Oral argument — KEYnetik, Inc. v. Samsung Electronics Co., Ltd.

The United States Court of Appeals for the Federal Circuit is now open and in session. God save the United States and this Honorable Court. Our first case this morning is number 20-1270, Keynetik, Inc. v. Samsung Electronics. Mr. Beam. Thank you and good morning, Your Honors. This is Edward Beam, arguing on behalf of Appellant Keynetik, Incorporated. We are appealing the Board's use of the broadest possible interpretation in erroneously construing the sleep command limitation to not be logically correlated to the steps of the step-motion algorithm and the erroneous construction of the term sleep by allowing motion sensed during sleep to be executed upon reactivation of the step-motion algorithm. The Board interpreted the claims devoid of any context of what the patent teaches and, indeed, the Board's construction is inconsistent with the intrinsic record. Although this Court has held that the correct inquiry is, quote, not whether the specification prescribes or precludes some broad reading of the claim, that is precisely what the Board did here by adopting Samsung's argument that, quote, nothing in the claims or the specification precludes using data sensed during the sleep period in commands that are executed after the system is reactivated. The claim invention includes a step-motion algorithm which requires that the algorithmic steps of Claim 22 to be logically and conditionally related to each other as a set of instructions in accord with the specification and the intrinsic record. What do you mean by logical relationship? Do you mean by that that the motion instruction has to bring about the sleep period after it's executed? Yes. The step-motion algorithm is an algorithm which is a process of rules that are followed by a computing system to problem solve different operations. Here, the step-motion algorithm solves the problem. All the term algorithm requires is that steps be executed in order. It doesn't, in and of itself, require a logical relationship, right? As part of that logical relationship, the specification also conditions- No, no. Wait, try to answer my question. Sure. In terms of plain and ordinary meaning, algorithm just means the steps are performed in order. The term algorithm itself doesn't require a logical relationship. Well, I would say that they must be performed in order. There is a logical relationship that the algorithm is performed in a step-by-step procedure to solve for a problem or accomplish some sort of end. Yes, they must be performed. A facilitator would understand that the algorithm accepts an algorithm would be followed in the recited order. Otherwise, it would not solve the entire problem. They have to be done in the recited order. But you're saying that there has to be something more here besides being performed in order. You say that there has to be a logical relationship. But the meaning of algorithm only requires that they could be performed in order, right? Well, they are performed in order, but we also are saying that they are conditioned upon each other, that one must follow the other. Are you saying it has to be immediate, though? That it has to follow it immediately? Yes, that it's followed after the- For example, the execution, the sleep command is conditioned upon the execution command. If we refer to our blue brief at page 29, figure 12 of the 146, it shows the stepwise algorithm. It's the simple step motion algorithm implemented on top of the motion input algorithm. And the execution of step 1204, then a sleep command is sent to a motion input algorithm at 1206. So that is sent is an affirmative statement. And so it conditions the- Well, the board agreed that the one had to cause the other, but the board said that there could be intervening acts between the cause and the result. Right? They tried to go down that road that there could be intervening acts, and certainly there could be things added within the algorithm because of the comprising language. But you can't render the existing claims a nullity. The sleep command must follow the execution command. Can I ask you a question? Are we talking about sleep command implementation or the after the command is executed? Judge Clevenger, you're breaking up a bit. I'm sorry. Can you hear me now? Try that. Yes. Are you talking about the sleep command limitation or are you talking about the after the command is executed limitation? Or are you mixing them up? The sleep command limitation is step two of the step motion algorithm, which includes the after command is executed language that you're referring to. So step two is instructions to send a sleep command to the motion input algorithm after the command is executed. And that command is executed in step one of the algorithm. Yes, I understand. But I understood your argument about the after not being limited solely to temporal, but it also has a conditional requirement. Correct. And that's what I understood you to be talking about in your logical relationship, whether or not the algorithm has to be performed in order. Well, I think it's logical and conditional because the way the board has construed it, after can mean any time after. So there's no sort of, it's temporal only and no end in time. So indeed, under their construction, it still may not have happened. Here, we're saying that not only is it logically correlated, but it's conditioned on the execution of the command from step one. And as that is done, the sleep command, instructions to send a sleep command are executed. I don't understand the board to be saying that you could have the sleep command take place before the motion command is executed. Are you suggesting that? No, I'm not suggesting that it's before. Certainly not. I'm just saying that the instructions to send the sleep command must happen after the execution of the command from step one. Does the board agree with that? That the sleep command has to happen afterwards, temporarily? Well, temporarily I think is the broadest possible interpretation of the term after because they're providing such a, again, allowing the after to occur at any possible time later, an undefined time. Here, we're saying it's conditioned. What in the specification would lead us to the conclusion that there can't be a great gap of time or a lot of other activity that occurs between the command and the sleep function actually occurring? Well, again, I'd refer you back to Figure 12, which is an affirmative use of the isSense language, demonstrating the stepwise function of the algorithm. Certainly, the purpose of the invention is to block returning motion. Hello? Oh, I'm sorry. And the delaying of the execution is not consistent with the sleep command or the specification. That means the specification explains that the invention must allow the system to readjust without cursor moving on the screen by ignoring the motion that is returning, which means that the sleep command must be executed after the execution of the command from the motion input algorithm of step one. But what does that have to do with logical relationship? That seems to me to be an argument that after means immediately after or almost immediately after. Okay. Yeah, if it's not conditional, it defies the purpose of the claim to ignore returning motion. The fact is that upon execution of the command from step one, it must then go into a sleep command. I mean, that's the conditional relationship. If we don't follow that conditional relationship... But are you asking us to import the word immediately into the claim? Well, the specification teaches that in order for a returning motion to be ignored, the sleep command must conditionally follow the execution. I don't know necessarily if we need to import anything. I think the specification clearly teaches that, and there's nothing in the specification that teaches, frankly, away from that, that teaches otherwise. Well, this is Judge Clevenger. What's wrong? I'm looking at the claim for the solution in appendix 10 to 11, and they say that the after means that there's a temporal relationship between the executed command and the sleep command that requires the executed command to occur, and then at some later point in time, the command is issued. So there is a conditional relationship. It's just that there can be a long period of time. So... Yeah. I don't understand, but with all due respect, I do not understand what you mean by a conditional or a causal relationship, because the interpretation says that the command requires the executed command first to occur, so that the second command is conditioned on the occurrence of the previous event. Well, the prior R that is cited here is the condition on the... I'm not interested in prior R. I'm interested in what you're telling me is supposed to happen that isn't covered by the claim construction. Isn't there a conditional relationship between the executed command and the sleep command? Yes, and there must be, because, again, But I think within the definition that the board gave us, what more do you want? Well, the board gave us only a temporal relationship, which means it could happen at any time, meaning that the returning motion that the invention is seeking to block would otherwise be sensed, and commands would be issued thereon. The sleep command has to issue after the execution of the command from step one. There cannot be this sort of lag in time between step one and step two, otherwise returning motions... Wait, wait. Now you're saying that there is a relationship, but it allows a lag in time. The presiding judge was asking earlier, as well as Judge O'Malley, what's wrong with the interpretation that says there is a point in time between the executed command and the sleep command? Why does the existence of a period of time negate a condition, a causal push? Because the sleep command is triggered by the execution of the command from step one. Let's assume it's triggered immediately, or it's triggered at an undetermined time. It's still triggered by, and therefore a condition won't work. But if it's not triggered immediately, if it doesn't immediately follow, then the returning motions that are sought to be ignored by the invention, as taught by the specification as the purpose of the invention, would be not ignored. It would go against the teaching of the specification. Right. So your point is that if it's construed the way the board construed it, then it totally negates the entire point of the invention, right? Absolutely, yes, Your Honor. So what you're saying is that after ought to be construed to mean immediately after. Yes, Your Honor. And again, there's nothing in the specification... ...to be ignored? We did not ask that the immediate... We did say that it should be conditionally... argue that it's conditionally related, and that it should be, you know, follow upon the execution. But immediately or follow, as the board interpreted it. I mean, I understand you to have admitted that the board's definition has a conditional element. It's just that it has an open-ended time for performing the condition. No, I believe that it does not have a conditional element, that the board's construction is only temporal, and that we have always argued that there is a conditional requirement as well. I mean, there's a logical... Well, you can read the interpretation at the bottom of page 10 going to 11 to say that it indicates a temporal relationship between these two events, meaning that time is open-ended. I'm sorry. I don't know if I followed your question. The board's interpretation, as I read it, says that they recognize that there is a temporal relationship between an executed command and a sleep command that requires the one to occur before the other. It's temporal in nature in that it's open-ended, but it's still conditional, because the relationship between the two is conditional. I would argue that if it is indeed conditional, then one would follow the other. But why would it necessarily follow immediately? Well, it would follow immediately. The algorithm, as stated, there is no room for interpretation by the computer device that's running it. If the instructions to execute the command are sent, then the executions to send the sleep command follow because of the executing of the command from step one. There is no intervening time. It is conditioned upon the... I'm sorry. You broke up, sir. Yes. Figure 12. Figure 12. Right. But when you look into this, when you read the specification and the purpose of the invention is specifically to ignore returning motion, and the only way to ignore returning motion is to immediately have step two send instructions for the sleep command after step one is executed. If there is any sort of time delay in there, again, undefined time delay, not even described in the specification, that would allow returning motion to not be ignored. I mean, it would go completely against the purpose of the invention. Well, that makes sense. But what it seems to turn on is not whether it's temporal or logical, in your view, but on whether the thing has to happen immediately or soon thereafter. And what I'm asking is you didn't tell the board that their construction had to mean immediately. Well, no. We said it could not be a time difference. We said it couldn't be days, weeks. I mean, the time here is completely undefined. And again, the specification speaks to the fact that if there is any gap in time. But a temporal construction doesn't necessarily mean that it can be forever. A temporal construction could mean temporally, but immediately, or temporally, or soon. But the specification, reading specification, tells one of skills in the art that it's immediate, because you have to ignore returning motion. I mean, the motion, the device is always in motion. There are calculations flying in. Any sort of time gap between step one and step two is going to allow that returning motion to come in and, frankly, corrupt the system. Fine, but it doesn't depend on whether it's a logical relationship or a temporal relationship. The construction you're urging is that it should be immediate. But as I understand it, you didn't argue that to the board. You said instead that it couldn't be temporal and had to be logical, whatever that means. Well, it's temporal, logical because it's an algorithm, temporal because it happens after, but it's also conditioned upon. And that is the key thing we're arguing, that once the instructions to execute the command from step one come, step two happens immediately thereafter because it's conditioned upon receiving that command from step one. And, again, it's an algorithm that there is no space to make a decision otherwise. It's a stepwise function. Step one happens, and step two happens as a response to step one. Okay. Well, unless my panel members have other questions for Mr. Beam, I think we'll hear from Mr. Citron. Are there any further questions? Can I just ask one thing? Perhaps it would help Mr. Beam in the time that he has. Looking at the patent owner's short reply to the board on this issue, sir, Mr. Beam, and I noticed that on page two of your short reply, you did complain that the petitioner's view provided that the sleep command need only be sent at some indefinite time after the command is executed. And I wonder if you further developed that argument before the board. So what I'm citing in your short reply at least is some indication that you were worried about the length of time after the command is executed. This goes to Judge Dyke's question about whether, the presiding judge's question about whether you raised this amount of time issue before the board. That's my only point. Okay. Mr. Beam, we'll give you two minutes for rebuttal. Thank you, Your Honor. Mr. Citron? Yes, good morning, Your Honors. Can you hear me? Yes, we can hear you. Okay. May it please the court, my name is Philip Citron, and I'm arguing on behalf of Apelli Samsung. The argument today focuses on whether the board erred by not construing the claim 22 so that the, quote, instructions to send the sleep command are conditioned upon, whereas noted today must occur immediately after, the prior instructions to execute a command from the motion input algorithm. As the board explained at 8.10 to 11, neither the claims nor the specifications support such a narrow construction. The board's construction here is... Doesn't the specification, I mean, the language doesn't say may be sent at some future point in time. The language says that it is sent, right? That's correct, Your Honor. And that's what figure 12 says. That's correct, Your Honor. But the language that's used in the specification, which is very similar to the term after that's used in the claim, is that it's sent following execution of the motion input command. So looking at the claims and that explanation in the specification, that term doesn't imply or require some sort of conditional relationship. It implies some temporal relationship, which is the same conclusion... But the broadest reasonable construction still has to be a logical construction in light of the specification as a whole, doesn't it? Yes, Your Honor. That's absolutely correct. And we believe the board's construction here is consistent with the specification and is a reasonable interpretation based on both the claims and the specification. Wouldn't it completely negate the whole point of the invention? We don't believe it would, Your Honor. And what Kinetic attempted to do in its briefing is focus on really one sentence in the specification, which states that the forces caused by returning motion must be ignored. But there's a few issues with that argument that, as an initial matter, as Your Honors know, the claims themselves... I'm totally confused. Look, as Judge O'Malley points out, this invention, even if you regard it as a temporal limitation, suggests that the sleep period has to happen soon thereafter. It can't be a matter of days or weeks. That makes no sense in terms of the invention. So even if you view it as a temporal limitation, there has to be some notion in the temporal limitation that it's happening pretty soon thereafter. And what I was trying to get at with Mr. Beam is whether that construction, which seems to make some sense, was one that Kinetic asked the board to make. I mean, do you agree that the temporal limitation... ...is a short period of time that we're talking about rather than weeks, right? Your Honor, I don't think that the claims or the specification necessarily require that. And here's why. There is a discussion of what step motion means in columns 13 and 14 of 146 patent, which is at Appendix 55. And it defines it simply as allowing readjustment of the system between steps without using sliding zero code, which is a prior art algorithm that's used. It then goes on to explain that using permissive language, that you can allow forces caused by returning motion to be ignored. And that's at the bottom of column 13, lines 63 to 65. In column 14, it then goes on to describe, and this is at line 3, the simple step motion algorithm has an additional feature based on the introduction of an intensity timeout, also known as a sleep period. And if you go further down... But you're not answering the question. I mean, the question is, I mean, it makes sense that the temporal period of time, if that is the construction, has to be a very short period of time. And I don't understand how you can really dispute that. The question is, was that requirement one that was argued to the board? We had a case recently, an Uber case, where the language was responsive to, and again, involved what the period of time was. We looked at the specification to determine what the period of time was. And so here, it seems to me to make sense to say, if you're going to have a temporal limitation, which is what it is, you look at the specification to see how long that period can be, and here it has to be pretty short. And I just don't understand how you can argue otherwise. The question is whether that construction was presented to the board. Sure, Your Honor. So I don't believe that precise construction was argued before the board. That is, does it need to occur immediately after a prior motion command? That was not presented before the board. I don't believe that term was even used before the board. And one point I think I'd like to make that is important here is that, even if a short period of time is required, or even if it must be immediate, which is what Kinetic is arguing today, this limitation would still be disclosed by the combination of Neguero and Liberty, which we explain in our briefing, because in that combination, the sleep command does occur upon a current movement of the cursor and looks at the velocity, how fast it's moving. But that would require some fact-finding that we shouldn't be doing in the first instance. If we disagree with the construction, shouldn't there be a remand? So for this particular factual issue, the board did actually make these factual findings. That's in Appendix 21 to 22. And they cite to the teachings of Liberty itself and also to the expert testimony provided by Samsung's experts. So I don't think for this particular factual issue a remand would be necessary. I think the record is complete, and the court can make a determination based on the record. Let me back up to the claim construction. So you would agree with your friend on the other side, who argued to the board, that your proposed construction, which put no temporal limit, left it wide open, that it could be years even between the execution and when the command is adopted, right? Hey, the claims are broad enough to encompass a longer period of time, and the specification is open about the types of movements. It says that you can change the orientation any way the user wants. So there is some vagueness in the specification about how long the time period would need to be, depending on that needs to be excluded. But in the construction, there is no time limit. So it could be years, right? Correct, Your Honor, that is correct. There is no limitation in the board's construction. It merely requires that it recur after the prior movement command is executed. That's correct. One additional point I'd like to go back to, and this is going back to what the purpose is, is that, again, the idea of ignoring motion, which is the sentence that Kinetic focuses on, that is one feature that's described in the specification, but it's not the, quote, additional feature of the sleep period, which is what is actually claimed in that feature. All that happens is that the input algorithm is paused, and that need only occurs at some point after. So I just want to make that one last point there. So at the end of the day, I think the board's construction is not only trying to preclude other types of constructions. I think it is consistent with the claims and the specification here. So with that, Your Honor, unless there are any additional questions, I will give back my time. I haven't got any further. Okay, hearing no further questions, we'll hear from Mr. Beam for two minutes. Thank you, Your Honors. And to address Judge Kleppinger's comment, the appellant did address the fact that there's no time limitations in their blue brief at the bottom of page 30, moving into 31. But I believe, you know, to address Judge O'Malley's comments about, you know, the period of time, you know, I think the specification clearly teaches that the time has to be short enough to ignore the returning motion. You know, we make clear, the patentee makes clear in the specification that the simple step motion algorithm requires... That makes sense, but did you ask the board to do that? Did you ask for that construction? Yes, we asked that it would, that the steps be conditioned on one another. Well, condition doesn't limit the time period. Condition could still be weak. You know, that doesn't get you anywhere. Mr. Beam? Yes. Following up on the presiding judge's question, in your patent office, your patent owner's initial response to the petition, you made the argument that the steps needed to be performed as you said, because otherwise you'd be allowing forces caused by the returning motion to be ignored. So I see that that argument was being made, and that argument relates to the period of time, doesn't it? I'm trying to get at whether or not you gave enough information to the board to let it understand that you were arguing against this unreasonably long period of time. Yes, thank you. And we definitively were, and in part because we were arguing against what the board construed as this belay period, right, where you could take in motion over a period of time and use it. Without ignoring the returning motion, immediately following the motion command, the motion is not ignored, and the teaching of the specification would be corrupted. And I would also say there's no teaching that the return motion is not ignored. I don't have in front of me at the moment of all the briefing that was below, but I'm looking at both the patent owner's response, as well as as I said earlier, sir, reply on the issue. And you certainly were complaining that SAMSUNG's interpretation left an indeterminate amount of time after, before the command is executed. And you had been arguing that you were certainly wanted to avoid the allowing the forces caused by the returning motions to be ignored. Yeah, absolutely. I think the open ended period of time makes the construction that the, you know, board, you know, settled on, you know, not applicable to the claim to mention. All right. Unless there are further questions. Hearing none. Thank you, Mr. Bean. Thank both council and cases submit. Thank you. Your honors.