dissenting.
The court today rewrites the claims of the ’688 patent, adopting a “claim construction” that states the invention more broadly than did the patentee. The court then holds its broadened claims anticipated by the prior art, on which the patentee has provided an improvement that is not shown in the prior art.
The court errs in its analysis of the subject matter that is claimed, and strays from the substantial evidence standard of review of PTAB findings of fact. From the court’s finding of invalidity based on “anticipation,” over-turning the finding of the Board, I respectfully dissent.
Discussion
The court’s rejection of the Board’s finding that the claims are not anticipated is based on an incorrect understanding of the claims, coupled with an unwarranted enlargement of the references. The ’688 patent is directed to an improvement in blender technology, an improvement explained in the specification and by unrebutted expert testimony. My colleagues make their own findings, construe the claims to broadly include the prior art, and then invalidate the claims based on their unduly broad claim construction.
The description in the specification, and the guidance of the prosecution history, negate the majority’s construction of the claims to include the prior art. The process described and claimed in the ’688 patent is not shown in the prior art. The Board’s findings are supported by substantial evidence, and require affirmance. My colleagues’ de novo findings are contrary to the record, overstep our appellate role, and are incorrect in fact and law.
The claimed subject matter, including the settling speed and the blender sequence, are not shown in the prior art
My colleagues criticize the absence of “construction” of the term “settling speed” in the Board’s final decision.1 Maj. Op. 1374-75. The Board did not err in holding that this term did not require “construction.” Such a holding by the expert PTO Board is not grounds for discarding the Board’s findings and ignoring the expert *1380and documentary evidence presented to and discussed by the Board.
The Board also stated that: “Our decision ultimately does not hinge, on the precise contours of a construction of ‘settling speed.’ ” Board Op. 7. In its initial act to institute review, the Board described the settling speed “in accordance with the patent’s disclosure, i.e., as a speed at which the cutter assembly has slowed enough to allow the blender contents to be processed again.” Homeland Housewares, LLC v. Whirlpool Corp., IPR2014-00877, 2014 WL 5585266, at *5. My colleagues do not ascribe error to this finding, but simply discard it in favor of their own definition of settling speed based upon generalized blender operations of the prior art, particularly prior art from 1972 as discussed in Wulf, through which my colleagues reach their ultimate finding of invalidity based on “anticipation.”
No error has been shown in the Board’s treatment of “settling speed” in the ’688 patent, a treatment based on the specification, the prior art, and'unrebutted expert testimony. My colleagues do not discuss the substantial evidence that supports the Board’s ruling. Contra 5 U.S.C. § 706(2) (the appellate task is to “hold unlawful and set aside agency action, findings, and conclusions found to be ... unsupported by substantial evidence.”). Here, the Board’s findings are plainly supported by substantial evidence.
The Board correctly found that the Wulf reference does not anticipate the ‘688 claims
“Anticipation under 35 U.S.C. § 102 is a question of fact,” and “[w]e review the Board’s factual findings for substantial evidence,” Kennametal, Inc. v. Ingersoll Cutting Tool Co., 780 F.3d 1376, 1381 (Fed. Cir. 2015).
The panel majority finds that “[i]n light of [ ] teachings from the Wulf specification, Figure 25 discloses a settling speed limitation consistent with our construction, as well as the other elements of the pulsing cycle in claim 1” and thus finds anticipation. Maj. Op. 1378. The panel majority omits the specific limitations of the claim, and restates the ’688 invention in generalized terms that do not distinguish it from Wulf. However, the blender systems in Wulf and in the ’688 patent are not the same, and the Wulf method is distinguished in the ’688 claims. The Board recognized the distinction, and the Board’s finding that the ’688 claims are not anticipated by Wulf is supported by substantial evidence in the record before the Board.
In finding that Wulf anticipates, the panel majority ignores the limitations in the ’688 claims, and also violates “[t]he requirement that the prior art elements themselves be ‘arranged as in the claim’ means that claims cannot be ‘treated ... as mere catalogs of separate parts, in disregard of the part-to-part relationships set forth in the claims and that give the claims their meaning.’ ” Therasense, Inc. v. Becton, Dickinson & Co., 593 F.3d 1325, 1332 (Fed. Cir. 2010) (quoting Lindemann Mas-chinenfabrik GmbH v. Am. Hoist & Derrick Co., 730 F.2d 1452, 1459 (Fed. Cir. 1984)).
Wulf indeed shows some automation of blender action, and the Board cited the unrebutted expert testimony in finding that Wulf lacked the automated pulsing of the ’688 patent. The majority does not acknowledge the basis of the Board’s decision, and instead cites Wulf s discussion of a 1972 Swanke patent as background information to fill any anticipatory gaps in Wulf. The Board found that “the manual process described by Wulf as background information does not anticipate the independent claims, which require ‘automatically controlling a rotational speed of the cutter assembly’ to effect the recited puls*1381ing.” Board Op. 13. That finding is supported by substantial evidence.
My colleagues do not explain where the procedure claimed in the ’688 patent is shown in Wulf, including in the discussion of the Swanke patent and the general operation of blenders in 1972. The panel majority instead imputes to Wulfs general reference to high and low motor speeds, the hindsight knowledge of the advance described and claimed in the ’688 patent. As the Board discussed, Wulf does not show the ’688 method whereby each pulse includes a predetermined operating speed, a deceleration phase to the settling speed, and an acceleration from the settling speed to the operating speed—all limitations in claim 1. Nor does Wulf show “automatically controlling a rotational speed of the cutter assembly to effect a pulsing,” in claim 1. The distinctions from Wulf were found and relied upon by the Board and are not disputed by my colleagues; instead, they are ignored. However, “[t]he standard for lack of novelty, that is, for ‘anticipation,’ is one of strict identity.” 1-3 Chisum on Patents § 3.02 (citing sources).
“A patent is invalid for anticipation when the same device or method, having all of the elements contained in the claim limitations, is described in a single prior art reference.” Crown Operations Int’l, Ltd. v. Solutia Inc., 289 F.3d 1367, 1375 (Fed. Cir. 2002). The majority’s reliance on general background of blender operation does not establish anticipation of the claims, whose limitations are not shown in either Wulf or the background references cited in Wulf. The Board’s finding that the blender operation claimed in the ’688 patent is not described in the prior art, is supported by substantial evidence.
Wulfs Figure 25 does not show the method of the ’688 claims
My colleagues focus on Figure 25 of Wulf, stating that the functions marked in the figure anticipate the ’688 claims. The Board received expert testimony on Figure 25, with Mr. Faerber testifying that ‘Wulf gives no indication of whether any of these speeds ... would cause items to settle around the cutter,” and that Wulf “only discloses items settling when the motor stops.” J.A. 430-31 (citing Wulf, col. 2, 11. 15-19 (“[t]he user may have to stop the blending process to dislodge the ice or to assure the ice is coming into contact with the blades.”)). The Board observed that the expert testimony was “unrebutted.” The testimony remains unrebutted.
Instead, the panel majority mischarac-terizes the expert testimony. Wulfs Figure 25 shows both a “ramp down to low” and then “low for 5 seconds,” and Mr. Faerber testified that “the ‘low’ speed cannot be the settling speed because it “is maintained for at least five seconds ... [and] there is no reason to maintain a settling speed for so long.” J.A. 430. The majority states that the specification contradicts Mr. Faerber; that is incorrect, as shown in the Board’s discussion.
The Board found that the specification teaches deceleration to the settled condition via the settling speed, not deceleration and operation at the settled condition via the settling speed. Mr. Faerber explained that “[i]t would not be useful to maintain the settling speed for any significant length of time,” J.A. 420-21, since “speeds typical of settling speeds do not efficiently comminute the blender contents,” J.A. 412. Mr. Faerber explained that where a speed is maintained in accordance with Wulf, it is “likely too high to be considered a settling speed.” J.A. 420-21. Mr. Faerber’s testimony and the Board’s findings are not in conflict with the specification; it is the majority’s finding that is in conflict with the specification.
*1382The panel majority also brushes aside Mr. Faerber’s testimony regarding the relative slow and high speeds of a blending program for powdered drinks. Maj. Op. 1355-56. The inability of powered drink mix to settle “even [at] very low speeds” reinforces his testimony that the “low” speed in Figure 25 is not a “settling speed” since nothing would “settle.” This is not merely an absence of an intended function; it is lack of a structural limitation, as the Board found.
The Board’s findings are supported by substantial evidence; the majority does not show otherwise.2 As reiterated in Apple Inc. v. Samsung Electronics Co., Ltd., “such fact findings are indisputably the province of the [fact-finder].” 839 F.3d 1034, 1039 (Fed. Cir. 2016). The structure of the post-grant administrative process assigns to this court the traditional “appellate function ... limited to deciding the issues raised on appeal by the parties, deciding these issues only on the basis of the record made below, and ... requiring appropriate deference be applied to the review of factfindings.” Id.; see also Cutter v. Wilkinson, 544 U. S. 709, 718, n.7, 125 S.Ct. 2113, 161 L.Ed.2d 1020 (2005) (“[W]e are a court of review, not first view.”). The panel majority does not discuss the substantial evidence on which the Board’s decision was based, ignores the unrebutted expert evidence, and in general oversteps the appellate role.
Conclusion
Substantial evidence supports the Board’s finding of no anticipation. The ’688 patent embodies a novel distinction not shown in any cited reference, as the Board recognized. My colleagues present a flawed analysis and incorrect conclusion. I respectfully dissent.
. Homeland Housewares, LLC v. Whirlpool Corp., No. IPR2014-00877 (P.T.A.B. Oct. 21, 2015) (“Board Op.”).
. The majority also states that "[cjontrary to the dissent, in holding that Wulf does not anticipate, the Board never made any findings with respect to the length of the settling speed as a part of the claimed invention, nor approved that aspect of the Faerber opinion.” Maj. Op. 1378 n.4. The majority does not explain how these aspects affect the substantial evidence supporting the Board’s decision. The question on appeal is not whether the Board discussed the majority's selected issues; the question is whether the Board’s findings in support of its decision are supported by substantial evidence. The record demonstrates that they are.