Case: 21-2348   Document: 48     Page: 1    Filed: 01/20/2023




        NOTE: This disposition is nonprecedential.


   United States Court of Appeals
       for the Federal Circuit
                 ______________________

  LKQ CORPORATION, KEYSTONE AUTOMOTIVE
            INDUSTRIES, INC.,
                Appellants

                            v.

  GM GLOBAL TECHNOLOGY OPERATIONS LLC,
                   Appellee
            ______________________

                       2021-2348
                 ______________________

     Appeal from the United States Patent and Trademark
 Office, Patent Trial and Appeal Board in No. IPR2020-
 00534.
                  ______________________

                Decided: January 20, 2023
                 ______________________

    MARK A. LEMLEY, Lex Lumina PLLC, New York, NY,
 argued for appellants. Also represented by MARK P.
 MCKENNA; ANDREW HIMEBAUGH, BARRY IRWIN, IFTEKHAR
 ZAIM, Irwin IP LLC, Chicago, IL.

    JOSEPH HERRIGES, JR., Fish & Richardson P.C., Minne-
 apolis, MN, argued for appellee. Also represented by JOHN
 A. DRAGSETH; NITIKA GUPTA FIORELLA, Wilmington, DE;
 LAURA E. POWELL, Washington, DC.
Case: 21-2348      Document: 48     Page: 2    Filed: 01/20/2023




 2                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


                    ______________________

     Before LOURIE, CLEVENGER, and STARK, Circuit Judges.
            Opinion for the court filed PER CURIAM.
        Additional views filed by Circuit Judge LOURIE.
     Opinion concurring in part and concurring in judgment
                 filed by Circuit Judge STARK.
 PER CURIAM.
      LKQ Corp. and Keystone Automotive Industries, Inc.
 (collectively, “LKQ”) appeal from a final written decision of
 the U.S. Patent and Trademark Office Patent Trial and Ap-
 peal Board (“the Board”) holding that LKQ failed to show
 by a preponderance of the evidence that U.S. Patent
 D797,625 (the “’625 patent”) was anticipated or would have
 been obvious over the cited prior art before the effective fil-
 ing date. See LKQ Corp. v. GM Glob. Tech. Operations
 LLC, IPR2020-00534, Paper 28 (P.T.A.B. Aug. 4, 2021)
 (“Decision”), J.A. 1–60. For the reasons provided below, we
 affirm.
                         BACKGROUND
     GM Global Technology Operations LLC (“GM”) owns
 the ’625 patent, which is directed to an “ornamental design
 for a vehicle front fender” as shown below.
Case: 21-2348    Document: 48      Page: 3    Filed: 01/20/2023




 LKQ CORPORATION v.                                         3
 GM GLOBAL TECHNOLOGY OPERATIONS LLC




     GM manufactures and sells automotive vehicles. LKQ
 sells automotive body repair parts for most mainstream ve-
 hicle models available, including front fenders for vehicles
 manufactured by GM. GM and LKQ had previously been
 parties to a license agreement, under which LKQ was
 granted a license to many of GM’s design patents. The li-
 cense agreement expired in February 2022 following a
 breakdown of renewal negotiations, after which GM sent
 letters to LKQ’s business partners alleging that the now
 unlicensed LKQ parts infringe its patents.
     LKQ petitioned for inter partes review of the ’625 pa-
 tent, asserting that it was anticipated by U.S. Patent
 D773,340 (“Lian”) and would have been obvious over Lian
 alone or in combination with the design of the 2010 Hyun-
 dai Tucson as disclosed in a promotional brochure (“Tuc-
 son”). The Board issued a final written decision concluding
 that LKQ had not demonstrated by a preponderance of the
 evidence that the ’625 patent was anticipated or would
 have been obvious before the effective filing date. Decision,
 J.A. 1–60.
Case: 21-2348      Document: 48      Page: 4     Filed: 01/20/2023




 4                                      LKQ CORPORATION v.
                       GM GLOBAL TECHNOLOGY OPERATIONS LLC


     First, the Board determined that the ordinary observer
 would include both retail consumers who purchase replace-
 ment fenders and commercial replacement part buyers.
 Decision, J.A. 20–23. In so doing, the Board emphasized
 that the “’625 design claims a ‘vehicle front fender,’ not a
 vehicle in total.” Decision, J.A. 21.
     Second, applying that understanding of the ordinary
 observer, the Board concluded that, although there were
 some similarities between the claimed design and Lian,
 there were a number of key differences, including: (1) the
 wheel arch shape and terminus, (2) the door cut line, (3)
 the protrusion, (4) the sculpting, (5) the first and second
 creases, (6) the inflection line (i.e., third crease), and (7) the
 concavity line. Decision, J.A. 30–45. The Board found that
 these differences in the claimed features contributed to dif-
 ferent overall appearances in the design. Id. For example,
 the Board found that the claimed design’s sculpting and
 crease pattern contributed to a smooth, curved overall ap-
 pearance, whereas Lian depicted substantially linear, an-
 gled lines. Decision, J.A. 43. The Board was ultimately not
 persuaded that an ordinary observer would be deceived
 into purchasing the Lian front fender supposing it to be the
 claimed fender, finding no anticipation. Decision, J.A. 45.
      Third, applying the tests established in Rosen and
 Durling, the Board found that LKQ failed to identify a suf-
 ficient primary reference, and therefore failed to prove ob-
 viousness by a preponderance of the evidence. Durling v.
 Spectrum Furniture Co., Inc., 

 (Fed. Cir. 1996);
 In re Rosen, 

 (C.C.P.A. 1982); Decision, J.A.
 48–58. Because the Board found that Lian did not qualify
 as a proper primary reference under Rosen, the Board did
 not turn to Durling step two and look beyond Lian to Tuc-
 son. Decision, J.A. 58.
    In summary, the Board concluded that LKQ had not
 demonstrated that the claimed design of the ’625 patent
 was anticipated or would have been obvious before the
Case: 21-2348    Document: 48     Page: 5    Filed: 01/20/2023




 LKQ CORPORATION v.                                        5
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 effective filing date. LKQ appealed. We have jurisdiction
 under 

(a)(4)(A).
                        DISCUSSION
     LKQ raises two main challenges on appeal. First, LKQ
 contends that the Board erred in finding that the ordinary
 observer would include only retail consumers who pur-
 chase replacement fenders and commercial replacement
 part buyers, and, ultimately, in finding no anticipation.
 Second, LKQ contends that the Rosen and Durling tests on
 which the Board relied in its obviousness analysis have
 been implicitly overruled by the Supreme Court’s decision
 in KSR International Co. v. Telflex, Inc., 

 (2007). We address each argument in turn.
      We review the Board’s legal conclusions de novo and its
 factual findings for substantial evidence. Campbell Soup
 Co. v. Gamon Plus, Inc., 

, 1339 (Fed. Cir.
 2019). Anticipation is a question of fact reviewed for sub-
 stantial evidence. Id.; Int’l Seaway Trading Corp. v.
 Walgreens Corp., 

 (Fed. Cir. 2009). The
 ultimate determination of obviousness is reviewed de novo,
 and any underlying factual findings are reviewed for sub-
 stantial evidence. Campbell Soup Co. v. Gamon Plus, Inc.,
 

, 1275 (Fed. Cir. 2021). Substantial evidence
 is “such relevant evidence as a reasonable mind might ac-
 cept as adequate to support a conclusion.” Consol. Edison
 Co. v. N.L.R.B., 

 (1938).
                              I
     As an initial matter, GM raises a threshold issue, con-
 tending that LKQ lacks standing to appeal and forfeited
 the opportunity to demonstrate otherwise by not present-
 ing any evidence or argumentation on the issue in its open-
 ing brief. Standing requires an “injury in fact,” meaning
 evidence of “an invasion of a legally protected interest
 which is (a) concrete and particularized, and (b) actual or
 imminent, not conjectural or hypothetical.” Lujan v. Defs.
Case: 21-2348    Document: 48      Page: 6    Filed: 01/20/2023




 6                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 of Wildlife, 

 (1992) (quotation marks and
 citations omitted). If a party is “currently engaged in con-
 duct creating a substantial risk of infringement,” standing
 may exist. Gen. Elec. Co. v. Raytheon Techs. Corp., 

, 1341–42 (Fed. Cir. 2020).
     We find that LKQ did establish that it has standing.
 LKQ asserted in its opening brief that it is currently en-
 gaged in conduct creating a substantial risk of infringe-
 ment. Appellant Br. at 1. For purposes of evaluating
 standing, we must accept this representation as true.
 Amerigen Pharm. Ltd. v. UCB Pharma GmBH, 

 (Fed. Cir. 2019). And, indeed, GM does not dis-
 pute this assertion. That LKQ supplied additional facts in
 its reply brief in response to GM’s challenge does not
 change the facts themselves or render LKQ’s original as-
 sertion insufficient. LKQ represents that it sells a replace-
 ment fender for GM’s 2018–2022 Chevrolet Equinox, which
 GM has asserted comprises a front fender that practices
 the ’625 patent. See Appellant Reply Br. at 2–3. Indeed,
 GM sent a letter to at least one of LKQ’s business partners
 alleging that the now unlicensed LKQ parts, including the
 replacement fender for the 2018–2022 Chevrolet Equinox,
 infringe its patents. 

 For purposes of evaluating stand-
 ing, we accept these representations as true. Amerigen,
 

. LKQ is therefore currently engaged in
 conduct creating a substantial risk of infringement, and we
 find that it has standing to appeal.
                              II
     We next consider LKQ’s challenge to the Board’s deter-
 mination that LKQ failed to meet its burden to prove that
 the ’625 patent was anticipated by Lian. LKQ first argues
 that the Board erred in not finding that original retail ve-
 hicle purchasers are the correct ordinary observer. LKQ
 notes that replacement parts are merely a downstream
 consequence of prior whole vehicle sales, that front fenders
 are not sold in the first instance as replacement products,
Case: 21-2348    Document: 48     Page: 7    Filed: 01/20/2023




 LKQ CORPORATION v.                                        7
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 and that replacement products make up a small subset of
 total fender sales. LKQ further asserts that that alleged
 error led the Board to place undue emphasis on allegedly
 trivial differences between the claimed design of the ’625
 patent and Lian that would not have mattered to the cor-
 rect ordinary observer, and that the Board failed to
 properly compare the designs as a whole.
     GM responds that the Board correctly defined the ordi-
 nary observer. GM emphasizes that the ’625 patent is not
 directed to the design of a whole vehicle, but rather a de-
 sign for a specific part of a vehicle. Therefore, GM argues
 that the ordinary observer must be the person who pur-
 chases that part or is otherwise sufficiently interested in
 that part, not necessarily the vehicle as a whole. GM fur-
 ther argues that, regardless of the ordinary observer, the
 Board’s holding that the claimed design of the ’625 patent
 is not anticipated by Lian was supported by substantial ev-
 idence.
     We agree with GM that the Board’s finding that the
 ordinary observer would include both retail consumers who
 purchase replacement fenders and commercial replace-
 ment part buyers who purchase replacement fenders was
 supported by substantial evidence. Defining the ordinary
 observer is a fact-dependent inquiry. See Arminak & As-
 socs., Inc. v. Saint-Gobain Calmar, Inc., 

 (Fed. Cir. 2007) (abrogated on other grounds by Egyp-
 tian Goddess, Inc. v. Swisa, Inc., 

 (Fed. Cir.
 2008)). It is through the lens of the ordinary observer that
 the anticipation inquiry is viewed. Gorham Co. v. White,
 

 (1871). The Board’s finding that the ordi-
 nary observer here includes repair professionals and vehi-
 cle owners purchasing replacement parts, and does not
 include retail vehicle purchasers, is well-grounded in the
 record, LKQ itself recognized that repair shop mechanics
 and auto parts store employees who order replacement
 parts may be ordinary observers of front fenders. J.A.
 182–83. And LKQ cannot disagree that there is a market
Case: 21-2348     Document: 48     Page: 8    Filed: 01/20/2023




 8                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 for the claimed fender alone, given that its business is sell-
 ing individual replacement parts. See Appellant Reply Br.
 at 2. That GM sells whole vehicles to retail consumers, ra-
 ther than fenders to car manufacturers, is of no conse-
 quence. As the Board noted, the ordinary observer test
 requires consideration of the features visible in the “normal
 use” lifetime of the accused product, i.e., “from the comple-
 tion of manufacture or assembly until the ultimate destruc-
 tion, loss, or disappearance of the article.” Decision, J.A.
 18 (quoting Int’l Seaway, 

). See also
 Contessa Food Prods. v. Conagra, Inc., 

, 1380
 (Fed. Cir. 2002) (“[T]he ‘ordinary observer’ analysis is not
 limited to those features visible during only one phase or
 portion of the normal use lifetime of an accused product.”).
 The “normal use” would thus include repair or replace-
 ment.
     Most notably, the claimed design of the ’625 patent is
 not the design of a whole vehicle, but rather a design of a
 specific part of a vehicle. J.A. 61–64. In Arminak and Key-
 stone, we recognized that the ordinary observer can be the
 purchaser of a component part of a later-assembled product
 as opposed to the purchaser of the later-assembled product.
 Arminak, 501 F.3d at 1321–24; KeyStone Retaining Wall
 Sys., Inc. v. Westrock, Inc., 

, 1451 (Fed. Cir.
 1993). The same is true here. In this case, substantial ev-
 idence supports the Board’s conclusion that the ordinary
 observer is a person who purchases that part or is other-
 wise sufficiently interested in that part, not the vehicle as
 a whole. See Egyptian Goddess, 

.
     When determining whether a design patent is antici-
 pated, we consider:
     [I]f, in the eye of an ordinary observer, giving such
     attention as a purchaser usually gives, two designs
     are substantially the same, if the resemblance is
     such as to deceive such an observer, inducing him
     to purchase one supposing it to be the other . . . .
Case: 21-2348     Document: 48         Page: 9   Filed: 01/20/2023




 LKQ CORPORATION v.                                            9
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 Gorham, 

. 1 Under this standard, substantial
 evidence supported the Board’s finding of no anticipation.
      The Board concluded that the design claimed in the
 ’625 patent and Lian design created different overall im-
 pressions. The Board pointed to several differences be-
 tween the designs, namely: (1) the wheel arch shape and
 terminus, (2) the door cut line, (3) the protrusion, (4) the
 sculpting, (5) the first and second creases, (6) the inflection
 line (i.e., third crease), and (7) the concavity line. Decision,
 J.A. 30–45. And LKQ’s claim that the Board failed to
 properly compare the designs as a whole is without merit.
 The Board explained how the aforementioned differences
 contributed to the overall appearance of both the ’625 pa-
 tent and Lian, noting, for example, the claimed design’s
 smooth, curved overall appearance, and Lian’s substan-
 tially linear, angled overall appearance. Decision, J.A. 43.
 We agree and find that the Board’s finding of no anticipa-
 tion was supported by substantial evidence.
                                 III
     LKQ next argues that the Supreme Court’s decision in
 KSR implicitly overruled the long-standing Rosen and
 Durling tests for obviousness of design patents, and there-
 fore the Board’s decision applying such test must at least
 be vacated and remanded. For design patent obviousness,
 Durling outlines a two-step analysis. First, it must be de-
 termined whether a primary reference, also known as a
 Rosen reference, exists with characteristics “basically the
 same” as the claimed design by discerning the visual



     1   While the test articulated in Gorham was with re-
 gard to infringement, we have found “that the same tests
 must be applied to infringement and anticipation,” as to
 “prevent an inconsistency from developing between the in-
 fringement and anticipation analyses.” Int’l Seaway, 

.
Case: 21-2348    Document: 48     Page: 10    Filed: 01/20/2023




 10                                    LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 impression of the design as a whole. Durling, 

; Rosen, 673 F.2d at 390–91. Second, if a satisfactory
 primary reference exists, the court must consider whether
 an ordinary designer would have modified the primary ref-
 erence to create a design with the same overall visual ap-
 pearance as the claimed design. 

 This test safeguards
 against a challenger picking and choosing features from
 multiple references to create something entirely new, fun-
 damentally changing the overall visual impression of the
 original designs. See In re Jennings, 

 (C.C.P.A. 1950) (explaining that one must start with
 “something in existence—not with something that might be
 brought into existence by selecting individual features
 from prior art and combining them”).
     While KSR does not address Rosen or Durling, LKQ ar-
 gues that its holding implicates the test established in
 these cases. GM argues that LKQ has forfeited this argu-
 ment by not raising it before the Board, and even if it were
 not forfeited, KSR does not overrule Rosen or Durling.
                              A
     As a general proposition, a federal appellate court does
 not consider issues not raised before the lower tribunal.
 Singleton v. Wulff, 

 (1976). The exact
 phrasing of the argument need not have been used below
 “so long as it can be said that the tribunal was ‘fairly put
 on notice as to the substance of the issue.’” Nike Inc. v.
 Adidas AG, 

 (Fed. Cir. 2016) (citing
 Consolidation Coal Co. v. United States, 

 (Fed. Cir. 2003) (quoting Nelson v. Adams, 

 (2000) (“But [issue preservation] does not demand
 the incantation of particular words; rather, it requires that
 the lower court be fairly put on notice as to the substance
 of the issue.”))).
     But even if an issue was not presented below, there is
 no absolute bar to considering and deciding the issue on
 appeal, as forfeiture is a matter of discretion. Harris Corp.
Case: 21-2348    Document: 48     Page: 11   Filed: 01/20/2023




 LKQ CORPORATION v.                                         11
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 v. Ericsson Inc., 

 (Fed. Cir. 2005) (“An
 appellate court retains case-by-case discretion over
 whether to apply waiver.”). In Icon Health and Fitness,
 Inc. v. Strava, Inc., we set forth some relevant considera-
 tions to guide the exercise of such discretion. 

 (Fed. Cir. 2017) (citing Automated Merch. Sys.,
 Inc. v. Lee, 

, 1379 (Fed. Cir. 2015)). These
 considerations include whether the issue involves a pure
 question of law and refusal to consider it would result in a
 miscarriage of justice, and whether the issue presents sig-
 nificant questions of general impact. 

     In this case, appellants argue that KSR, 

,
 overrules this court’s long-standing precedents, Rosen, 

, and Durling, 

, which outline the test
 for determining design patent obviousness. In their peti-
 tion to the Board, the appellants raised their KSR argu-
 ment in the following manner:
     As a final matter, the Federal Circuit has, to date,
     been silent regarding whether the Supreme Court’s
     opinion in KSR Int’l Co. v. Teleflex, Inc., 

 (2007), governs design patent obviousness,
     other than to say that it was “not obvious that the
     Supreme Court necessarily intended to exclude de-
     sign patents from the reach of KSR.” Titan Tire
     Corp. v. Case New Holland, Inc., 

,
     1384–85 (Fed. Cir. 2009). 2 In a design patent obvi-
     ousness analysis, to modify a primary reference, a
     secondary reference must be “so related [to the


     2   In KSR, the Supreme Court rejected the “rigid,
 mandatory formula” embodied in the Federal Circuit’s
 then-prevailing “teaching suggestion motivation” require-
 ment as improperly limiting the obviousness inquiry and,
 instead, prescribed a more flexible approach to obviousness
 and motivation to combine prior art teachings. KSR, 550
 U.S. at 420–22.
Case: 21-2348     Document: 48      Page: 12    Filed: 01/20/2023




 12                                     LKQ CORPORATION v.
                       GM GLOBAL TECHNOLOGY OPERATIONS LLC


      primary reference] that the appearance of certain
      ornamental features in one would suggest the ap-
      plication of those features to the other.” Titan Tire,
      

 (quoting Durling, 

). To the extent that the “so related” test oper-
      ates to unduly limit the scope of design patent ob-
      viousness, such an overly restrictive view would
      run afoul of KSR’s proscription against rigid re-
      strictions on the scope of an obviousness analysis.
 J.A. 174.
     We consider that this assertion in the petition is suffi-
 cient to preserve the argument that LKQ seeks to make,
 that KSR overrules our current test for design patent obvi-
 ousness. While not extensive, LKQ’s argument enunciates
 concern that the applied obviousness test was “overly re-
 strictive” in light of KSR, which is the heart of its argument
 on appeal. To be sure, LKQ could have made its argument
 in depth. However, such arguments would have been to an
 unreceptive audience, as the Board already had heard and
 rejected the argument. See, e.g., Johns Manville Corp. v.
 Knauf Insulation, Inc., IPR2015-1453, 2017 Pat. App.
 LEXIS 3419 at *63-64 (P.T.A.B. Jan. 11, 2017) (concluding
 that there was no suggestion that KSR is “germane to an
 obviousness determination in design patents, or has any
 effect upon the relevance of the Durling analysis,” and
 that, regardless, the “basically the same test” is consistent
 with KSR).
     Further, this challenge to Rosen and Durling is a pure
 question of law that undoubtedly presents a significant
 question of general impact, favoring the exercise of our dis-
 cretion to hear and decide this issue. See Icon Health, 

. Moreover, we may decide to apply the correct
 law even if the parties did not argue it, so long as the issue
 is properly before the court. Forshey v. Principi, 

, 1356–57 (Fed. Cir. 2002) (en banc) (superseded by
 statute on other grounds, Flores v. Nicholson, 476 F.3d
Case: 21-2348    Document: 48      Page: 13    Filed: 01/20/2023




 LKQ CORPORATION v.                                         13
 

 (Fed. Cir. 2006)). Given the aforementioned
 considerations and in our discretion, we decline to find
 LKQ’s KSR argument forfeited.
                               B
      Appellant argues that the Supreme Court, in reviewing
 Teleflex, Inc. v. KSR Int’l Co., 119 Fed. App’x 282 (Fed. Cir.
 2005), overruled Rosen and Durling. The Court stated,
 “[t]he obviousness analysis cannot be confined by a formal-
 istic conception of the words teaching, suggestion, and mo-
 tivation . . . .” KSR, 

. The Court did note
 that the teaching, suggestion, and motivation test “cap-
 tured a helpful insight,” but it found it overly “rigid.” 

. However, KSR did not involve or discuss design
 patents, which is the type of patent we have here before us
 and that was addressed by Rosen and Durling.
     And we note that in the more than fifteen years since
 KSR was decided, this court has decided over fifty design
 patent appeals. In these appeals, this court has continu-
 ally applied Rosen and Durling just as it had in the decades
 preceding. Notably, the correctness of our current law in
 light of KSR was raised in only two of those over fifty ap-
 peals, only tangentially, and not addressed in our decisions
 in those two cases. This challenge is therefore an outlier.
     We, as a panel, cannot overrule Rosen or Durling with-
 out a clear directive from the Supreme Court. Deckers
 Corp. v. United States, 

 (Fed. Cir. 2016)
 (explaining panels are “bound by prior panel decisions until
 they are overruled by the court en banc or the Supreme
 Court”). See also California Inst. Of Tech. v. Broadcom
 Ltd., 

, 990–91 (Fed. Cir. 2022) (explaining the
 panel only had authority to overrule circuit precedent with-
 out en banc action given the Supreme Court’s reasoning
 undermining said circuit precedent). As explained above, it
 is not clear the Supreme Court has overruled Rosen or
 Durling. The panel is therefore bound to apply existing law
 to this appeal.
Case: 21-2348    Document: 48     Page: 14    Filed: 01/20/2023




 14                                   LKQ CORPORATION v.
                     GM GLOBAL TECHNOLOGY OPERATIONS LLC


                              C
     Finally, applying existing law, we affirm the Board’s
 non-obviousness determination and conclude that its fac-
 tual findings are supported by substantial evidence.
      The Board subdivided the first part of the Durling test
 into two sub-steps. First, the Board concluded that LKQ
 failed to identify “the correct visual impression created by
 the patented design as a whole.” Durling, 

;
 Rosen, 673 F.2d at 390–91. The Board found that LKQ’s
 proposed construction “offers a starting point, but does not
 go far enough, in describing the overall appearance of the
 claimed design.” Decision, J.A. 14–17, 48. The Board
 noted that additional elements of the claimed design, un-
 addressed by Lian, included (1) a consistently curved door
 cut line and a circular wheel arch, (2) a consistent width
 lower rearward terminus defined between the wheel arch
 and door cut line, (3) a top protrusion having an upper sur-
 face having a thickness and consistent width that mirrors
 the inside surface of the u-shaped notch, (4) a gently curv-
 ing first and second crease extending forward from the door
 cut line and defining a concavity line there between, and
 (5) below the first and second crease, an inflection line ex-
 tending between the wheel arch and door cut line. Deci-
 sion, J.A. 14–17, 50.
     In reaching these determinations, the Board credited
 the testimony of GM’s expert witness and other witnesses.
 

 We cannot find that the Board erred in valuing certain
 testimony over others. In re Jolley, 

 (Fed. Cir. 2002) (where “two different, inconsistent conclu-
 sions may reasonably be drawn from the evidence in rec-
 ord, an agency’s decision to favor one conclusion over the
 other is the epitome of a decision that must be sustained
 upon review for substantial evidence”). The factual find-
 ings here that informed the Board’s articulation of the vis-
 ual impression created by the claimed design are supported
 by substantial evidence. However, we, like the Board,
Case: 21-2348    Document: 48      Page: 15    Filed: 01/20/2023




 LKQ CORPORATION v.                                         15
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 proceed to consider whether LKQ identified a Rosen refer-
 ence that is “basically the same” as the claimed design.
      LKQ failed to show that Lian creates “basically the
 same” visual impression as the claimed design. Durling,
 

; Rosen, 673 F.2d at 390–91. The Board
 found while Lian and the claimed design shared certain el-
 ements, “a complete visual comparison of the overall ap-
 pearance of both designs sufficiently shows that the two
 designs are quite different.” Decision, J.A. 53. For exam-
 ple, the Board noted the claimed design’s smooth, curved
 overall appearance as compared to Lian’s substantially lin-
 ear, angled overall appearance. See, e.g., id. at 54. Here
 too, the Board credited the testimony of GM’s witnesses, in
 addition to considering the images presented from the per-
 spective of an ordinary designer. Id. at 53–58. We likewise
 find the Board’s finding that Lian failed to create “basically
 the same” visual impression as the claimed design is there-
 fore supported by substantial evidence.
     We therefore affirm the Board’s conclusion that LKQ
 did not show that the ’625 patent would have been obvious
 over the cited references.
                         CONCLUSION
      We have considered LKQ’s remaining arguments but
 find them unpersuasive. For the foregoing reasons, the de-
 cision of the Board is affirmed.
                         AFFIRMED
Case: 21-2348    Document: 48      Page: 16   Filed: 01/20/2023




         NOTE: This disposition is nonprecedential.


    United States Court of Appeals
        for the Federal Circuit
                   ______________________

   LKQ CORPORATION, KEYSTONE AUTOMOTIVE
             INDUSTRIES, INC.,
                 Appellants

                              v.

   GM GLOBAL TECHNOLOGY OPERATIONS LLC,
                    Appellee
             ______________________

                         2021-2348
                   ______________________

     Appeal from the United States Patent and Trademark
 Office, Patent Trial and Appeal Board in No. IPR2020-
 00534.
                  ______________________

 LOURIE, Circuit Judge, additional views.
     The panel resolves this case by determining that sub-
 stantial evidence supports the Board’s conclusion that the
 claims of the ’625 patent are not unpatentable as antici-
 pated or obvious. I agree. I write separately to respond to
 the assertion by appellant that the Board improperly relied
 on our predecessor court’s Rosen decision for its rationale
 in deciding its case. Appellant argues that Rosen, what-
 ever its validity at the time it was decided, is inconsistent
 with, and hence was implicitly overruled, by the Supreme
 Court in KSR. The panel resolves the case without having
Case: 21-2348    Document: 48      Page: 17     Filed: 01/20/2023




 2                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 to reach this issue, but I comment separately on the argu-
 ment appellant raises, but which we do not reach.
      Rosen was decided by the Court of Customs and Patent
 Appeals in 1982, sitting with its usual en banc panel of
 Judges Markey, Rich, Baldwin, Miller, and Nies, hardly a
 group unversed in patent law. Rosen, in reversing the Pa-
 tent Office’s Board of Appeals, stated that “there must be a
 reference, a something in existence, the design character-
 istics of which are basically the same as the claimed design
 in order to support a holding of obviousness.” In re Rosen,
 

 (C.C.P.A. 1982) (citing In re Jennings,
 

 (C.C.P.A. 1950)).
     Appellant argues that the Supreme Court in reviewing
 Teleflex, Inc. v. KSR Int’l Co., 

 (Fed. Cir.
 2005), a non-precedential decision of ours that recited al-
 most as boilerplate the longstanding requirement that, in
 evaluating a combination of references asserted to render
 a claimed invention obvious, there must be some teaching,
 suggestion, or motivation shown in the prior art to combine
 the references, overruled Rosen. The Court stated, “[t]he
 obviousness analysis cannot be confined by a formalistic
 conception of the words teaching, suggestion, and motiva-
 tion . . . .” KSR Int’l Co. v. Teleflex Inc., 

 (2007). The Court did note that the teaching, suggestion,
 and motivation test “captured a helpful insight,” but it
 found it overly “rigid.” 

. That essentially is the
 broad concept that appellant now asserts has overruled an
 almost forty-year-old design patent decision that this court
 has been bound to follow and has continued to follow in the
 decade since KSR was decided.
     I disagree. First, KSR did not involve design patents,
 which is the type of patent we have here before us. Utility
 patents and design patents are distinct types of patents.
 Utility patents protect processes, machines, manufactures,
 compositions of matter, and improvements thereof. But
 their inventions must be “useful.” 

. A design
Case: 21-2348    Document: 48     Page: 18    Filed: 01/20/2023




 LKQ CORPORATION v.                                         3
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 patent protects ornamentation, the way an article looks, 

. See M.P.E.P. § 1502.01. Claims and claim
 construction in design patents are thus quite different com-
 pared with utility patents. See Richardson v. Stanley
 Works, Inc., 

 (Fed. Cir. 2010) (question-
 ing how “a court could effectively construe design claims,
 where necessary, in a way other than by describing the fea-
 tures shown in the drawings.”); Egyptian Goddess, Inc. v.
 Swisa, Inc., 

 (Fed. Cir. 2008) (en banc)
 (citing Dobson v. Dornan, 

 (1886)) (finding
 that a design is better represented by an illustration than
 a description).
     While 

, which deals with obviousness,
 does not differentiate between types of inventions, and
 hence applies to all types of patents, the considerations in-
 volved in determining obviousness are different in design
 patents. Obviousness of utility patents requires consider-
 ations such as unexpected properties, utility, and function.
 Design patents, on the other hand, relate to considerations
 such as the overall appearance, visual impressions, art-
 istry, and style of ornamental subject matter. Ornament is
 in the eyes of the beholder. Functional utility is objective.
     What is the utility or function of something that is or-
 namental? To be sure, it may also be functional and have
 use, as do the fenders in the case before us, but that func-
 tion is beside the point when considering whether its orna-
 mental features would have been obvious. And what is an
 unexpected property in the context of design patents, for it
 must be related to ornament, as function is not to be con-
 sidered in evaluating obviousness of a design?
     Obviousness of an ornamental design thus requires dif-
 ferent considerations from those of a utility invention.
 KSR did not address any of these considerations, and it did
 not even mention design patents. Surely, it did not intend
 to speak to obviousness of designs, and what was said
 about a test long applied to utility inventions was not
Case: 21-2348    Document: 48      Page: 19    Filed: 01/20/2023




 4                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 indicated to apply to design patents. It cannot reasonably
 be held to have overruled a precedent of one of our prede-
 cessor courts involving a type of patent it never mentioned.
      A comprehensive review of obviousness in design pa-
 tents, Overcoming the “Impossible Issue” of Nonobvious-
 ness in Design Patents, supports the view that “KSR has
 very limited, if any, applicability to design patents.” Janice
 M. Mueller and Daniel Harris Brean, 99 KY. L. J., 419, 518
 (2011). It further states that, “[m]ost of the KSR Court’s
 discussion is completely irrelevant to what design patents
 protect.” 

 It notes that “[t]he subject matter of a design
 patent is fundamentally different from a utility patent, as
 evidenced by the separate statutory provisions that define
 the eligible subject matter of each type of patent.” 

 “In-
 deed, a patented design ‘need not have any practical util-
 ity,’ the antithesis of a utility invention.” 

     Second, while Rosen may have overstated its point in
 adding to the quoted Jennings language such as that the
 primary reference must have design characteristics that
 are “basically the same” as those of the claimed design, this
 statement hardly reflects the rigidity the Court was con-
 demning in KSR. 

.
     And finally, Rosen was not essentially incorrect. In any
 obviousness analysis, the question is whether the claimed
 invention was obvious, but obvious over what. One has to
 start from somewhere. See K/S Himpp v. Hear-Wear
 Techs., 

 (Fed. Cir. 2014) (finding that
 an examiner may not rely on his common knowledge with-
 out a supporting prior art reference) (citing M.P.E.P.
 § 2144.03).
      In any event, whether Rosen was right or wrong, or
 over-simplified, the Supreme Court, in KSR, reviewing a
 decision involving a utility patent, did not address the basic
 concept of beginning an obviousness analysis in design pa-
 tents by looking for a “something in existence.” And, espe-
 cially, it did not overrule Rosen.
Case: 21-2348    Document: 48      Page: 20   Filed: 01/20/2023




         NOTE: This disposition is nonprecedential.


    United States Court of Appeals
        for the Federal Circuit
                   ______________________

   LKQ CORPORATION, KEYSTONE AUTOMOTIVE
             INDUSTRIES, INC.,
                 Appellants

                              v.

   GM GLOBAL TECHNOLOGY OPERATIONS LLC,
                    Appellee
             ______________________

                         2021-2348
                   ______________________

     Appeal from the United States Patent and Trademark
 Office, Patent Trial and Appeal Board in No. IPR2020-
 00534.
                  ______________________

 STARK, Circuit Judge, concurring in part and concurring in
 judgment.
     I agree with the majority’s conclusion that the Board
 had substantial evidence to conclude LKQ failed to show
 the ’625 patent was anticipated or would have been obvi-
 ous. Accordingly, I join the majority’s anticipation analysis
 as well as its substantive obviousness analysis. Where I
 part ways is on the issue of forfeiture. The majority con-
 cludes that LKQ did not forfeit its argument that KSR
 overruled our precedent’s requirement that a party claim-
 ing a design patent is obvious must, as the first step in its
Case: 21-2348    Document: 48      Page: 21   Filed: 01/20/2023




 2                                    LKQ CORPORATION v.
                     GM GLOBAL TECHNOLOGY OPERATIONS LLC


 invalidating efforts, identify a single reference providing
 “basically the same” visual impression as the patented de-
 sign. Because LKQ did not raise this issue before the
 Board, I would find it was forfeited. Accordingly, I do not
 believe we need to address the merits of LKQ’s KSR argu-
 ment. Therefore, I respectfully decline to join the relevant
 portions of the majority’s opinion, namely sections III.A
 and III.B.
                               I
      I would resolve this case without reaching the merits
 of LKQ’s KSR argument as I believe the argument was for-
 feited before the Board.
      As the majority notes, normally an issue is forfeited if
 it cannot “be said that the tribunal was fairly put on notice
 as to the substance of the issue.” Maj. Op. 10 (citing Nike
 Inc. v. Adidas AG, 

 (Fed. Cir. 2016))
 (internal citation and quotation marks omitted); see also
 Fresenius USA, Inc. v. Baxter Int’l, Inc., 

 (Fed. Cir. 2009) (“If a party fails to raise an argument
 before the trial court, or presents only a skeletal or unde-
 veloped argument to the trial court, we may deem that ar-
 gument [forfeited] on appeal.”); Wallace v. Dep’t of the Air
 Force, 

 (Fed. Cir. 1989) (“Ordinarily, ap-
 pellate courts refuse to consider issues not raised before an
 administrative agency.”). As we have previously stated, an
 issue being pressed on appeal “must be raised with suffi-
 cient specificity and clarity that the tribunal is aware that
 it must decide the issue.” Wallace, 

. Here,
 LKQ’s briefing before the Board was not nearly so specific
 and clear as to make the Board aware it had to decide the
 issue on which LKQ would now have us decide its appeal.
 In fact, to the contrary, the PTAB judges who heard this
 case would, I imagine, be surprised to see how this case has
 morphed into one about whether KSR overruled our “Rosen
 reference” requirement.
Case: 21-2348    Document: 48      Page: 22     Filed: 01/20/2023




 LKQ CORPORATION v.                                           3
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


                               A
      To demonstrate how unfair LKQ is being to the Board,
 I will first set out our current standard for evaluating de-
 sign patent obviousness challenges, the “Durling test.” See
 Durling v. Spectrum Furniture Co., 

 (Fed.
 Cir. 1996) (citing In re Rosen, 

 (CCPA
 1982)). The Durling test proceeds in two steps. First, what
 we have called a “Rosen reference” must be identified. To
 assess whether a Rosen reference exists, the reviewing tri-
 bunal must “(1) discern the correct visual impression cre-
 ated by the patented design as a whole; and (2) determine
 whether there is a single reference that creates ‘basically the
 same’ visual impression” as that patented design.” Durl-
 ing, 

. If a Rosen reference cannot be iden-
 tified, the obviousness analysis is done, and the validity
 challenge is defeated. See Rosen, 

. If, but
 only if, a Rosen reference exists, the analysis proceeds to a
 second step, in which “the trier of fact must determine
 whether, using secondary references, an ordinary designer
 would have modified the primary reference to create a de-
 sign that has the same overall visual appearance as the
 claimed design.” Campbell Soup Co. v. Gamon Plus, Inc.,
 

, 1275 (Fed. Cir. 2021). The secondary refer-
 ences relied on must be “so related [to the primary refer-
 ence] that the appearance of certain ornamental features
 in one would suggest the application of those features to
 the other.” Durling, 

 (internal citations and
 quotation marks omitted; alteration in original). If a sec-
 ondary reference does not meet this standard, it may not
 be relied on as part of an obviousness combination, and
 there is then no need even to consider whether a designer
 of ordinary skill would have been motivated to combine
 such a secondary reference with the Rosen primary refer-
 ence. See Rosen, 

.
     With this background, I will now repeat below the same
 excerpt from LKQ’s briefing before the Board that is in-
 cluded in the majority opinion—this also happens to be the
Case: 21-2348    Document: 48      Page: 23     Filed: 01/20/2023




 4                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 entirety of the briefing LKQ provided to the Board on the
 KSR issue. Thereafter, I will present what LKQ is arguing
 to us, in this appeal, about KSR’s impact on the Durling
 test.
     Here is what LKQ argued to the Board in its petition:
     [1] As a final matter, the Federal Circuit has, to
     date, been silent regarding whether the Supreme
     Court’s opinion in KSR Int’l Co. v. Teleflex, Inc., 

 (2007), governs design patent obvious-
     ness, other than to say that it was “not obvious that
     the Supreme Court necessarily intended to exclude
     design patents from the reach of KSR.” Titan Tire
     Corp. v. Case New Holland, Inc., 

,
     1384–85 (Fed. Cir. 2009). [2] In a design patent
     obviousness analysis, to modify a primary refer-
     ence, a secondary reference must be “so related [to
     the primary reference] that the appearance of cer-
     tain ornamental features in one would suggest the
     application of those features to the other.” Titan
     Tire, 

 (quoting Durling, 

). [3] To the extent that the “so related” test
     operates to unduly limit the scope of design patent
     obviousness, such an overly restrictive view would
     run afoul of KSR’s proscription against rigid re-
     strictions on the scope of an obviousness analysis.
 J.A. 174 (bracketed number alterations and emphasis
 added). 1 As can be seen, in sentence [1] LKQ correctly



     1    In the well-chosen words of GM, LKQ also “vaguely
 stated” (Ans. Br. 40) the following to the Board: “In KSR,
 the Supreme Court rejected the ‘rigid, mandatory formula’
 embodied in the Federal Circuit’s then-prevailing ‘teaching
 suggestion motivation’ requirement as improperly limiting
 the obviousness inquiry and, instead, prescribed a more
 flexible approach to obviousness and motivation to combine
Case: 21-2348    Document: 48      Page: 24    Filed: 01/20/2023




 LKQ CORPORATION v.                                          5
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 stated we have not definitively decided if KSR applies to
 design patent cases, in sentence [2] LKQ made a point ex-
 clusively directed to the second step of the Durling test, and
 in sentence [3] LKQ made clear that it challenged Federal
 Circuit doctrine only “to the extent” that the second step of
 our Durling test is too rigid for KSR.
     But here is how LKQ describes the issue presented to
 our court in this appeal:
     Whether, in finding that the claimed design was
     not obvious under 

:
     the Board erred in applying Rosen and demanding
     a single primary reference as a predicate to the ob-
     viousness analysis despite the Supreme Court’s re-
     jection of such bright-line rules in connection with
     the obviousness analysis in KSR Int’l Co. v. Tele-
     flex, Inc., 

 (2007) . . . .
 Op. Br. 6 (emphasis added). Elsewhere in its briefing to us,
 LKQ writes: “Rosen cannot survive KSR,” “The Rosen re-
 quirement to identify a single reference that is ‘basically
 the same’ as the claimed design is not consistent with
 KSR,” and “Rosen’s requirement that a design patent chal-
 lenger identify, as a predicate to the obviousness analysis,
 a ‘primary reference’ that is ‘basically the same’ as the
 claimed design represents an even more rigid bright-line



 prior art teachings.” J.A. 174. As even LKQ acknowledges,
 this statement—like other general statements LKQ made
 about the holding of KSR—does no more than imply that
 KSR maybe overruled unspecified aspects of our design pa-
 tent obviousness doctrine. See Rep. Br. 17 (citing J.A. 174-
 75, 205-06, 1124, 1126-27, 1132-33); see also 

 (LKQ ar-
 guing “its obviousness grounds and evidence” as presented
 to Board “assumed KSR may supersede Rosen and Durl-
 ing”) (first emphasis added).
Case: 21-2348    Document: 48      Page: 25    Filed: 01/20/2023




 6                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 barrier than the TSM test did.” Op. Br. 50, 54 (emphasis
 omitted and capitalization modified).
      That the KSR issue LKQ presses here was not fairly
 presented to the Board is evident by comparing the above
 excerpts from the briefing. In the three-sentence argument
 LKQ made to the Board, LKQ plainly and unambiguously
 challenged the post-KSR vitality of the Durling test’s “so
 related” second step, and only that portion of the Durling
 test. LKQ articulated no challenge to the Rosen reference
 requirement, remaining entirely silent as to whether this
 first step of the Durling test survives KSR. In this way,
 LKQ directed the Board away from having to assess
 whether the Rosen requirement remains viable.
     In this appeal, by contrast, LKQ completely shifts its
 focus to the first step of the Durling test, directing its at-
 tack squarely at our Rosen reference requirement. The re-
 peated, explicit arguments to us—that the need for a
 primary reference that is visually “basically the same” as
 the patented design cannot withstand the Supreme Court’s
 analysis in KSR—stands in striking contrast to the KSR
 challenge presented to the Board, which omitted any refer-
 ence to Rosen or the first step of the Durling test.
     A party’s articulation of its arguments tells the tribu-
 nal what issues it is being asked to decide, as well as what
 issues are not presented. I do not think the Board, or any
 tribunal we review, should, ordinarily, face the potential of
 being reversed on an issue no party asked it to address.
 Fairness to the tribunal of first instance—and fairness to
 the opposing party—requires that we hold parties to their
 obligation to raise arguments clearly at the first available
 opportunity. See In re Google Tech. Holdings LLC, 

, 864 (Fed. Cir. 2020) (“The very word ‘review’ pre-
 supposes that a litigant’s arguments have been raised and
 considered in the tribunal of first instance. To abandon
 that principle is to encourage the practice of ‘sandbagging’”
 suggesting or permitting, for strategic reasons, that the
Case: 21-2348    Document: 48      Page: 26    Filed: 01/20/2023




 LKQ CORPORATION v.                                          7
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 [lower tribunal] pursue a certain course, and later—if the
 outcome is unfavorable—claiming that the course followed
 was reversible error.”).
                               B
      Nor am I persuaded that the situation here is so excep-
 tional that we should exercise our discretion to address it.
 In Forshey v. Principi, 

 (Fed. Cir.
 2002) (en banc) (superseded by statute on other grounds),
 we enumerated four limited circumstances in which it is
 appropriate to consider forfeited issues: (1) “[w]hen new
 legislation is passed while an appeal is pending, courts
 have an obligation to apply the new law if Congress in-
 tended retroactive application even though the issue was
 not decided or raised below;” (2) “when there is a change in
 the jurisprudence of the reviewing court or the Supreme
 Court after consideration of the case by the lower court;”
 (3) “appellate courts may apply the correct law even if the
 parties did not argue it below and the court below did not
 decide it, but only if an issue is properly before the court;”
 and (4) “where a party appeared pro se before the lower
 court, a court of appeals may appropriately be less strin-
 gent in requiring that the issue have been raised explicitly
 below.”
     Only the third possibility is even potentially relevant
 here. Yet, Forshey makes clear that this exception is for
 situations when an issue is clearly presented but neither
 party argues for the legally correct answer. Forshey uses
 the example of the standard of proof. See 

. So,
 for example, if one party argues for preponderance of the
 evidence and the other argues for proof beyond a reasona-
 ble doubt, the court is free to apply the legally correct
 standard of clear and convincing evidence (in a case where
 this is, in fact, the correct standard). See 

 Here, the
 panel unanimously agrees that the Board committed no le-
 gal error in finding LKQ failed to identify a Rosen reference
 as required under Durling. Thus, the parties are not
Case: 21-2348    Document: 48     Page: 27    Filed: 01/20/2023




 8                                     LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 asking us to decide between adopting a forfeited position
 and applying the incorrect law. This is not a case in which
 affirming the Board requires us to adopt or approve of the
 application of a self-evidently incorrect legal standard. In-
 deed, as the majority opinion observes, the Board applied
 the legal standard we have applied in every one of the more
 than 50 post-KSR design patent cases we have decided.
      Our unbroken, and largely unquestioned, application
 of the Durling test in dozens of post-KSR design patent ob-
 viousness cases confirms, for me, that we have strong
 grounds for not exercising our discretion to reach out and
 decide the forfeited challenge in this case. The majority
 notes that the issue of the correctness of our current law in
 light of KSR has only been raised twice before this appeal,
 which suggests it may not actually be a “significant ques-
 tion[] of general impact.” Icon Health & Fitness, Inc. v.
 Strava, Inc., 

 (Fed. Cir. 2017). There
 also seems to be nothing about this case that makes avoid-
 ing the question of KSR’s impact result in a “miscarriage of
 justice.” See 

     For all these reasons, I would not exercise our discre-
 tion to decide an issue LKQ forfeited below. 2
                              C
     LKQ argues any forfeiture should be excused because
 the Board was powerless to decide that KSR abrogated the
 Rosen reference requirement. See Rep. Br. 20 (“Unlike this
 Court, the Board had no power to determine that Rosen or
 Durling had been overruled.”); 

 (citing Koninklijke
 Philips Elecs. N.V. v. Cardiac Sci. Operating Co., 

 (Fed. Cir. 2010); Merck & Co. v. Kessler, 

   I might view the discretionary decision as to
 whether to excuse the forfeiture differently if this case is
 reheard en banc. That, of course, is not an issue I need to
 decide today.
Case: 21-2348    Document: 48     Page: 28    Filed: 01/20/2023




 LKQ CORPORATION v.                                         9
 

 (Fed. Cir. 1996)). The authorities LKQ re-
 lies on for this proposition do not actually support the view
 that an administrative agency lacks the power to recognize
 that the Supreme Court has overruled what had been bind-
 ing circuit court precedent. In fact, the portion of Kon-
 inklijke quoted by LKQ, that “[j]udicial precedent is as
 binding on administrative agencies as are statutes,” 

, instead clarifies that the Board is required to
 follow KSR, just as this court must.
     I believe the Board—like a district court 3 or a panel of
 our court 4—has the power, indeed an obligation, to recog-
 nize when the Supreme Court has overruled previously
 binding precedent, and in such an instance (which is likely
 to be exceedingly rare) all of us are bound to follow the Su-
 preme Court, not the overruled circuit court caselaw. See
 Hyatt v. Heckler, 

 (4th Cir. 1986) (“The
 separation of powers doctrine requires administrative
 agencies to follow the law of the circuit whose courts have
 jurisdiction over the cause of action. In the absence of a
 controlling decision by the Supreme Court, the respective
 courts of appeals express the law of the circuit.”).
    If KSR truly overruled Durling, expressly or implicitly,
 then the Board was bound to follow KSR as controlling Su-
 preme Court precedent regardless of what our decisions


     3    “Ordinarily, a trial court may not disregard its re-
 viewing court’s precedent. There are two narrow excep-
 tions: if the circuit’s precedent is expressly overruled by
 statute or by a subsequent Supreme Court decision.”
 Strickland v. United States, 

, 1338 n. 3 (Fed.
 Cir. 2005) (internal citations omitted).
     4    “[W]e are bound by prior panel decisions until they
 are overruled by the court en banc or the Supreme Court.”
 Deckers Corp. v. United States, 

 (Fed. Cir.
 2014); see also Texas Am. Oil Corp. v. U.S. Dep’t of Energy,
 

 (Fed. Cir. 1995) (en banc).
Case: 21-2348    Document: 48      Page: 29   Filed: 01/20/2023




 10                                    LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 said. This is the same principle that allows LKQ to argue,
 as it does, that this panel can depart from Durling without
 needing to go en banc. Oral Arg. at 4:57-5:10. Moreover, it
 is clear that LKQ believed the Board was empowered to di-
 rectly apply KSR, as LKQ specifically argued step two of
 Durling had been modified, if not abrogated outright, and
 therefore asked the Board not to follow step two of Durling.
     Relatedly, the court’s opinion today excuses LKQ’s fail-
 ure to clearly raise the issue it now presses on appeal be-
 cause the argument would have been made “to an
 unreceptive audience,” citing a previous Board decision
 (which was nonprecedential) rejecting a KSR challenge to
 current law. Maj. Op. 12 (citing Johns Manville Corp. v.
 Knauf Insulation, Inc., 2017 Pat. App. LEXIS 3419, at *63-
 64 (P.T.A.B. Jan. 11, 2017)). I disagree with this reasoning,
 both because the Board is always bound to follow control-
 ling Supreme Court caselaw—notwithstanding its own
 prior rulings—and because we know LKQ understood its
 KSR contentions were not futile, since it chose to present
 one of them to the Board.
    Again, then, I would hold that LKQ forfeited the KSR
 argument it urges on us.
                              II
     Because my colleagues find no forfeiture, they address
 the impact of KSR on our existing obviousness doctrine
 with respect to design patents, concluding “it is not clear
 the Supreme Court has overruled Rosen or Durling.” Maj.
 Op. 13. I agree it is not clear.
     Despite the lack of clarity, I think it worth noting that
 there is, at minimum, substantial tension between the Su-
 preme Court’s holding in KSR and our Durling test. In
 KSR, 

, the Supreme Court explained that
 “when a court transforms [a] general principle into a rigid
 rule that limits the obviousness inquiry . . . it errs.” See
 also 

 (“Throughout this Court’s engagement with
Case: 21-2348    Document: 48      Page: 30    Filed: 01/20/2023




 LKQ CORPORATION v.                                         11
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


 the question of obviousness, our cases have set forth an ex-
 pansive and flexible approach . . . .”). A strong case can be
 made that the step one Rosen reference requirement is pre-
 cisely the type of limiting, rigid rule KSR faulted: if a de-
 sign patent challenger fails to identify a Rosen reference,
 the obviousness analysis stops. In this way, the Durling
 test can prevent the consideration of a combination of prior
 art references. See Durling, 

 (“Before one
 can begin to combine prior art designs, however, one must
 find a single reference, a something in existence, the design
 characteristics of which are basically the same as the
 claimed design.”) (emphasis added; internal quotation
 marks omitted). It also may deprive the tribunal of the op-
 portunity to consider other factors that often drive the
 analysis in utility patent obviousness cases. This all seems
 to conflict with KSR’s instruction that “[o]ften, it will be
 necessary for a court to look to interrelated teachings of
 multiple patents; the effects of demands known to the de-
 sign community or present in the marketplace; and the
 background knowledge possessed by a person having ordi-
 nary skill in the art.” 

.
     Similarly, KSR says that “a court can take account of
 the inferences and creative steps that a person of ordinary
 skill in the art would employ.” 

 But the requirement to
 identify a Rosen reference, and the consequence that the
 analysis proceeds no further if the challenger fails to do so,
 appears to prevent consideration of the ordinary designer’s
 creativity, at least in cases like the one before us now.
     Additional rigidity that might not be appropriate in a
 post-KSR world arguably exists at step two of the Durling
 test, which limits consideration solely to prior art refer-
 ences that are “so related” to the identified Rosen reference
 that “the appearance of certain ornamental features in one
 would suggest the application of those features to the
 other.” Durling, 

 (internal citations omit-
 ted). This is essentially the KSR challenge that LKQ pre-
 sented to the Board yet only presses generally on appeal.
Case: 21-2348    Document: 48     Page: 31    Filed: 01/20/2023




 12                                    LKQ CORPORATION v.
                      GM GLOBAL TECHNOLOGY OPERATIONS LLC


 See Op. Br. 51 (“[T]his restrictive, formulaic test, and par-
 ticularly its requirement that the court identify a single
 prior art reference . . . is inconsistent with KSR.”) (empha-
 sis added); see also KSR, 

 (“Rigid, preventa-
 tive rules that deny factfinders recourse to common sense,
 however, are neither necessary under our case law nor con-
 sistent with it.”). 5
     Accordingly, I can see how one might conclude that
 KSR implicitly overrules the Durling test. Alternatively, I
 can also see how one might persuasively argue that design
 patents are so fundamentally different than utility patents
 that they warrant their own test for obviousness, a test
 that the Supreme Court would likely allow to coexist along-
 side the KSR test we apply to utility patents. Judge Lourie,
 in his additional views, admirably sets out how such a con-
 clusion might be reached.




      5   Another potential problem in our design patent ob-
 viousness test concerns claim construction. At least as the
 Board applied the law here, it imposed on LKQ the obliga-
 tion to articulate “’the correct visual impression created by
 the patented design as a whole’” and held that LKQ’s fail-
 ure to do so was itself a sufficient basis on which to reject
 LKQ’s obviousness challenge. J.A. 50-51 (quoting Durling,
 

). Neither the Board—nor GM, in arguing
 for affirmance—cites any of our own cases supporting the
 proposition that a patent challenger’s failure to articulate
 the correct construction of a design patent is a dispositive
 failing. If that is a correct understanding of our caselaw,
 an issue on which I do not today express a view, it would
 certainly be an additional rigidity reserved only for design
 patents. No one would argue, I think, that KSR or our ob-
 vious standards for utility patents hold that an obviousness
 challenge fails at the point at which the challenger pro-
 poses an incorrect claim construction.
Case: 21-2348    Document: 48       Page: 32   Filed: 01/20/2023




 LKQ CORPORATION v.                                         13
 GM GLOBAL TECHNOLOGY OPERATIONS LLC


      This case does not require us to decide the continuing
 vitality of the Durling test post-KSR because, as I have ex-
 plained, LKQ forfeited the only specific challenge it makes
 here to that test—the step one Rosen reference require-
 ment—by not presenting it to the Board. If (or when) we
 have to confront this issue, it should be in a case in which
 we have thorough and thoughtful briefing on the differ-
 ences between design and utility patents, and most prefer-
 ably a case in which we have the benefit of the tribunal of
 first instance’s analysis. All of this, of course, is lacking
 here.
                              III
      In sum, in my view, LKQ forfeited the KSR issue it now
 raises on appeal. I would not reach the question of whether
 the Supreme Court’s KSR decision overruled the first part
 of our Durling test, i.e., our Rosen reference requirement.
 Instead, I would proceed directly from finding the issue for-
 feited to considering whether the Board’s application of the
 Durling test was supported by substantial evidence, which
 I agree it was. For these reasons, I join the majority’s opin-
 ion except for parts III(A) and III(B) and I concur in affirm-
 ing the Board.