[00:00:05] Speaker 02:
May it please the court?

[00:00:07] Speaker 02:
You look familiar.

[00:00:13] Speaker 02:
In this case, the board misconstrued the term predicted road section coefficients as if the word predicted was absent from the claim language and further ignored.

[00:00:22] Speaker 01:
The difference is that you want to have predicted modified coefficients and they want to have predicted or the board wants to have predicted modified road section.

[00:00:31] Speaker 01:
Is that right?

[00:00:33] Speaker 02:
Google has made arguments that predicted should modify road sections.

[00:00:38] Speaker 02:
I don't read the board as saying that at all.

[00:00:41] Speaker 02:
They're saying that there's some prediction

[00:00:47] Speaker 02:
I think that the board fairly reading it is saying that the route is predicted.

[00:00:51] Speaker 02:
So I guess in a way it is the route sections on the route that is predicted, Your Honor, excuse me.

[00:00:56] Speaker 01:
So your argument is twofold.

[00:00:59] Speaker 01:
It's that predicted modifies coefficients and then you make another a leap that says, and therefore it must also be dynamic, right?

[00:01:10] Speaker 01:
Correct.

[00:01:11] Speaker 01:
Can you tell us about that leap?

[00:01:14] Speaker 02:
I can tell you about that leap, Your Honor.

[00:01:16] Speaker 02:
A prediction is something that must change over time.

[00:01:19] Speaker 02:
It has to be dynamic.

[00:01:20] Speaker 02:
It's in the very word predicted.

[00:01:21] Speaker 02:
You don't predict something that stands still.

[00:01:25] Speaker 02:
It's in the common language, I mean, it's the common definition of predictive.

[00:01:30] Speaker 02:
Speak up a bit, please.

[00:01:31] Speaker 02:
Yes, Your Honor.

[00:01:33] Speaker 02:
So that is, once the road section coefficients are predicted as is properly construed.

[00:01:40] Speaker 01:
Again, that goes to your, that you're using predicted to modify coefficients.

[00:01:45] Speaker 02:
Understood, Your Honor.

[00:01:46] Speaker 02:
And I, if I can address that as well, but I don't think it actually is a leap that they have to be dynamic because they are predicted.

[00:01:51] Speaker 02:
I think it flows directly, logically.

[00:01:55] Speaker 02:
any prediction is something that must change over time.

[00:01:59] Speaker 02:
And on the first point, whether predicted modifies road sections or coefficients, Google's expert made this clear when he was asked about the Matsumoto reference and which, because there's a lot of coefficients or multiplicative factors in the two references that are being combined here.

[00:02:25] Speaker 02:
And Google's expert Dr. Brash was asked, what in this mix can actually qualify as a predicted road section coefficient?

[00:02:36] Speaker 02:
And he said that the only piece of data in these two

[00:02:42] Speaker 02:
references that could qualify as a predicted road section coefficient is the mean required traveling time.

[00:02:50] Speaker 02:
This is at appendix 1149.

[00:02:51] Speaker 02:
I think that's our motion for observations.

[00:02:56] Speaker 02:
And then it's in appendix 880.

[00:02:59] Speaker 02:
The deposition testimony was there.

[00:03:02] Speaker 02:
That's because this is the only variable in the whole mix that could arguably, and we dispute this, that could arguably be dynamic.

[00:03:12] Speaker 02:
We dispute that it actually is because the entirety of Matsumoto is a static calculation.

[00:03:18] Speaker 02:
And there's these, I think it's expressed as three words, mean, traveling time, and four words in another place, mean, required, traveling time.

[00:03:25] Speaker 02:
There's no discussion of what this term actually means.

[00:03:29] Speaker 02:
It's a mystery.

[00:03:31] Speaker 02:
And it happens to be in a translated Japanese patent, and it talks totally about static factors, the number of turns,

[00:03:40] Speaker 02:
on a road segment, the distance of a road segment, the width of a road segment, and then there's this mean traveling time.

[00:03:48] Speaker 02:
And in the combination, which Google has challenged the claims with, the combination takes the entirety of the

[00:04:01] Speaker 02:
Zhu reference and tries to put it in these three or four words of Matsumoto, which we argue is reconstructing the claim.

[00:04:09] Speaker 02:
And it does not make any sense in the context of the two references.

[00:04:14] Speaker 02:
This combination starts out with the Zhu reference as the primary reference, and then uses some ideas from Matsumoto about hierarchically computing a root search, doing it in stages for different types of roads.

[00:04:29] Speaker 02:
OK.

[00:04:30] Speaker 02:
Then from there,

[00:04:32] Speaker 02:
they take Matsumoto, which is a different type of calculation.

[00:04:36] Speaker 02:
And you can see this in the Matsumoto reference in column one of Matsumoto.

[00:04:43] Speaker 02:
And this is at appendix 543.

[00:04:47] Speaker 02:
There is this formula for how Matsumoto

[00:04:58] Speaker 02:
identifies its root, and it's all these static variables with multiplicative factors.

[00:05:04] Speaker 02:
There's weighting coefficients for each of these that says it's greater than or equal to zero.

[00:05:09] Speaker 02:
There's somehow this big addition of

[00:05:14] Speaker 02:
a distance cost, a mean travel time cost, and a road width cost, each of which are multiplied by some factor of zero or greater than zero.

[00:05:26] Speaker 02:
And then there's a fourth term and these are all summed together and it's unclear how any of this would work.

[00:05:30] Speaker 02:
There's the number of turns and that's squared.

[00:05:33] Speaker 02:
They take that and they plug you into one of those terms and say, there is your predicted road section coefficient.

[00:05:44] Speaker 02:
And there's no description of how this would work, or why this would work, or what you would need to do to either of these references to make it work in this way.

[00:05:52] Speaker 02:
And it's circular, taking Zhu, modifying it with Matsumoto, and then somehow taking Matsumoto and modifying it with Zhu in this way.

[00:06:01] Speaker 02:
And Zhu purports to be a holistic system that comes up with roots and plans that.

[00:06:09] Speaker 02:
If you put that into, this would replace the entire cost function if you wanted to.

[00:06:14] Speaker 02:
It would not be any sort of multiplicative factor against which you multiply anything, which is the basis for what is a coefficient.

[00:06:22] Speaker 02:
It has to be a multiplicative factor in some sort of equation.

[00:06:26] Speaker 02:
And that's how it's disclosed in the 7.83 patent.

[00:06:28] Speaker 02:
The regular travel times and the current travel times are multiplicative factors that are, they're multiplied against the theoretical travel time to get you

[00:06:37] Speaker 02:
your root calculation.

[00:06:39] Speaker 02:
None of this is present here.

[00:06:48] Speaker 02:
So the logic does not add up that you can combine these references in this way, and it lacks substantial evidence, we argue.

[00:06:57] Speaker 02:
And I would ask you

[00:07:02] Speaker 02:
If counsel was going to get up here shortly and argue himself, if counsel can make sense of this equation and how these two fit together and where Zhu fits into Matsumoto, it would be a good question to ask of counsel because it's not been described anywhere.

[00:07:15] Speaker 02:
Further, Dr. Brash, Google's expert, made the statement that using historical information to determine a likely outcome is how Matsumoto's mean required traveling time works.

[00:07:27] Speaker 02:
And in our claim construction argument and in the context of the claims, you will see that these predicted road section coefficients are used to, they are used to compute a route search based on current and statistical section data.

[00:07:43] Speaker 02:
If you're only talking about past data and not any current data, then they do not qualify as the predicted road section coefficient in the challenged claims.

[00:07:51] Speaker 03:
You do understand that what the patent office determines the prior art says is a question of fact.

[00:07:58] Speaker 02:
I understand that, Your Honor.

[00:07:59] Speaker 03:
And that under our standard of review, if there's substantial evidence in the record to support patent office, you lose.

[00:08:11] Speaker 03:
Understood, Your Honor.

[00:08:12] Speaker 03:
It's basically very hard for us to sit up here and tell the Patent Office, you didn't understand the prior art as well as we do.

[00:08:24] Speaker 02:
Understood, Your Honor.

[00:08:25] Speaker 02:
I would say, in general, the gist of the final written decision misconstrues the term predicted and goes off from there on the tangent.

[00:08:32] Speaker 02:
And now they are pointing to save this decision and saying, well, there's one place where we point to where you can put Matsumoto in here, and that's the predicted road section coefficient.

[00:08:42] Speaker 02:
And that's what we argue against, not the overall findings of the Patanau case.

[00:09:05] Speaker 00:
May it please the Court.

[00:09:07] Speaker 00:
The central issue on appeal is whether the limitation reciting predicted road section coefficient should be so narrowly construed such that the term predicted can only mean dynamic or variable.

[00:09:21] Speaker 00:
The Board considered this argument, fully vetted intrinsic evidence, and determined that the claim language cannot be construed so narrowly.

[00:09:29] Speaker 00:
Specifically, in the final written decision, the Board noted that MACOR failed to provide any citation to evidence that would support its attempt to import a dynamic or variable restriction into the claim language.

[00:09:44] Speaker 00:
The board also explained that MACOR did not point to any portion of the specification showing that dynamic or variable coefficients are required by the claim.

[00:09:55] Speaker 00:
That's at A21.

[00:09:57] Speaker 00:
As a result, and after having itself reviewed the intrinsic evidence, the board found that the specification, quote, did not show any intent to limit the term predicted road section coefficients.

[00:10:09] Speaker 03:
Let me read you something and have you tell me what's wrong with it.

[00:10:15] Speaker 03:
There is a clear difference between the act of predicting something and using that something as a predictor, which the board clearly conflates.

[00:10:27] Speaker 03:
The coefficients are themselves predicted.

[00:10:30] Speaker 03:
Google makes its contrary position clear in its brief when it equates, quote, predicted road section coefficients, unquote,

[00:10:38] Speaker 03:
to road section coefficients that are, quote, capable of predicting or foretelling road section speeds.

[00:10:46] Speaker 03:
As the overall claim language of claim one and the specification of the 783 patent make clear, however, quote, predicted road section coefficients, unquote, are coefficients that are themselves predicted.

[00:10:59] Speaker 03:
and not merely coefficients that are used in a root search.

[00:11:04] Speaker 03:
That seemed to me to be the clearest statement of any of you as to what's going on with this predicted coefficient.

[00:11:12] Speaker 03:
What's wrong with their explanation?

[00:11:15] Speaker 00:
So let me try a three-part answer of that.

[00:11:17] Speaker 00:
First, Your Honor, regardless of whether predicted modifies coefficient, as Maker argues, or predicted modifies road section, the board's analysis and holding and determination still stand, because there's simply nothing in the term predicted that, I'm sorry, there's nothing in the specification.

[00:11:33] Speaker 00:
There's no lexicography.

[00:11:35] Speaker 00:
There's nothing that would satisfy the stringent and

[00:11:38] Speaker 00:
demanding standard that a patentee set forth a specific definition that departs from the ordinary meaning.

[00:11:44] Speaker 00:
Maker has not pointed to any of that.

[00:11:46] Speaker 00:
That would require, predicted, regardless of whether it modifies coefficient or modifies road section, there's nothing in the intrinsic evidence that limits that term solely to dynamic or variable.

[00:11:57] Speaker 00:
Second, your honor,

[00:11:59] Speaker 00:
There's nothing in a patent that ever describes predicting a coefficient.

[00:12:03] Speaker 00:
The entirety of the patent is predicting an optimal route search.

[00:12:13] Speaker 00:
The way you predict an optimal route search, you forecast what are the best

[00:12:17] Speaker 00:
individual road sections or links that constitute that route.

[00:12:22] Speaker 00:
So that really shows that what predicted is modifying is the road sections, not the predicted coefficients.

[00:12:28] Speaker 00:
Third, I would suggest, Your Honors, that regardless of whether, again, predicted modifies coefficient or modifies road sections, the Board found that even under MAKOR's argument that the Matsumoto reference

[00:12:44] Speaker 00:
disclosure of a mean required travel time expressly discloses the predicted road coefficient as Maker characterizes it.

[00:12:53] Speaker 00:
In particular, in its brief, Maker repeatedly says that a predicted road coefficient must be an empirical speed coefficient.

[00:13:03] Speaker 00:
The Board

[00:13:04] Speaker 00:
having evaluated, as Your Honor noted under substantial evidence, having evaluated the substance of the prior art, having credited, and having evaluated and credited Dr. Brasch's unchallenged testimony, concluded that the mean required travel time in Matsumoto is precisely what is described as a empirical speed coefficient in the 783 patent.

[00:13:27] Speaker 00:
So I think, Your Honor's regard,

[00:13:30] Speaker 00:
There's a dispute, but ultimately, regardless of whether predicted modifies coefficient or a road section, the judgment should be affirmed.

[00:13:38] Speaker 03:
Your answer is it can be a static as well as a dynamic.

[00:13:43] Speaker 00:
Right.

[00:13:44] Speaker 00:
Absolutely, Roger.

[00:13:46] Speaker 00:
There's nothing that limits predicted to constantly changing.

[00:13:49] Speaker 00:
It certainly can.

[00:13:50] Speaker 00:
That's certainly one form of a prediction, but it doesn't need to be that way.

[00:13:55] Speaker 03:
Okay.

[00:13:55] Speaker 03:
Thank you.

[00:13:57] Speaker 00:
Unless your honors have any questions, I'll get you back on track, and I ask that you affirm the Board of Judgment below.

[00:14:02] Speaker 00:
Thank you.

[00:14:10] Speaker 02:
Your Honor, just to address a couple of points raised by counsel, nothing in the patent discusses predicting a route or predicting a search.

[00:14:19] Speaker 02:
It talks about predicting conditions within that route and generating the best, fastest route.

[00:14:26] Speaker 02:
Predicting the route itself doesn't make sense.

[00:14:29] Speaker 02:
There's no, it doesn't make sense in the context of the patent.

[00:14:36] Speaker 02:
There's no predicting a route and then weighing it against something.

[00:14:41] Speaker 02:
Council brought up empirical speed coefficients, and there's a little confusion discussing in the briefing where empirical is used by us to mean real world data.

[00:14:54] Speaker 02:
In the patent, empirical speed coefficients are the regular times.

[00:14:59] Speaker 02:
And I would just point out that under the language of the challenge claims that the

[00:15:07] Speaker 02:
Predicted road section coefficients are used to allow the route search to be computed based on current and statistical section data.

[00:15:17] Speaker 02:
If you are limiting the Matsumoto mean required traveling time to statistical data, then you are not meeting the language of the claims.

[00:15:26] Speaker 02:
It's lacking.

[00:15:31] Speaker 01:
If your honors don't have any questions.