VL Collective IP, LLC v. Unified Patents, LLC, No. 24-1890 (Fed. Cir. Feb. 20, 2026)

CHEN, Circuit Judge.

Appellant VideoLabs Collective IP, LLC (VideoLabs) appeals the Patent Trial and Appeal Board’s (Board’s) Final Written Decision in No. IPR2022-01086 finding claims 1, 3, 5, 9, 12–13, and 15 of U.S. Patent No. 8,605,794 (’794 patent) unpatentable under 35 U.S.C. § 102 and 35 U.S.C. § 103 in view of Sonohara.¹ Unified Pats., LLC v. VL Collective IP LLC, No. IPR2022-01086, 2023 WL 8717013, at *20 (P.T.A.B. Dec. 18, 2023) (Decision). VideoLabs challenges the Board’s claim construction, the Board’s allowance of a first-time Reply argument, and the Board’s on the merits finding that the claims are unpatentable. Because we agree with the Board’s analysis, we affirm.

BACKGROUND

The ’794 patent is directed to synchronizing data segments consisting of audio and visual content. The invention discloses synchronization of “content-related . . . data segments” “by way of the predefinable assignment rule,” which assigns a first data segment to a second data segment and enables the segments to be output chronologically. ’794 patent col. 2 ll. 36–43, 46–52.

Independent claim 1 is representative of the challenged claims and recites:

A method for synchronizing content-related first data segments of a first data file and content-

3related second data segments of a second data file, the method comprising:

sequentially outputting, by a device for synchronizing content-related data, the content-related first data segments and the content-related second data segments according to their chronological sequence in such a way that each of the content-related second data segments is output together with an associated one of the content-related first data segments on the basis of an assignment rule for assigning each one of the content-related second data segments to one of the content-related first data segments.

Id. at col. 7 ll. 45–57 (emphases added). Dependent claims 20 and 21 include an additional limitation to independent claims 1 and 9, respectively, that “the assignment rule is not based on a timestamp.” Id. at col. 9 ll. 1–4.

On June 7, 2022, Unified Patents, LLC (Unified) filed a petition for inter partes review arguing that claims 1, 3, 5, 9, 12–13, 15, and 20–21 of the ’794 patent were unpatentable in view of Sonohara. J.A. 84–85.

Sonohara teaches a “motion picture reproducing apparatus” which “synchronizes the image data and the sound data” of a file. Sonohara, at Abstract. It discloses a composite motion-picture file generated by combining an image file and sound file. Id. at col. 4 ll. 12–20; see also id. FIG. 4. The composite file contains a header, an image “track” of segmented image data, and a sound “track” of segmented sound data. Id. at col. 1 ll. 51–58; see also id. FIG. 1. When the image data and sound data are processed and given track numbers, representing the order of the data within the tracks, “the sound data and the image data are coupled4 with each other.” Id. at col. 4 ll. 45–48; id. at col. 6 ll. 40– 46.

The petition mapped Sonohara’s segmented image and sound data in the composite file to the ’794 patent’s content-related first and second data segments. The petition specified that the segments were “of a [first and] second data file”—as required in claim 1’s preamble—because the image and sound data in the composite file originated from a separate image and sound file. J.A. 121.

On December 19, 2022, the Board instituted inter partes review. J.A. 1046. At that time, the Board noted that the petitioner had not submitted any claim constructions beyond the plain and ordinary meanings of the claim terms. Id. at 1054. Though VideoLabs had proposed certain constructions in its Preliminary Response, the Board “[did] not find it necessary to construe any term.” Id. at 983, 1054.

In its Patent Owner Response, VideoLabs advocated for a negative limitation of the term “assignment rule” based on instances in the prosecution history where the applicant distinguished the ’794 patent from the prior art. Decision, 2023 WL 8717013, at *8 (citing J.A. 1166–67). Specifically, VideoLabs argued to the Board that the Applicant had disclaimed “assignment rules in which the content-related data segments are assigned to one another using exact timing information.” Id. Thus, VideoLabs proposed that “assignment rule” should be construed as “assignment rule for assigning each one of the content-related second data segments to one of the content-related first data segments, where in the assignment is not performed by using exact timing information for the content-related data segments.” Id. (emphasis in original).

VideoLabs also proposed that claim 1’s “content-related . . . data segments” term be construed as “data segments ordered in a file such that they will present the file’s contents in their intended presentation order when5 rendered sequentially.” Id. at *7 (citing J.A. 1148). According to VideoLabs, the ’794 patent inventors had acted as their own lexicographer by defining this claim term in the specification. Id. (quoting ’794 patent col. 4 ll. 14–17 (“The term ‘content-related’ is understood to mean that first and second data segments have a syntactical meaning within the respective data file.”)). But instead of copying the asserted definition in the specification, VideoLabs’s proposed construction reflected its own understanding of “syntactical meaning.” Id.

In the Petitioner’s Reply, in addition to arguing that VideoLabs’s constructions were incorrect, Unified also argued, in the alternative, that Sonohara nevertheless discloses the “syntactical meaning” of the “content-related . . . data segments.” See J.A. 1500–03. According to Unified, Sonohara’s composite file—containing image and sound data segments mapped to the ’794 patent’s “content-related . . . data segments”—has syntax information (i.e., header and track identification numbers). J.A. 1500. Unified further noted that the composite file is assigned those syntaxes during the file production process which creates the composite file from the two originating files. Id. Thus, the originating image and sound files likewise have syntactical meaning. Id.

In its Final Written Decision, the Board declined to construe the term “assignment rule,” but adopted the definition in the ’794 specification to construe “content-related . . . data segments. Decision, 2023 WL 8717013, *7– 9. The Board then mapped Sonohara’s segmented image and sound data to the as-construed “content-related . . . data segments” and held that Unified had shown that Sonohara anticipated claims 1, 3, 5, 9, 12–13, and 15. Id. at *16–17, *20. In holding dependent claims 20 and 21 were not shown to be unpatentable, the Board determined that Sonohara’s “assignment rule” was “based on” two different timestamps. Id. at *19.

6After VideoLabs’s request for rehearing to the Board was denied, VideoLabs appealed. See J.A. 51. We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A) and 35 U.S.C. § 141(c).

STANDARD OF REVIEW

We review the Board’s decisions under the standards in the Administrative Procedure Act (APA), 5 U.S.C. § 706. We review “the Board’s compliance with the procedural requirements of the [APA] de novo.” Axonics, Inc. v. Medtronic, Inc., 75 F.4th 1374, 1380 (Fed. Cir. 2023). “We review the [Board’s] factual determinations for substantial evidence and its legal determinations de novo.” In re Nuvasive, Inc., 842 F.3d 1376, 1379 (Fed. Cir. 2016). Substantial evidence is “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Id. at 1380 (quoting In re Applied Materials, Inc., 692 F.3d 1289, 1294 (Fed. Cir. 2012)).

DISCUSSION

I. Assignment Rule

In declining to construe the term “assignment rule” with a negative limitation, the Board noted that disavowal must be “clear and unmistakable.” See Decision, 2023 WL 8717013, at *8 (citing Omega Engineering, Inc. v. Raytek Corp., 334 F.3d 1314, 1325 (Fed. Cir. 2003)). The Board determined that the standard for prosecution disclaimer had not been met.

Though VideoLabs proposed a negative limitation focused on assignment rules using “exact timing information,” the Board indicated that the phrase was “used nowhere in the prosecution history.” Id. at *9. Instead, the Board agreed with Unified that the prosecution history was “more fairly characterized as ‘the applicant argu[ing] that7 claim 1 was distinguishable over Shin² [] and Rosenau³ [] because these references consider time information for synchronization, rather than using an assignment rule for synchronization.’” Id. (first alteration in original). Thus, the Board declined to add the negative limitation and exclude assignment rules using “exact timing information.” Id. at *10.

Before this Court, VideoLabs argues for the first time that the Board’s construction of “assignment rule” was in error due to its inclusion of “techniques using timestamps.” Appellant Br. 42.

This argument is forfeited. It is well-established that “a party may not introduce new claim construction arguments on appeal or alter the scope of the claim construction positions it took below.” Conoco, Inc. v. Energy & Env’t Int’l, L.C., 460 F.3d 1349, 1358–59 (Fed. Cir. 2006) (citing Interactive Gift Exp., Inc. v. Compuserve Inc., 256 F.3d 1323, 1346–47 (Fed. Cir. 2001)). VideoLabs acknowledged below that a negative limitation disclaiming only “timestamps” is a substantively different claim construction position than it took before the Board. See J.A. 2197– 98 (explaining “[e]xact timing information is broader than time stamps”).

But “a position not presented in the tribunal under review will not be considered on appeal in the absence of exceptional circumstances.” In re Google Tech. Holdings LLC, 980 F.3d 858, 863 (Fed. Cir. 2020). VideoLabs has not raised any exceptional circumstances that would warrant considering its new argument, and we thus decline to consider it.

8II. “Content-Related . . . Data Segments”

The Board construed the term “content-related . . . data segments” as “segments that have a syntactical meaning within the respective data file,” in accordance with the specification’s definition. Decision, 2023 WL 8717013, at *7–8; see ’794 patent col. 4 ll. 14–17. The Board, however, rejected VideoLabs’s construction to the extent that it added “additional gloss.” Id. at *7. In the Board’s view, it was apparent that the ’794 applicant “chose to limit its definition of ‘content-related’ in terms of the broader phase ‘syntactical meaning’ rather than through narrower specific reference to presentation order.” Id. at *8.

On appeal, VideoLabs does not meaningfully dispute the Board’s construction, but argues the Board erred by (1) considering Unified’s argument, raised for the first time in its Reply, that Sonohara discloses the “syntactical meaning” portion of the construction for “content-related . . . data segments,” and (2) determining that Sonohara disclosed “content-related . . . data segments” as construed. Appellant Br. 27–41. We address each in turn.

A. “Syntactical Meaning” Reply Argument

In its Board rehearing request, VideoLabs contended that the Board impermissibly allowed Unified to address “syntactical meaning” in the Petitioner’s Reply when Unified’s Petition made no such argument. J.A. 52. The Board disagreed. Citing Axonics, the Board determined that once the Patent Owner proposed a new construction for the “content-related . . . data segments” in its Patent Owner Response, the Petitioner had the right to respond to the construction in its Reply, including whether Sonohara discloses “content-related . . . data segments” with syntactical meaning in its data file. Id. at 53–54.

Before this Court, VideoLabs argues that this case does not fall within the Axonics rule, because Unified did not9 address the “content-related” requirement in the petition, “under any construction.” Appellant Br. 32. According to VideoLabs, Unified’s petition lacked discussion of how the segments identified in Sonohara have “syntactical meaning”—a “content-related . . . data segment” requirement— a failure of proof which should have barred Unified from “fill[ing] the gap” in Reply. Id. at 30.

We agree with the Board that Axonics applies. When VideoLabs raised its claim construction argument after the Institution stage in which the Board declined to construe the term “content-related . . . data segments,” Unified was entitled to an opportunity to respond. See Axonics, 75 F.4th at 1383. Unified’s discussion of “syntactical meaning” within Sonohara relied on the same (only) embodiment in the patent, see Sonohara col. 3 ll. 48–49 (“An embodiment of the invention will now be described with reference to the accompanying drawings.”), and relied upon portions of Unified’s argument in the petition. Compare J.A. 1500 (discussing in Petitioner’s Reply that “syntactical meaning” is assigned to the composite file during the file production process), with J.A. 122–23 (discussing in Petition that the originating image and sound file undergo file production processing to form the composite file).

VideoLabs’s arguments that Unified should have “clearly anticipated”⁴ the construction of “content-related . . . data segments” based on the patent’s specification and therefore should have presented argument in the Petition under that construction is one the Axonics decision does not endorse. See Axonics, 75 F.4th at 1383 n.10 (stating that there is “no support” for a rule which draws a line between claim constructions that could be “clearly10 anticipated” and those that are not). Thus, the Board’s decision to review Unified’s mapping of Sonohara’s “image data” and “sound data” to the ’794 patent’s “content-related . . . data segments” was in accordance with its obligations under the APA.

B. Merits Argument

The Board found that one could infer syntactical meaning in Sonohara’s originating files (i.e., the claimed “first data file” and “second data file”), because the composite file, containing header and track information (i.e., “syntactical meaning”), was formed from the originating image and sound files. Decision, 2023 WL 8717013, at *17. In the order denying rehearing, the Board further explained the originating files had syntactical meaning “because the files have an order of data according to a chronological sequence, and that this syntactical meaning is imparted by the data and track identifiers to the file containing the image data segments and sound data segments.” J.A. 55.

VideoLabs’s final argument disputes the Board’s determination, on the merits, that Sonohara discloses “content-related . . . data segments” as construed. VideoLabs argues that the Board erred because it inferred “syntactical meaning” in Sonohara’s originating image file and sound file, even though the file production process creates a separate composite file with header and track information. Appellant Br. 36. According to VideoLabs, it is the composite file alone which contains “syntactical meaning,” and there is no “syntactical meaning within the [originating] data file.” Id. at 36, 38–41 (emphasis added).

The Board’s findings are supported by substantial evidence. Unified’s expert explained the “syntactical meaning” in the original image and sound files is the same as the “syntactical meaning” in the composite file, because the composite file “is created from image and sound portions” of the original files during the file production process. J.A. 1558 ¶ 55; J.A. 1554–57 ¶¶ 52–54. Unified’s expert further11 explained that the composite file includes “syntax information including header information with track and data identifiers identifying the relationship between the image and audio portions” of the file. Decision, 2023 WL 8717013, at *17 (citing J.A. 1554–58 ¶¶ 52–55). The Board reasonably relied upon this testimony from Unified’s expert to support its finding that a skilled artisan would infer “syntactical meaning” in the originating files after reviewing the file production process and header and track information in the composite file. Id.; see also J.A. 55–59. Thus, we affirm the Board’s analysis.

CONCLUSION

We have considered the remainder of VideoLabs’s arguments and find them unpersuasive. For the foregoing reasons, we affirm.

AFFIRMED