[00:00:00] Speaker 04:
We will hear argument next in number 231019, Roku Against Universal Electronics.

[00:00:14] Speaker 03:
Good morning, Your Honours, and may it please the Court.

[00:00:17] Speaker 03:
This appeal centres on a single dispositive claim construction issue, as to the meaning of the phrase, data that functions to identify.

[00:00:25] Speaker 03:
Now that's a non-technical term that has broad meaning, meaning that the data must be used to facilitate the function of identifying

[00:00:33] Speaker 03:
without regard to whether additional processing is performed on the data or whether other information is also used.

[00:00:41] Speaker 04:
One of the things that I guess I found rather scarce in the briefing is references to sources, whether it's case law from this context or others, about a phrase functions to.

[00:01:01] Speaker 04:
And for example,

[00:01:03] Speaker 04:
This would come up in 112.6 context.

[00:01:10] Speaker 04:
I'm not sure whether we would say in that context that something performs a function if it really enables it.

[00:01:20] Speaker 04:
In fact, I think we have said not.

[00:01:24] Speaker 04:
And your word facilitate is rather close to this enable idea.

[00:01:31] Speaker 03:
Yes, Your Honor, so we didn't cite any extrinsic evidence such as dictionary definitions or the like for the plain meaning of the phrase functions, too.

[00:01:40] Speaker 03:
And as far as we could tell, this court has not construed that phrase in other cases.

[00:01:45] Speaker 03:
To answer Your Honor's questions about Section 112,

[00:01:49] Speaker 03:
That's a slightly different context in which we're determining whether a claim term has sufficient disclosure for a function in the specification.

[00:02:02] Speaker 03:
But here, we're talking about HDMI functionality.

[00:02:06] Speaker 03:
And the claim expressly recites and is directed to that functionality.

[00:02:20] Speaker 04:
UVD sends out or sends the word pause to describe its current state.

[00:02:28] Speaker 04:
That functions to identify something other than pause, namely play or some controllable function

[00:02:36] Speaker 04:
that many users would want to exercise next when they have that state.

[00:02:41] Speaker 03:
That's exactly right, Your Honor.

[00:02:43] Speaker 03:
That's what the prior chart in reference discloses.

[00:02:46] Speaker 03:
And if we think about it, that makes sense, because the home theater device, which is the entity that is controlling the appliance, can't know about how to control the appliance unless it knows how the appliance is currently being controlled.

[00:03:00] Speaker 03:
And that's what the specification of the 486 patent discloses, Your Honor.

[00:03:04] Speaker 00:
Do you think that's one embodiment in the specification?

[00:03:08] Speaker 03:
I do, Your Honor.

[00:03:09] Speaker 03:
The passage from column 14 through column 15 is a single continuous embodiment.

[00:03:17] Speaker 03:
That embodiment is describing at appendix 63, column 14, line 30,

[00:03:25] Speaker 03:
It is describing icon metadata or icon information.

[00:03:31] Speaker 03:
And everybody agrees, the board included, that icon metadata is a type of data that functions to identify.

[00:03:38] Speaker 03:
And then, at appendix 64, column 15, line 45,

[00:03:44] Speaker 03:
The patent says that ICON metadata can convey the appliance's, quote, current status.

[00:03:51] Speaker 03:
And that is exactly what the prior Chardon discloses, Your Honor, and that's significant.

[00:03:56] Speaker 00:
Can I ask you another question before you move on to the question?

[00:04:00] Speaker 00:
Yes, Your Honor.

[00:04:01] Speaker 00:
I wanted to ask you, is your argument that the board improperly read in something like

[00:04:10] Speaker 00:
functions to directly identify or functions alone to identify, those kinds of words.

[00:04:17] Speaker 03:
Yes, Your Honor.

[00:04:19] Speaker 03:
Not only did the board read out the phrase functions, too, which is a broad phrase, the board read in the word itself.

[00:04:27] Speaker 03:
And in our view, the board's construction means that the data has to itself communicate.

[00:04:33] Speaker 03:
And alone.

[00:04:34] Speaker 00:
Alone.

[00:04:34] Speaker 03:
And directly.

[00:04:35] Speaker 03:
And directly.

[00:04:36] Speaker 03:
It's an extraordinarily narrow interpretation, Your Honor.

[00:04:39] Speaker 03:
I believe the only way to satisfy that interpretation would be if the appliance tells the home theater device, my controllable function is x.

[00:04:47] Speaker 03:
And as I was saying in the beginning, that's inconsistent with that.

[00:04:51] Speaker 00:
So I would emphasize that it says a controllable function, right?

[00:04:54] Speaker 00:
That's exactly right.

[00:04:55] Speaker 00:
It doesn't say the controllable function or a specific controllable function.

[00:04:59] Speaker 00:
I don't even know how you would say it.

[00:05:01] Speaker 00:
But it does say a controllable function.

[00:05:03] Speaker 03:
Well, that's correct, Your Honor.

[00:05:05] Speaker 03:
Any controllable function.

[00:05:07] Speaker 03:
And as long as a DVD player says that I am in a pause state, the TV uses that information to infer that a controllable function is play.

[00:05:17] Speaker 03:
It doesn't have to be the only controllable function.

[00:05:20] Speaker 03:
And if we think about it, this has to be the case because the claim is directed to HDMI functionality.

[00:05:27] Speaker 03:
And under standard HDMI protocol, the home theater device is the entity that determines appliance capabilities using the claimed data.

[00:05:38] Speaker 03:
UEI itself has conceded that point both before this court at page 38 of its red brief, as well as before the board at appendix 165

[00:05:50] Speaker 03:
and 245.

[00:05:52] Speaker 03:
And that's the key concession.

[00:05:53] Speaker 03:
Because if the home theater device identifies the controllable function using the claim data, then there is no reason for the claim data to itself communicate the claim, the controllable appliance.

[00:06:08] Speaker 03:
That context supports our construction, Your Honors, and the board ignored it.

[00:06:12] Speaker 01:
Those embodiments, though, that would be captured under the board's narrow construction would also still be within the scope of the claims under your broader construction.

[00:06:21] Speaker 03:
That's exactly right, Your Honor.

[00:06:22] Speaker 03:
Our construction encompasses and is broader than UEI's narrower construction.

[00:06:27] Speaker 03:
That the board adopted and I think that goes to show why the board's reading of the specification cannot be correct Because the board noted that our or I'm sorry that the specification doesn't preclude you guys construction and that there's support for both parties construction Constructions and yet the board used that to determine that there's ambiguity in the specification

[00:06:51] Speaker 03:
But I would submit, Your Honor, if the specification is broad enough to support both parties' constructions, then it supports the broader construction.

[00:06:59] Speaker 03:
Or stated differently, if the specification doesn't preclude our construction, it necessarily precludes the board's construction.

[00:07:07] Speaker 00:
What about an argument that isn't it kind of odd to have the claim, say, data, for example, saying pause.

[00:07:19] Speaker 00:
would function to identify a different function than the function that the data is actually identifying?

[00:07:26] Speaker 00:
Does that make sense?

[00:07:27] Speaker 00:
Yes, Your Honor.

[00:07:28] Speaker 00:
I think I understand.

[00:07:28] Speaker 00:
The pause is actually causing you to identify play as opposed to pause identifying pause.

[00:07:34] Speaker 03:
I understand the question, Your Honor.

[00:07:35] Speaker 03:
And I don't think there is anything inconsistent or awkward with that

[00:07:40] Speaker 03:
conclusion, I think a message can expressly identify content and in so doing function to identify other content.

[00:07:49] Speaker 03:
And I think it has to be that way because the home theater device will take as input the status data and will use that input to determine perhaps by looking at a lookup table or some algorithm that the output that is the controllable function while in that pause state is play.

[00:08:10] Speaker 03:
I don't see the inconsistency in that, Your Honor.

[00:08:13] Speaker 03:
I will note that UEI's expert also conceded during his deposition that claim nine of the 486 patent covers that very embodiment, where an appliance conveys that it is an apostate.

[00:08:28] Speaker 03:
And in so doing, the home theater device determines that the controllable function is play.

[00:08:36] Speaker 04:
Just getting back to the language that is following up on

[00:08:39] Speaker 04:
which Judge Stoll was saying, you're reading the phrase receiving at the home theater device data that functions to identify a controllable function to mean

[00:08:57] Speaker 04:
the same as receiving at the home theater device data that the home theater device uses to identify a controllable phone.

[00:09:07] Speaker 04:
That's correct, your honor, and just to be... I guess it seems to me that the difference between those two is in

[00:09:18] Speaker 04:
Who or what is doing the identifying?

[00:09:22] Speaker 03:
I think that's an accurate description, Your Honor.

[00:09:25] Speaker 03:
And I'll note again that the UEI acknowledged in its brief at page 38 that the home theater device has identified the controllable function in that receiving step.

[00:09:36] Speaker 03:
Again, they acknowledged that before the board at appendix 165 and 245.

[00:09:41] Speaker 03:
And that goes to Your Honor's point.

[00:09:42] Speaker 03:
If the home theater device is the entity that is identifying the controllable function using the claim data,

[00:09:48] Speaker 03:
there is no reason for the data to itself have to expressly and directly identify the controllable function.

[00:09:59] Speaker 03:
I just did want to get back briefly to the point that UEI's own expert admitted at appendix 3139 during his deposition that claim nine covers that Chardon embodiment that we've been discussing.

[00:10:12] Speaker 03:
And if claim nine covers Chardon, then claim one must as well.

[00:10:16] Speaker 03:
The board erred by a concern.

[00:10:18] Speaker 04:
One of the mysteries for me, speaking personally in this case, is what claim nine means.

[00:10:26] Speaker 04:
I have no idea how to explain that.

[00:10:29] Speaker 03:
Yes, your honor.

[00:10:30] Speaker 03:
So claim nine specifies that the controllable function is a media rendering function that is provided via use of the appliance and UEI Well Sure, yes, your honor we read that to mean that the controllable function is

[00:10:54] Speaker 03:
depends on how the appliance is currently being used.

[00:10:57] Speaker 03:
UEI itself has conceded that point in its red brief, I believe, at page 36.

[00:11:07] Speaker 03:
Yes.

[00:11:08] Speaker 03:
The provided via use of the controllable appliance language could indicate

[00:11:14] Speaker 03:
that the controllable function simply exists as a result of the current state of the controllable appliance.

[00:11:21] Speaker 03:
I agree with that reading, Your Honor, and if that's true, then the data that functions to identify the controllable function must be broad enough to somehow account for the current state

[00:11:32] Speaker 03:
of the appliance.

[00:11:33] Speaker 03:
And that's exactly our argument.

[00:11:35] Speaker 03:
That's exactly what we see in the specification in column 15.

[00:11:38] Speaker 03:
And that's exactly what UEI's expert has testified is covered by the claim.

[00:11:43] Speaker 03:
To date, UEI has never tried to reconcile its expert's testimony with the board's construction.

[00:11:50] Speaker 01:
UI says you're not arguing that the prosecution history supports your construction.

[00:11:56] Speaker 01:
Is that correct?

[00:11:58] Speaker 03:
I would say that is correct with one tiny caveat, and that is at appendix 864 in the prosecution history, the applicants don't use the phrase data that functions to identify, they use a broader phrase, data indicative of.

[00:12:13] Speaker 03:
I think that choice of words actually does support our construction because

[00:12:18] Speaker 03:
data indicative of is broader than the board's very narrow construction.

[00:12:22] Speaker 03:
Aside from that caveat, I would agree with your honor.

[00:12:25] Speaker 01:
So most of the prosecution history is relevant to you explaining why it doesn't support their construction as opposed to actually helping theirs?

[00:12:33] Speaker 03:
That's correct, your honor.

[00:12:34] Speaker 03:
Clearly there is a dispute between the parties as to whether the board relied on a disclaimer theory without finding that the heightened standard had applied.

[00:12:43] Speaker 03:
applies.

[00:12:44] Speaker 03:
For the reasons set forth in our briefing, I think that's exactly what the board did, and UEI doesn't have a good response.

[00:12:52] Speaker 00:
You're right, there's disclaimer, but there's also just relying on the prosecution history in order to be able to understand the meaning of a claim term.

[00:12:59] Speaker 00:
You've got a separate argument on that point.

[00:13:02] Speaker 03:
That's correct.

[00:13:03] Speaker 03:
That's exactly correct, Your Honor.

[00:13:05] Speaker 03:
We're not saying that the file history cannot be considered, it should be considered, as intrinsic evidence.

[00:13:10] Speaker 03:
But the probative value of the file history depends on the clarity of the claims, the specification, and the file history.

[00:13:17] Speaker 03:
Here, the file history is muddled.

[00:13:20] Speaker 03:
The applicant's remarks at Appendix 864 comprise a single paragraph that span the entire page.

[00:13:26] Speaker 03:
The board itself noted that the applicant intermingled multiple bases for distinguishing the prior art.

[00:13:32] Speaker 03:
And as I noted to Judge Stark, the applicant didn't even use the claim phrase at issue here.

[00:13:38] Speaker 03:
So even if this court disagrees that we're in a world where the only way the file of history can narrow the scope of the claims is through disclaimer, I think the probative value of the file of history is just very weak in this case, given how muddled and ambiguous it is.

[00:13:52] Speaker 00:
And emphasizing it particularly, I think

[00:13:55] Speaker 00:
Sakai was being distinguished primarily because it wasn't because there was a separate server that provided the information identifying the information identifying the controllable function is that right or it didn't come from a

[00:14:11] Speaker 03:
Yes, your honor is exactly correct.

[00:14:15] Speaker 03:
At Appendix 864, the applicant was distinguishing Sakai state data not based on the nature of the data itself, but which device within Sakai's architecture uses the data.

[00:14:27] Speaker 03:
The claims require the home theater device to use the data to add an icon to the user interface.

[00:14:32] Speaker 03:
But in Sakai, the server unit doesn't do that.

[00:14:35] Speaker 03:
it forwards the data onto a different device, the remote control, which then adds the icon to the user interface.

[00:14:42] Speaker 03:
That is the distinction that the applicant seems to be making at Appendix 864 as best as I can tell.

[00:14:50] Speaker 03:
Even to the extent that the applicant were distinguishing Sakai state data based on the nature of the data itself, that's different from Chardon state data because as we noted in our briefing, Sakai state data refers to the state of a menu

[00:15:04] Speaker 03:
on the television.

[00:15:06] Speaker 03:
That is, if you can navigate a menu on the television by clicking up, down, or left or right,

[00:15:12] Speaker 03:
then the up, down, or left, right buttons will appear on the remote control.

[00:15:16] Speaker 03:
In that sense, Sakai uses state data.

[00:15:19] Speaker 03:
That's completely different from Chardon state data, and UEI has never addressed that.

[00:15:24] Speaker 01:
The board says at 821 on the claim construction that the patent owner's position on claim construction is supported by Dr. Turnbull's testimony.

[00:15:32] Speaker 01:
Does that mean we need to reach the extrinsic evidence?

[00:15:35] Speaker 01:
and review it deferentially?

[00:15:37] Speaker 03:
No, Your Honor, I don't think so.

[00:15:38] Speaker 03:
I believe that's the only place in the entire analysis where the board even cited, or maybe one of two places, where the board cited the expert testimony.

[00:15:46] Speaker 03:
And even there, there's no fact finding to which this court could give deference under Teva or Knowles.

[00:15:54] Speaker 03:
This expert testimony merely pertains to how the expert would read the intrinsic record, not as to the meaning of the term in the art or any extrinsic source of evidence.

[00:16:05] Speaker 03:
I would submit that to the extent any extrinsic evidence is probative, it's UEI's expert's deposition that I, testimony that I noted earlier, where he acknowledges that the claim nine does cover Chardon.

[00:16:19] Speaker 03:
Thank you.

[00:16:20] Speaker 04:
Thank you, Your Honors.

[00:16:35] Speaker 02:
Good morning, Your Honors.

[00:16:36] Speaker 02:
May it please the Court?

[00:16:38] Speaker 02:
The Board's construction of the data that functions to identify limitation was correct, whether you look at the plain language of the claims, whether you look at the specifications, and whether you look at the prosecution history, which wasn't correct.

[00:16:55] Speaker 02:
If you look at the plain language of the claims, Your Honor mentioned the words functions to.

[00:17:00] Speaker 02:
It says what the data does, what the purpose of it is.

[00:17:03] Speaker 02:
It identifies

[00:17:05] Speaker 02:
a controllable function of a controllable appliance.

[00:17:09] Speaker 02:
Now if you look at the claim language, it's very important that this data comes from the controllable appliance itself.

[00:17:19] Speaker 02:
It's in the receiving step, the first step.

[00:17:21] Speaker 02:
Receiving at the home theater device from a controllable appliance, et cetera, data that functions to identify.

[00:17:29] Speaker 02:
So the board looked at that language, and then citing Phillips,

[00:17:34] Speaker 02:
appendix page 19 of the board's decision.

[00:17:36] Speaker 02:
It says, well, the claims themselves provide substantial guidance as to the meaning of particular terms.

[00:17:41] Speaker 02:
So let's look at the rest of the claim.

[00:17:43] Speaker 02:
So they looked at the other two steps, the automatically adding step and the causing step.

[00:17:49] Speaker 02:
And both of those steps have the following language, the controllable function of the controllable appliance that was identified by the data received from the controllable appliance.

[00:18:01] Speaker 02:
And the board concluded, OK,

[00:18:03] Speaker 02:
The antecedent basis for that is the claim element at issue from the receiving step.

[00:18:10] Speaker 02:
So that must mean that the data itself identifies the controllable function of the controllable appliance.

[00:18:17] Speaker 01:
I don't understand where the data itself comes in.

[00:18:20] Speaker 01:
How is the automatically step or the causing step

[00:18:24] Speaker 01:
indicating a narrower meaning to data that functions to identify.

[00:18:31] Speaker 01:
I just don't see what precludes the broader understanding that the appellant is arguing for.

[00:18:39] Speaker 02:
Because it says the data that identifies the controllable function, not some other data.

[00:18:47] Speaker 02:
If you look at Roku's construction, they want to read it to mean data that can be used in connection with other information.

[00:18:54] Speaker 02:
Well, the claims don't say that.

[00:18:56] Speaker 01:
I guess that's a better way to ask the question.

[00:18:59] Speaker 01:
Why couldn't these later steps be performed if at the receiving step it was the data from the controllable appliance plus from some other source?

[00:19:14] Speaker 02:
Well, some other data, and I'll show you in the specification, Your Honor, exactly where, some other data

[00:19:20] Speaker 02:
could come into play in these subsequent steps, but not in the receiving steps.

[00:19:25] Speaker 02:
If you look at the figure that the parties had all this debate about during the IPR, Figure 15, the specification.

[00:19:32] Speaker 01:
I do want to get to the specification, but you're starting appropriately with the claim language.

[00:19:36] Speaker 01:
I've missed the logical step in the board's writing, and I'm missing it as you're articulating it now.

[00:19:44] Speaker 01:
how it is you're seeing the automatically adding and the causing step are somehow supporting the narrow construction at the receiving step?

[00:19:54] Speaker 02:
Well, it supports it for two reasons.

[00:19:56] Speaker 02:
One, based on the language, that was identified by the data received from the controllable appliance.

[00:20:04] Speaker 02:
Those words are in the automatically adding and causing step.

[00:20:10] Speaker 02:
The antecedent basis for that is the claim element at issue in the receiving step.

[00:20:15] Speaker 02:
It's saying the data, the data itself received from the controllable appliance is what identifies the controllable function, not some other information.

[00:20:26] Speaker 00:
That data alone identifies, and it doesn't say that data directly identifies.

[00:20:34] Speaker 00:
It doesn't say that.

[00:20:35] Speaker 00:
So I agree with you that it removes that functions to identify, right?

[00:20:42] Speaker 00:
I mean, that's the most work it does.

[00:20:44] Speaker 00:
But even so, it removes functions to.

[00:20:47] Speaker 00:
Still, the phrase data that identifies a controllable function is broad.

[00:20:52] Speaker 00:
How do you respond to that?

[00:20:54] Speaker 02:
The way I respond to it, Your Honor, is that I don't believe it is broad, especially when you look at the specification, which is part of the intrinsic record, which informs the claim language.

[00:21:04] Speaker 02:
It's not broad.

[00:21:06] Speaker 02:
If Your Honors will permit me, let's look at figure 15.

[00:21:10] Speaker 04:
Can I ask one more language question?

[00:21:16] Speaker 04:
The board, well,

[00:21:25] Speaker 04:
without the term function, that if the claim simply said data that identifies a controllable function, that too would permit use of other information based on that data to identify the controllable function.

[00:21:51] Speaker 04:
Did Roku ever concede that

[00:21:53] Speaker 04:
Even if the words functions to were not there, that the analysis would be any different.

[00:22:00] Speaker 02:
They did not.

[00:22:01] Speaker 02:
But Your Honor, it's important.

[00:22:03] Speaker 04:
If I remember right, I don't know whether you were at the board.

[00:22:08] Speaker 04:
I was not.

[00:22:08] Speaker 04:
Oh, I think in the transcript of the board, the APJ asked that question.

[00:22:14] Speaker 04:
And I can quite remember what the answer was.

[00:22:20] Speaker 04:
any disparity in usage between the receiving step and the adding and causing steps would disappear?

[00:22:29] Speaker 04:
Each one of them would permit the home theater appliance and device to do the identifying based on the data.

[00:22:40] Speaker 02:
Well, the home theater device doesn't do the identifying.

[00:22:43] Speaker 02:
The claim language says

[00:22:45] Speaker 02:
the claim language says receiving at the home theater device from a controllable appliance in communication with the home theater device via use of HDMI connection data that functions to identify.

[00:23:02] Speaker 02:
So that data comes from the controllable appliance and that's important because it was

[00:23:10] Speaker 02:
those amendments to the causing and automatically adding steps during prosecution that help the board reach the conclusion that the data itself that comes from the controllable client has to do the task and not in combination with some other data.

[00:23:27] Speaker 02:
And the specification supports that.

[00:23:29] Speaker 04:
OK.

[00:23:31] Speaker 02:
So there's a nice reproduction of it on appendix page five of the board's decision.

[00:23:41] Speaker 02:
And if you look at the top, you have reference numerals 1502, 1504, 1506, 1508.

[00:23:50] Speaker 02:
And let's stop there for a moment.

[00:23:55] Speaker 02:
And 1508 says, icon metadata available from the appliance.

[00:24:00] Speaker 02:
That's a query.

[00:24:01] Speaker 02:
That's a question.

[00:24:02] Speaker 02:
Now, the parties agreed below

[00:24:06] Speaker 02:
that the term icon metadata or icon information, as the written description calls it, is an example of the data that functions to identify in the claims.

[00:24:18] Speaker 02:
So it wants to know at step 1508, is this identifying data available from the appliance?

[00:24:29] Speaker 02:
If it is, it goes directly down to where it says send to the smart device.

[00:24:37] Speaker 02:
That's the only part of this figure that asks the question, does the data that identifies the controllable function come from the appliance as the claims require?

[00:24:49] Speaker 02:
Roku wants to rely on another part of this figure where the answer to that question at 1508 is no.

[00:25:00] Speaker 02:
So if the data cannot and does not come from the controllable appliance, what happens?

[00:25:07] Speaker 02:
It goes through this other route until it gets to 1518, where it says, obtain ICON metadata for the appliance, not from the appliance.

[00:25:21] Speaker 02:
And then what happens?

[00:25:22] Speaker 02:
This ICON metadata is retrieved from an external server 207.

[00:25:29] Speaker 02:
It doesn't come from the appliance.

[00:25:32] Speaker 02:
It comes from a database that has information about older devices.

[00:25:37] Speaker 02:
So what claim one covers is the situation where the query is made and data that identifies a controllable function comes from the controllable appliance itself.

[00:25:52] Speaker 02:
And the board, in its decisions, noted those words as being important.

[00:26:00] Speaker 02:
Because during prosecution, the causing and the automatically adding steps

[00:26:08] Speaker 02:
were amended to clearly point out that the data that identifies must come from the controllable appliance.

[00:26:15] Speaker 01:
And if you look back to the figure, it seemed to me like through this flow, you can practice claim one using data in addition to data coming from the appliance, which is exactly what would be captured additionally by their construction, but is not captured by your narrower construction.

[00:26:38] Speaker 01:
So what am I missing?

[00:26:39] Speaker 02:
But Your Honor, the claim requires that the data that does the identifying, that identifies

[00:26:44] Speaker 02:
control book function must come from the controllable appliance.

[00:26:49] Speaker 01:
You say it must, it must alone, but I don't see how this figure helps you get there.

[00:26:55] Speaker 02:
Because that, as we call it in our brief, that branch of it that goes to the right to 1508, 1508 asked the question, is there icon data, icon metadata,

[00:27:07] Speaker 01:
What evidence do we have that you don't look at the other branch in understanding this claim language?

[00:27:14] Speaker 02:
Because the other branch, the evidence that we have that shows that you don't look at the other branch is because there's nothing in the other branch where ICON information or ICON metadata comes from the appliance itself.

[00:27:28] Speaker 02:
It doesn't.

[00:27:29] Speaker 02:
It says, at 1518, it says, obtain ICON metadata for the appliance, which is stored in some external database, which is not the controllable appliance.

[00:27:40] Speaker 02:
And if you look at the specific citations to the specification that Roku relies on, and if we go to appendix page 63, column 14,

[00:28:03] Speaker 02:
beginning at line 35 with the words, if the icon information is available, meaning if the data that identifies the controllable function is available, what happens next?

[00:28:17] Speaker 02:
And it says, the icon information may be sent to the smart device by the appliance and or retrieved by the smart device

[00:28:25] Speaker 02:
using other information provided by the appliance as appropriate.

[00:28:29] Speaker 02:
That's what they're relying on, the words other information, but the identification of the controllable function has happened already because it says ICON information.

[00:28:37] Speaker 02:
you're talking about transferring that information somewhere else.

[00:28:41] Speaker 02:
You can use other data to do that.

[00:28:42] Speaker 01:
So you read that as excluding a situation where other information is used in addition to the information coming from the appliance itself.

[00:28:54] Speaker 02:
Sure, because there is nothing in this specification and in the cited portions of it that Roku relies on that says

[00:29:05] Speaker 02:
If the icon information is coming from the controllable appliance, which is what the claim is, other information can be used with that to identify the controllable function.

[00:29:15] Speaker 02:
It doesn't say that in the specification.

[00:29:18] Speaker 02:
But it doesn't say you can't do that.

[00:29:19] Speaker 02:
Would you acknowledge that?

[00:29:20] Speaker 02:
There's no embodiment of it that says you can do that.

[00:29:28] Speaker 02:
I want to address, in the two minutes I have, this issue about Claim 9 and Roku's expert supposedly conceding something.

[00:29:38] Speaker 02:
And I think the page of the appendix that's important here is counsel cited A-31-38 and 39.

[00:29:50] Speaker 02:
And so the expert, UI's expert, is being asked about Claim 9.

[00:30:00] Speaker 02:
And when he's talking about claim nine, your honors can read it as well as I can.

[00:30:05] Speaker 02:
Yes, sir.

[00:30:06] Speaker 02:
It's appendix page 3138 over to 3139.

[00:30:09] Speaker 02:
Here's the question and continues on to 3139.

[00:30:12] Speaker 02:
So when we get there, UEI's expert is talking about what this

[00:30:26] Speaker 02:
They're not talking about the data that identifies limitation.

[00:30:30] Speaker 02:
He's talking about what could this media rendering function be.

[00:30:38] Speaker 02:
He's not talking about what is encompassed by the data that functions to identify limitation.

[00:30:48] Speaker 02:
And in addition, Roku has argued that claim nine covers state data.

[00:30:53] Speaker 02:
But during prosecution,

[00:30:56] Speaker 02:
the patentee did distinguish the Sakai reference based on its use of state data, and because this state data did not come from the controllable appliance, distinguished it on two grounds.

[00:31:10] Speaker 02:
So if their position is claim nine covers state data, that's really far afield to what we're talking about here, because not only did the patentee distinguish prior art based on state data, when the board took the claim construction and analyzed

[00:31:26] Speaker 02:
the references cited by Roku, they said, well, here, look, the HDMI 1.3a prior art, it has state data.

[00:31:35] Speaker 02:
That can't identify a controllable function.

[00:31:39] Speaker 02:
So the prosecution history is important.

[00:31:44] Speaker 02:
And if I can point in my... Could I just ask you before?

[00:31:47] Speaker 01:
3133.

[00:31:48] Speaker 01:
Yes.

[00:31:48] Speaker 01:
You're an expert, Dr. Turnbull.

[00:31:53] Speaker 01:
It seems to me like he's agreeing that the identification doesn't have to be done just from a controllable device.

[00:32:00] Speaker 01:
The question at line 15, so data that functions to identify a controllable function of the controllable device has to describe the icon itself that's going to be added to the user interface?

[00:32:12] Speaker 01:
answer from your expert, no.

[00:32:15] Speaker 01:
It just needs to be able to help identify that or identify that icon.

[00:32:21] Speaker 01:
Is that not a problem?

[00:32:22] Speaker 02:
I don't think it's a problem at all, Your Honor.

[00:32:24] Speaker 02:
I'll tell you why.

[00:32:25] Speaker 02:
It's because the question asked about the icon.

[00:32:29] Speaker 01:
The icon comes in in the next step.

[00:32:31] Speaker 01:
But the icon, I thought you already told us, is agreed in connection with Figure 15, at least, to be an example

[00:32:39] Speaker 01:
of the claim term that we're talking about.

[00:32:41] Speaker 02:
Icon information, the parties agreed, is an example of the data that functions to identify.

[00:32:47] Speaker 02:
What he's asking about here is the way the question was framed is what you do to automatically add the icon to the user interface.

[00:33:00] Speaker 02:
Look at the top.

[00:33:01] Speaker 02:
This is the question they left out.

[00:33:04] Speaker 02:
Page 76, lines four to seven.

[00:33:07] Speaker 02:
question, right?

[00:33:08] Speaker 02:
So on its face, the data has to identify a controllable function of the controllable clients, correct?

[00:33:14] Speaker 02:
And it says, yes.

[00:33:16] Speaker 01:
And then, but it doesn't have, what I'm driving at is, it doesn't have to specifically identify or to describe the icon itself.

[00:33:24] Speaker 01:
Answer, I think, it has to be associated with the icon.

[00:33:27] Speaker 02:
Well, the specification would certainly not enable an embodiment that covers

[00:33:35] Speaker 02:
Roku's claim construction because the entirety of column 14 and 15 of the written description and figure 15 show that the only time that data that identifies a controllable function is used in the claims comes from the controllable device, from the controllable appliance, not from some remote server.

[00:34:00] Speaker 02:
It just doesn't.

[00:34:01] Speaker 02:
And I know I'm over my time, but I just want to point

[00:34:04] Speaker 02:
your honors, to the prosecution history.

[00:34:10] Speaker 02:
That one part of the prosecution history that counsel was talking about, it's appendix page 864.

[00:34:25] Speaker 02:
And yes, this is a long paragraph, but one of the things that

[00:34:31] Speaker 02:
the patentee argued at the bottom was, in short, it is respectfully submitted that the state data that is used in the system of Sakai is not and would not be considered by one of skill and the art to be the claim data indicative of a controllable function of a controllable appliance.

[00:34:47] Speaker 02:
So if their position, which it appears to be in the brief, is that claim nine covers state data, the patentee said it doesn't,

[00:34:56] Speaker 02:
in the prosecution history and the board said state data cannot identify a controllable function.

[00:35:04] Speaker 02:
Thank you.

[00:35:17] Speaker 03:
Thank you, Your Honor.

[00:35:18] Speaker 03:
I'll try to keep this brief.

[00:35:20] Speaker 03:
Judge Stark, I agree with your reading of their expert deposition testimony.

[00:35:24] Speaker 03:
He did concede the claim nine covers the embodiment that Chardon discloses, that is Chardon state data, even to the extent that the applicant distinguished

[00:35:34] Speaker 03:
distinguished Sakai state data during the file history as I've explained and as we set forth in our briefing that is different.

[00:35:42] Speaker 03:
It is fundamentally different from Chardon state data and my colleague on the other side has not addressed that distinction.

[00:35:49] Speaker 03:
With regard to the file history I think

[00:35:52] Speaker 03:
to the extent that we can glean anything from the file history it's that the phrase data that functions to identify was already in the claims when the claims were rejected as anticipated and if the applicant truly had intended to narrow the scope of that term to cover data that itself identify it could have done so expressly and that's not what it did it added other claim language to other limitations and I think the board's

[00:36:17] Speaker 03:
conclusion that the applicant intended to narrow the disputed phrase by adding language to other phrases simply is incorrect.

[00:36:28] Speaker 03:
I also heard from my colleague on the other side that the home theater device is not the entity that identifies the controllable function.

[00:36:36] Speaker 03:
That's wrong.

[00:36:37] Speaker 03:
Again, we've pointed out in their briefing at page 38, as well as in the record below, they conceded that it is indeed the home theater device that identifies the controllable function using the claim data.

[00:36:48] Speaker 04:
Can I come back to one language point?

[00:36:51] Speaker 04:
Yes, you can.

[00:36:52] Speaker 04:
I think in our earlier discussion, you said that you are, in fact, reading or receiving

[00:37:01] Speaker 04:
mean the same thing as a rewritten clause connection data that the home feeder device uses to identify.

[00:37:10] Speaker 04:
And if it means that, then isn't there some tension, I think as the board was suggesting, between the different phrases in the automatically adding and causing, because there it says this function was identified

[00:37:27] Speaker 04:
by the data, not by the home theater device using the data?

[00:37:33] Speaker 03:
No, Your Honor, I don't think there is tension, because those downstream limitations, of course, refer back up to the antecedent limitation.

[00:37:41] Speaker 03:
And they cannot and do not change the nature of the data itself or what it conveys.

[00:37:45] Speaker 03:
Rather, they merely clarify that the controllable function was identified in the antecedent step, however it was identified.

[00:37:53] Speaker 03:
And the board simply over read those downstream limitations to suggest that they somehow narrow the scope of the antecedent limitation.

[00:38:02] Speaker 03:
That's inconsistent with basic antecedent basis law.

[00:38:06] Speaker 03:
But to your honor's point about whether

[00:38:09] Speaker 03:
we would be in a different situation if the claim had read data that identifies.

[00:38:13] Speaker 03:
I think the answer would be no.

[00:38:15] Speaker 03:
And I think that's what we said below at the oral hearing, as Your Honor noted.

[00:38:20] Speaker 03:
Data that identifies is still sufficiently broad.

[00:38:24] Speaker 03:
It doesn't preclude data that is used to identify.

[00:38:28] Speaker 03:
But even setting that aside, the claim language that actually appears

[00:38:31] Speaker 03:
is unambiguously broader than data that identifies.

[00:38:36] Speaker 03:
And the board's construction is unambiguously narrower.

[00:38:39] Speaker 03:
I could say a word about the specification.

[00:38:42] Speaker 03:
I see my time is up, but I'm happy to answer questions if there are any.