlier to confirm this, and he's asked if it's part
`of the NCN, and he said yes.
`"So the Kalman filter is predicting the
`position of the DigiFIN node?
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`10:40:41
`
`10:40:56
`
`10:41:09
`
`10:41:25
`
`10:41:41
`
`Ex. PGS 1020
`
`
`
`Case 4:09-cv-01827 Document 449 Filed in TXSD on 07/30/12 Page 144 of 390
`Direct-Triantafyllou/By Mr. LoCascio
`
`1352
`
`"It's predicting the position" --
`THE COURT: No, you're going too fast.
`THE WITNESS: I'm sorry.
`BY MR. LOCASCIO:
`The second passage you refer to --
`Q.
`A.
`I'm sorry.
`That's okay.
`Q.
`-- Mr. MacNab was asked: "So the Kalman
`filter is predicting the positions of the DigiFIN node:
`And his answer was: "It's predicting the
`positions of all nodes."
`And then on the next question, he was
`asked: "The Kalman filter is predicting the position of
`the DigiFIN devices, that's the birds?"
`And he answered?
`A.
`"Yes." So the nodes correspond to the device. The
`fins, the DigiFINs that are spread around.
`So based on the information of how they use it and
`Q.
`the algorithms itself, and Mr. MacNab and other testimony
`you identified the limitation of a prediction unit as
`being met?
`A.
`Exactly.
`Now, I want to ask you about this particularly
`Q.
`because there's specific software that runs this Kalman
`filter; right?
`
`123456789
`
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Johnny C. Sanchez, RMR, CRR - jcscourtreporter@aol.com
`
`10:41:46
`
`10:41:55
`
`10:42:05
`
`10:42:21
`
`10:42:33
`
`Ex. PGS 1020
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
