In Re: Hodges

United States Court of Appeals
for the Federal Circuit
______________________

IN RE: KENNETH ANDREW HODGES,
Appellant
______________________

2017-1434
______________________

Appeal from the United States Patent and Trademark
Office, Patent Trial and Appeal Board in No. 12/906,222.
______________________

Decided: February 12, 2018
______________________

STEVEN R. LEBLANC, Dority & Manning, P.A., Green-
ville, SC, argued for appellant.

MEREDITH HOPE SCHOENFELD, Office of the Solicitor,
United States Patent and Trademark Office, Alexandria,
VA, argued for appellee Andrei Iancu. Also represented
by NATHAN K. KELLEY, THOMAS W. KRAUSE, PHILIP J.
WARRICK.
______________________

Before LOURIE, O’MALLEY, and WALLACH, Circuit Judges.
Opinion for the court filed by Circuit Judge O’MALLEY.
Opinion concurring-in-part and dissenting-in-part filed by
Circuit Judge WALLACH.
O’MALLEY, Circuit Judge.
2 IN RE: HODGES

Appellant Kenneth Andrew Hodges appeals from ex-
amination of U.S. Patent Application No. 12/906,222
(“’222 application”), in which the Patent Trial and Appeal
Board (“Board”) affirmed the examiner’s rejection of all
claims as anticipated and/or obvious. See Ex Parte Ken-
neth Andrew Hodges, No. 2014-009710, 2016 WL 6441834
(P.T.A.B. Oct. 28, 2016). For the following reasons, we
reverse the Board’s anticipation determinations, vacate
its obviousness determinations, and remand for further
proceedings.
I. BACKGROUND
A. The ’222 Application
Entitled “System and Method for Operating a Drain
Valve,” the ’222 application is directed to a valve assem-
bly for draining contaminants, condensation, and other
fluids that adversely affect the efficiency and function of a
pressurized system. As shown in Figure 1 below, the
claimed assembly includes a single “valve body” 12, which
defines an inlet seat 18 and an outlet seat 20 downstream
thereto; two valves 14 and 16; and a sensor 40 for measur-
ing fluid pressure between the valves:
IN RE: HODGES 3

J.A. 48–52, 60. The ’222 application explains that the
sensor generates a “signal” 42 reflective of the fluid pres-
sure in the chamber between the valves and transmits the
signal to an indicator 44, such as a pressure gauge.
J.A. 51. The signal can then be compared to a predeter-
mined limit to generate a control signal, which, in turn,
can be used to control or adjust one or more of the two
valves to drain fluid as appropriate. J.A. 51–52.
The ’222 application contains 20 claims, 1 but the par-
ties agree that the claims at issue on appeal stand or fall
with independent claim 1, which recites:
A drain valve comprising:
a. a valve body, wherein said valve body
defines an inlet seat and a first outlet seat
downstream of said inlet seat;
b. a first member, wherein said first mem-
ber has a first position in sealing engage-
ment with said first outlet seat and a
second position separated from said first
outlet seat;
c. a second member, wherein said second
member has a first location in sealing en-
gagement with said inlet seat; and
d. a sensor downstream of said inlet seat,
wherein said sensor generates a signal re-
flective of a pressure downstream of said
inlet seat.
J.A. 56 (emphases added).

1 During prosecution, Hodges canceled claim 4 and
added claim 21.
4 IN RE: HODGES

B. The Patent Office’s Prior Art Rejections
Two prior art references are at issue on appeal:
(1) U.S. Patent No. 5,531,241 to Rasmussen (“Rasmus-
sen”), and (2) U.S. Patent No. 3,262,464 to Frantz
(“Frantz”).
Rasmussen discloses “[a] condensate removal device”
that senses pressure levels within the system and purges
condensate in response to such levels. Rasmussen col. 2,
l. 41; see id. col. 2, ll. 42–54. In particular, a sensor within
the device measures pressure changes corresponding to
condensate levels in the reservoir. Id. col. 4, ll. 21–28.
When the sensor measures a high condensate level, it
provides a signal that causes a plunger to move away
from a valve seat, thereby opening a drain path for the
condensate. Id. col. 4, ll. 40–48, 49–52; see also id. fig.7
(depicting a pressurized system that includes, inter alia,
drain valve 10, inlet port 17A, and an unlabeled valve
above inlet port 17A).
Frantz discloses “valves for draining condensate from
pressurized reservoirs.” Frantz col. 1, ll. 11–12. One
valve is automatic and the other is manual, see id. col. 2,
ll. 16–19, and both are “[m]ounted at specified positions in
the body,” id. col. 2, l. 15. Although Frantz teaches that
“it ordinarily will be preferred to actuate the automatic
valve by fluid pressure,” id. col. 2, ll. 42–44, it also dis-
closes an “air-actuated automatic valve [that] must have
[a] second element and seat to enable it to limit the dura-
tion of the draining,” id. col. 2, ll. 49–51. In the air-
actuated version, the automatic valve “preferably is a
differential piston with its larger or enlarged head fixed to
or rigid with the stem.” Id. col. 2, ll. 51–54.
The examiner found that: (1) Rasmussen anticipates
claims 1–3, 5–8, and 15–20 of the ’222 application under
35 U.S.C. § 102(b), (2) Frantz anticipates claims 1, 2, 9,
10, 14–18, 20, and 21 under § 102(b), and (3) the two
references collectively render all claims obvious under
IN RE: HODGES 5

§ 103. 2 The Board affirmed the examiner’s rejections, and
Hodges appealed the Board’s ruling with respect to claims
1–3, 5–8, and 21 only. We have jurisdiction under 28
U.S.C. § 1295(a)(4)(A).
II. DISCUSSION
A. The Board’s Finding that Rasmussen Anticipates
the Claims Is Unsupported by Substantial Evidence
As described above, the Board affirmed the examiner’s
rejection of certain claims as anticipated by Rasmussen.
A prior art reference anticipates a patent’s claim when
the four corners of the document “describe every element
of the claimed invention, either expressly or inherently,
such that a person of ordinary skill in the art could prac-
tice the invention without undue experimentation.”
Spansion, Inc. v. Int’l Trade Comm’n, 629 F.3d 1331, 1356
(Fed. Cir. 2010) (internal quotation marks omitted).
“Anticipation is a question of fact that we review for
substantial evidence.” Blue Calypso, LLC v. Groupon,
Inc., 815 F.3d 1331, 1341 (Fed. Cir. 2016); see also Redline
Detection, LLC v. Star Envirotech, Inc., 811 F.3d 435, 449
(Fed. Cir. 2015) (noting that we review the Board’s “fac-
tual findings for substantial evidence and its legal conclu-
sions de novo”).

2 Congress amended §§ 102 and 103 when it passed
the Leahy-Smith America Invents Act (“AIA”). Pub. L.
No. 112-29, §§ 3(b)–(c), 125 Stat. 284, 285–87 (2011). But,
because the ’222 application was filed in October 2010 and
never contained a claim having an effective filing date on
or after March 16, 2013 (the effective date of the statutory
changes enacted in 2011), or a reference under 35 U.S.C.
§§ 120, 121, or 365(c) to any patent or application that
ever contained such a claim, pre-AIA §§ 102 and 103
apply. See id. § 3(n)(1), 125 Stat. at 293.
6 IN RE: HODGES

At issue on appeal is claim 1’s recitation of an “inlet
seat” that is “define[d]” by a “valve body.” The examiner
found that Rasmussen inherently discloses such an inlet
seat within an unlabeled valve depicted in Rasmussen’s
Figure 7 above inlet port 17A. See Hodges, 2016 WL
6441834, at *2 (summarizing, and tacitly agreeing with,
the examiner’s finding that Rasmussen discloses “an inlet
seat (within the valve above condensate inlet port 17A)”);
see also J.A. 117 (examiner stating in his final rejection
that Rasmussen’s unlabeled valve “would inherently
include a member (second member) that seals against a
seat (inlet seat)”); J.A. 200 (examiner stating the same in
a non-final rejection); J.A. 206 (examiner stating in a non-
final obviousness rejection that Rasmussen discloses an
“inlet seat (i.e., the seat inherent to the valve above [inlet
port] 17A)”). Hodges conceded this point during examina-
tion, and the Patent Office does not allege that the Board
erred in this regard. 3
Thus, as the Board noted, the propriety of the exam-
iner’s rejection turns on whether Rasmussen’s unlabeled
valve containing the inlet seat is “define[d]” by Rasmus-
sen’s valve body—i.e., whether it “can properly be consid-

3 See, e.g., J.A. 104 (Hodges stating during exami-
nation that Rasmussen discloses “a second member
(valve) that seals against an inlet seat (inside the valve)”);
J.A. 251 (in his brief to the Board, Hodges characterizing
the examiner as having found that Rasmussen discloses
an inlet seat “inside the unnumbered valve” “shown
schematically in Fig. 7 above [inlet port] 17A,” but not
contesting that finding); J.A. 256 n.1 (Hodges acknowledg-
ing in his brief to the Board that Rasmussen discloses an
“inlet seat in the valve above [inlet port] 17A”); Appellee’s
Br. 10 (Patent Office relying on Hodges’ concession that
Rasmussen discloses a sensor that is downstream “of the
inlet seat in the valve above [inlet port] 17A”).
IN RE: HODGES 7

ered to be a part of Rasmussen’s drain valve body.”
Hodges, 2016 WL 6441834, at *2. The Board found that it
could. In particular, the Board found that Rasmussen’s
unlabeled valve is “connected to, and therefore allow[s] or
prevent[s] flow into, inlet port 17A,” and that the seat of
the unlabeled valve would therefore “be ‘an internal part’
of and contained within the ‘outer casing’ of drain valve
10.” Id. That finding is unsupported by substantial
evidence.
As shown in the annotated version of Rasmussen’s
Figure 7 included in Hodges’ opening brief, the unlabeled
valve (shown in red above inlet port 17A) resides above
the housing 11 that contains the other valve components
(shown in yellow):

Appellant’s Br. 22 (citing Rasmussen fig.7 (annotated));
see also Appellee’s Br. 5 (annotating Rasmussen’s Fig-
ure 7 to point to the same valve); J.A. 118 (examiner’s
rejection pointing to the same valve). As shown, the
unlabeled valve—and, therefore, the inlet seat therein4—

4 The dissent states that we improperly “assume[]”
that Rasmussen’s inlet seat is located within the unla-
8 IN RE: HODGES

is not “an internal part” of and “contained within” the
outer casing of the drain valve. To the contrary, Figure 7
clearly shows that the valve is external to and outside
Rasmussen’s casing. Accordingly, the only permissible
factual finding that can be drawn from Rasmussen is that
the inlet seat within the unlabeled valve is not “define[d]”
by the “valve body,” as required by the claims. 5 See
Owens Corning v. Fast Felt Corp., 873 F.3d 896, 903 (Fed.
Cir. 2017) (reversing the Board’s obviousness determina-
tion that was contrary to the “one permissible factual
finding” that could be drawn from the record).
As sole support for its contrary findings, the Board
asserted that “the positioning of Rasmussen’s unlabeled
valve is similar to the positioning of [the ’222 applica-

beled valve, and that this assumption is inconsistent with,
and mutually exclusive of, the Board’s “assumption” that
“the inlet seat is located at inlet port 17A.” Dissent at 5
(emphasis added). As we note above, however, the Board
and examiner found that Rasmussen’s inlet seat is con-
tained within its unlabeled valve, which, in turn, is locat-
ed above and connected to inlet port 17A. And, as also
described above, the parties do not dispute that finding.
5 In fact, it seems that the unlabeled valve is not
part of Rasmussen’s disclosed invention at all. The valve
is not depicted in any of Rasmussen’s other figures—even
though those same figures depict inlet port 17A and an
upper main cylinder “housing,” see, e.g., Rasmussen
figs.1–2, 6—nor is it described in Rasmussen’s specifica-
tion. This lack of description suggests that the valve is
not itself part of Rasmussen’s valve assembly, and that it
is depicted in Figure 7 merely to show how the assembly
interfaces with an external fluid pressure system. See id.
col. 3, ll. 3–4 (“FIG. 7 is a graphic illustration of a high
condensate level within the valve system installed on a
fluid pressure system.”).
IN RE: HODGES 9

tion’s] second member 16, which extends away from valve
body 12, ostensibly to allow second member 16 to be
controlled.” Hodges, 2016 WL 6441834, at *2 (emphasis
added). The Board neither supported its assertion of
similarity, however, nor explained how the positioning of
the unlabeled valve in Rasmussen’s Figure 7 would ena-
ble a skilled artisan to “practice the invention without
undue experimentation.” Spansion, 629 F.3d at 1356
(internal quotation marks omitted). And, even if Rasmus-
sen’s unlabeled valve is ostensibly “similar to” the ’222
application’s second member 16 in some respects, it is
different in the only respect that is relevant to the claims
at issue—i.e., it is not an internal part of or contained
within the valve body. We therefore conclude that sub-
stantial evidence does not support the Board’s anticipa-
tion findings. 6
The Patent Office shoulders the burden during initial
examination of establishing that the examined claims are
anticipated. See 35 U.S.C. § 102 (“A person shall be
entitled to a patent unless[] . . . .”); cf. In re Oetiker, 977
F.2d 1443, 1445 (Fed. Cir. 1992) (“If examination at the
initial stage does not produce a prima facie case of un-
patentability, then without more the applicant is entitled
to grant of the patent.”). Because the Board failed to meet
that burden here, and because the only permissible find-
ing that can be drawn from Rasmussen is that it does not
disclose the claim limitation at issue, we reverse the
Board’s anticipation determination with respect to Ras-
mussen as to appealed claims 1–3 and 5–8. See Corning,
873 F.3d at 903; In re Chudik, 851 F.3d 1365, 1371, 1373–

6 We also note that the examiner stated in his
obviousness rejection that it is at least “arguable” that
Rasmussen does not disclose “a valve body which defines
an inlet seat.” Hodges, 2016 WL 6441834, at *5 (quoting
examiner); see also, e.g., J.A. 25.
10 IN RE: HODGES

75 (Fed. Cir. 2017) (reversing the Board’s anticipation
rejection during examination and concluding from the
prior art figures that the references do not disclose the
relevant limitations without modification, and noting
that, “[t]hough our review of an anticipation finding is
deferential, we have not hesitated to reverse the Board
when substantial evidence does not support its findings”);
In re Rambus, Inc., 753 F.3d 1253, 1258 (Fed. Cir. 2014)
(reversing the Board’s anticipation findings during reex-
amination based in part on our review of the prior art
figures); cf. Arendi S.A.R.L. v. Apple Inc., 832 F.3d 1355,
1366 (Fed. Cir. 2016) (reversing the Board’s obviousness
determination and noting that “this is not a case where a
more reasoned explanation than that provided by the
Board can be gleaned from the record”); In re Rouffet, 149
F.3d 1350, 1359 (Fed. Cir. 1998) (“Absent any proper
motivation to combine part of Levine’s teachings with
Freeburg’s satellite system, the [Patent Office’s] rejection
of Rouffet’s claim over these references was improper and
is reversed.”); Synopsys, Inc. v. ATopTech, Inc., 685
F. App’x 951, 955–56 (Fed. Cir. 2017) (reversing the
Board’s anticipation ruling where the portions of the prior
art reference relied upon by the Board did not support its
findings). 7

7 The dissent asserts that we may reverse the
Board’s patentability determinations in only two circum-
stances—i.e., when the Board (1) “committed legal error
and no further factual findings are required,” or (2) “made
erroneous factual findings and only one permissible
factual finding exists.” Dissent at 2–3 (internal quotation
marks omitted). Whether or not the dissent is correct
that these are the only two circumstances in which rever-
sal is permitted, its conclusion that this case does not fit
within the second circumstance is incorrect. As described
above, the Board’s finding that the inlet seat within
IN RE: HODGES 11

B. The Board’s Finding that Frantz Anticipates
the Claims Is Similarly Unsupported by
Substantial Evidence
The Board also affirmed the examiner’s determination
that Frantz anticipates the claims. In particular, the
Board agreed with the examiner that Frantz’s piston stem
14 8 and piston head 18 collectively constitute the claimed
“sensor” because they sense pressure insofar as they move
in response to the pressure applied thereto. Hodges, 2016
WL 6441834, at *3. The Board also adopted the examin-
er’s construction of “signal” as “an act, event, or the like
that causes or incites some action.” Id. Based on that
construction, the Board determined that Frantz’s piston
stem and head generate a “signal,” as required by the
claims, in the form of a mechanical force determined by
the pressure in the valve chambers. Id.
As an initial matter, we note that the Patent Office
does not attempt to defend the Board’s anticipation find-
ing as to Frantz on appeal. See Appellee’s Br. 2 n.5 (“The
USPTO is not addressing the Frantz anticipation rejec-
tion on appeal.”). This is for good reason, as the Board’s
anticipation finding is predicated on an erroneous con-
struction of “signal,” and Frantz does not disclose a sensor

Rasmussen’s unlabeled valve is contained in Rasmussen’s
housing is plainly contrary to the only permissible factual
finding that can be drawn from Rasmussen itself. This
case therefore fits squarely within the second circum-
stance identified by the dissent.
8 Although the Board refers to Frantz’s “stem 14” as
“piston 14,” see, e.g., Hodges, 2016 WL 6441834, at *3, we
use the term “stem” here for consistency with Frantz’s
disclosures, see Frantz col. 2, ll. 52–54 (“[T]he automatic
valve 6 preferably is a differential piston with its larger or
enlarged head 18 fixed or rigid with the stem 14 . . . .”
(emphases added)).
12 IN RE: HODGES

that generates and transmits a signal under any reasona-
ble construction of that term.
During examination, claim terms are given their
broadest reasonable interpretation consistent with the
specification as understood by those of ordinary skill in
the art. In re Am. Acad. of Sci. Tech Ctr., 367 F.3d 1359,
1364 (Fed. Cir. 2004). We review claim construction de
novo except for subsidiary factual findings based on
extrinsic evidence, which we review for substantial evi-
dence. Teva Pharm. USA, Inc. v. Sandoz, Inc., 135 S. Ct.
831, 841–42 (2015).
The ’222 application explains that the sensor “may
transmit the signal to an indicator, such as a pressure
gauge or alarm system, to provide a visual or audible
indication of the operability of the drain valve,” and that a
controller can “compare[] the signal to a predetermined
limit and generate[] a control signal based on this com-
parison.” J.A. 51 (numerals omitted). It is clear from this
description that the signal must at least be capable of
being compared to a “predetermined limit.”
The Board’s construction is inconsistent with this de-
scription. The Board construed “signal” as encompassing
“an act, event, or the like that causes or incites some
action,” which would encompass virtually any mechanical
component within the valve drain that moves in response
to the flow of fluid through the drain. The Board at-
tempted to justify its construction by noting that the ’222
application specification refers to a pressure gauge that
can provide a visual indication of the operability of the
drain valve, and concluded from that passage that the
recited signal can constitute “movement of the needle of
pressure gauge 44 in response to the sensed pressure.”
Hodges, 2016 WL 6441834, at *3–4 (citing J.A. 51). Such
a reading of the ’222 application is unreasonable. The
portion of the specification to which the Board cites states
that the sensor “generates a signal,” which, in turn, can
IN RE: HODGES 13

be transmitted to a pressure gauge “to provide a visu-
al . . . indication of the operability of the drain.” J.A. 51.
In other words, the recited “signal” causes the pressure
gauge to provide a visual indication. The passage does
not suggest, however, that the signal itself can constitute
“movement of” the pressure gauge’s needle “in response to
the sensed pressure,” as the Board concluded. Id.
The Board’s strained interpretation of “signal” is
therefore unreasonably broad and inconsistent with the
’222 application. As such, it does not accord with the
broadest reasonable interpretation standard. See In re
Smith Int’l, Inc., 871 F.3d 1375, 1382–83 (Fed. Cir. 2017)
(stating that “the Board cannot construe the claims so
broadly that its constructions are unreasonable under
general claim construction principles,” and that giving
claim terms “a strained breadth in the face of the other-
wise different description in the specification [is] unrea-
sonable” (internal quotation marks omitted));
TriVascular, Inc. v. Samuels, 812 F.3d 1056, 1062 (Fed.
Cir. 2016) (“While the broadest reasonable interpretation
standard is broad, it does not give the Board an unfet-
tered license to interpret the words in a claim without
regard for the full claim language and the written de-
scription.”).
Under any reasonable construction of “signal,”
Frantz’s piston stem and head combination cannot fairly
be characterized as a sensor that generates such a signal,
much less one that is “reflective of a pressure downstream
of [the] inlet seat,” as claimed by the ’222 application.
Indeed, the movement of Frantz’s piston cannot be com-
pared to a predetermined limit, as described in the ’222
application. The only permissible factual finding that can
be drawn from Frantz is that it does not disclose a sensor
that generates a “signal” under any reasonable construc-
tion of that term. Indeed, the Patent Office’s refusal to
defend the Board’s decision only underscores the lack of
evidence supporting the Board’s finding, and we will not
14 IN RE: HODGES

give the Patent Office a second chance to reject the claims
on grounds that it is unwilling or unable to defend on
appeal.
We therefore reverse the Board’s anticipation deter-
mination based on Frantz as to appealed claims 1–2 and
21. See Smith, 871 F.3d at 1382–84 (reversing the
Board’s anticipation finding predicated on an unreasona-
bly broad claim construction); In re Skvorecz, 580 F.3d
1262, 1267–68 (Fed. Cir. 2009) (reversing the Board’s
rejection of reissue claims predicated on an incorrect
construction, and noting that “[a]nticipation cannot be
found, as a matter of law, if any claimed element or
limitation is not present in the reference”).
C. The Board Did Not Provide an Adequate Evidentiary
Basis or Explanation for Its Determination that
the Claims Would Have Been Obvious
Finally, the Board determined that the claims would
have been obvious over Rasmussen in view of Frantz, see
Hodges, 2016 WL 6441834, at *5, and, alternatively,
Frantz in view of Rasmussen, see id. at *5–6. 9 We review
the Board’s obviousness determinations de novo and its
factual findings underlying those determinations for
substantial evidence. See In re Adler, 723 F.3d 1322,
1325 (Fed. Cir. 2013); In re Gartside, 203 F.3d 1305, 1316
(Fed. Cir. 2000).

9 At oral argument, Hodges stated that he does not
contest motivation to combine. Oral Arg. at 4:44–55,
In re Hodges (No. 2017-1434), http://oralarguments.cafc.
uscourts.gov/default.aspx?fl=2017-1434.mp3 (Q: “You
don’t dispute the findings with respect to the motivation
to combine Rasmussen and Frantz, right?” A: “No, I
don’t, Your Honor, because I dispute the prima facie—the
existence of all the elements, and it’s based on . . . antici-
pation.”).
IN RE: HODGES 15

The Board concluded in a single paragraph that the
claims would have been obvious over Rasmussen in view
of Frantz because the unlabeled valve depicted in Ras-
mussen’s Figure 7 could be made part of Rasmussen’s
drain valve body such that the seat of the unlabeled valve
would be “‘an internal part’ of and contained within the
‘outer casing’ of drain valve 10.” Hodges, 2016 WL
6441834, at *5. But the Board did not explain how Ras-
mussen’s drain valve assembly could be so modified. In
another paragraph, the Board agreed with the examiner
that it also “would have been obvious to a skilled artisan
to include Rasmussen’s sensor in Frantz’s valve body for
the purposes of generating a control signal that purges a
high condensate level from Frantz’ [sic] [main] chamber.”
Id. (internal quotation marks omitted). But the Board did
not explain how such a sensor would be downstream of
the inlet seat and generate a signal reflective of a pres-
sure downstream of said inlet seat, as required by the
claims. Instead, the Board merely determined that the
claims would have been obvious over Frantz in view of
Rasmussen “[f]or the same reasons as presented above.”
Id. The Board made no findings, moreover, regarding the
obviousness factors laid out in Graham v. John Deere Co.
of Kan. City, 383 U.S. 1 (1966).
Our review of the Board’s decision “is rooted not just
in the law of obviousness but in basic principles of admin-
istrative law.” Personal Web Techs., LLC v. Apple, Inc.,
848 F.3d 987, 992 (Fed. Cir. 2017). The Board must
therefore “explicate its factual conclusions, enabling us to
verify readily whether those conclusions are indeed sup-
ported by ‘substantial evidence’ contained within the
record.” Gartside, 203 F.3d at 1314; see also Power Inte-
grations, Inc. v. Lee, 797 F.3d 1318, 1326–27 (Fed. Cir.
2015) (holding that the Board’s reasoning must be set out
“in sufficient detail to permit meaningful appellate re-
view”). The Board did not do so here. Unlike the Board’s
anticipation determinations, which contravene the only
16 IN RE: HODGES

permissible findings that can be drawn from the prior art
under the proper constructions of the relevant claim
terms, the Board’s obviousness determinations involve
“potentially lawful but insufficiently or inappropriately
explained” factual findings, In re Van Os, 844 F.3d 1359,
1362 (Fed. Cir. 2017) (internal quotation marks omit-
ted)—i.e., that Rasmussen’s valve assembly could be
modified such that the unlabeled valve above inlet port
17A could become an internal part of the drain valve body
and upstream of the sensor and outlet seat, or, alterna-
tively, that Frantz’s drain valve could be modified so as to
include a sensor downstream of the inlet seat that gener-
ates a signal reflective of a pressure downstream of that
seat. The Board has failed to explain its reasoning to
allow us to determine whether such findings would be
lawful.
When faced with similarly deficient factual findings,
“we have consistently vacated and remanded for further
proceedings.” Id. (collecting cases); Icon Health & Fitness,
Inc. v. Strava, Inc., 849 F.3d 1034, 1046 (Fed. Cir. 2017)
(“Although the [Board] made a factual finding, this find-
ing did not have an adequate basis in the record, and the
[Board] did not adequately explain its reasoning. Thus,
we vacate and remand for additional [Board] findings and
explanation.”); Personal Web Techs., 848 F.3d at 994
(vacating and remanding “for the Board to reconsider the
merits of the obviousness challenge, within proper proce-
dural constraints”); see also Fla. Power & Light Co. v.
Lorion, 470 U.S. 729, 744 (1985) (explaining that, when
an agency’s fact-findings and explanation are deficient,
“the proper course, except in rare circumstances, is to
remand to the agency for additional investigation or
explanation”).
We therefore vacate the Board’s obviousness rejec-
tions with respect to appealed claims 1–3, 5–8, and 21,
and remand for further factual findings and explanation
on this score.
IN RE: HODGES 17

III. CONCLUSION
We have considered the parties’ remaining arguments
and find them unpersuasive. Accordingly, we reverse in
part and vacate in part the Board’s ruling, and remand
for proceedings consistent with this decision.
REVERSED IN PART, VACATED IN PART,
AND REMANDED
COSTS
No costs.
 United States Court of Appeals
for the Federal Circuit
______________________

IN RE: KENNETH ANDREW HODGES,
Appellant
______________________

2017-1434
______________________

Appeal from the United States Patent and Trademark
Office, Patent Trial and Appeal Board in No. 12/906,222.
______________________
WALLACH, Circuit Judge, concurring-in-part and dissent-
ing-in-part.
I agree with the majority’s decisions to: (1) reverse
the U.S. Patent and Trademark Office’s (“USPTO”) Pa-
tent Trial and Appeal Board’s (“PTAB”) determination
that U.S. Patent No. 3,262,464 (“Frantz”) anticipates the
asserted claims of U.S. Patent Application No. 12/906,222
(“the ’222 application”), see Maj. Op. 11–14; and (2) vacate
and remand the PTAB’s determination that the asserted
claims of the ’222 application would have been obvious
over U.S. Patent No. 5,531,241 (“Rasmussen”) in view of
Frantz, see id. at 14–16. However, I believe the majority
goes too far in reversing the PTAB’s anticipation finding
with respect to Rasmussen. See id. at 5–10.
When an agency fails to make requisite factual find-
ings or to explain its reasoning, “the proper course, except
in rare circumstances, is to remand to the agency for
additional investigation or explanation. The reviewing
court is not generally empowered to conduct a de novo
inquiry into the matter being reviewed and to reach its
2 IN RE: HODGES

own conclusions based on such an inquiry.” Fla. Power &
Light Co. v. Lorion, 470 U.S. 729, 744 (1985) (first empha-
sis added). By reversing the PTAB’s determination that
Rasmussen anticipates the asserted claims of the ’222
application, the majority engages in the very de novo
inquiry against which the Supreme Court has cautioned.
See Maj. Op. 5–10. Therefore, I respectfully dissent-in-
part.
DISCUSSION
I. This Court’s Authority to Reverse the PTAB’s Patenta-
bility Determinations Is Limited
In accordance with Supreme Court guidance, “when
the [PTAB]’s action is potentially lawful but insufficiently
or inappropriately explained, we have consistently vacat-
ed and remanded for further proceedings.” In re Van Os,
844 F.3d 1359, 1362 (Fed. Cir. 2017) (internal quotation
marks omitted) (collecting cases). Our precedent, howev-
er, has identified two “rare circumstances” where we may
reverse the PTAB’s ultimate patentability determina-
tions, generally because the relevant facts are undisputed.
Fla. Power & Light, 470 U.S. at 744. First, we may
reverse when the PTAB has committed legal error and no
further factual findings are required. See Ariosa Diagnos-
tics v. Verinata Health, Inc., 805 F.3d 1359, 1365, 1366
(Fed. Cir. 2015) (stating that we may only decide a matter
in the first instance when it is “purely legal” and that “we
must not ourselves make factual and discretionary de-
terminations that are for the agency to make” (citations
omitted) (emphasis added)); see also In re Smith Int’l, Inc.,
871 F.3d 1375, 1384 (Fed. Cir. 2017) (reversing where the
PTAB’s “findings depended on an incorrect claim con-
struction” and where “[i]t [was] undisputed that [the prior
art references] do not teach or render obvious the missing
elements”); Smith & Nephew, Inc. v. Rea, 721 F.3d 1371,
1380 (Fed. Cir. 2013) (reversing where the PTAB’s obvi-
ousness determination “was mainly the result
IN RE: HODGES 3

of . . . analytical errors . . . , not the [PTAB]’s resolution of
factual questions” and “the facts [we]re largely undisput-
ed” (footnote omitted)). Second, we may reverse when the
PTAB made erroneous factual findings and “only one
permissible factual finding” exists. Owens Corning v.
Fast Felt Corp., 873 F.3d 896, 903 (Fed. Cir. 2017); Arendi
S.A.R.L. v. Apple Inc., 832 F.3d 1355, 1366 (Fed. Cir.
2016) (reversing where “a more reasoned explanation
than that provided by the [PTAB]” was unavailable based
on the record); cf. In re Rouffet, 149 F.3d 1350, 1358–59
(Fed. Cir. 1998) (applying this principle to motivation to
combine).
II. The Majority Exceeds Its Appellate Authority in Re-
versing the PTAB’s Determination that Rasmussen
Anticipates the Asserted Claims of the ’222 Application
The majority attempts to fit its reversal of the PTAB’s
determination that Rasmussen anticipates the asserted
claims of the ’222 application into the second of the rare
circumstances allowing reversal. 1 See Maj. Op. 10 n.7. I

1 The majority bases its reversal of the PTAB’s de-
termination that Frantz anticipates the asserted claims of
the ’222 application on the first circumstance, i.e., legal
error. See Maj. Op. 12. I agree with that decision be-
cause, as in In re Smith International, the majority cor-
rectly reverses the PTAB’s claim construction, a legal
issue, see Trs. of Columbia Univ. v. Symantec Corp., 811
F.3d 1359, 1362 (Fed. Cir. 2016) (“Claim construction is
ultimately a question of law that this court reviews de
novo.” (citation omitted)), and the USPTO does not con-
test that Hodges prevails under the rejected claim con-
struction, see In re Smith Int’l, 871 F.3d at 1384; see also
Appellee’s Br. 2 n.5 (“The USPTO is not addressing the
Frantz anticipation rejection on appeal.”). However, had
the USPTO contested whether Frantz anticipates the
asserted claims under the proper construction, the PTAB
4 IN RE: HODGES

believe the majority exceeds its appellate authority by
making an unsupported factual finding in the first in-
stance and by failing to demonstrate that no other factual
finding would be “permissible.” Owens Corning, 873 F.3d
at 903.
As an initial matter, the sole support for the majori-
ty’s finding is an “annotated version of Rasmussen’s
Figure 7,” Maj. Op. 7, but this annotated version was
provided neither by the PTAB nor by a person having
ordinary skill in the art—it was provided by Appellant
Kenneth Andrew Hodges’ (“Hodges”) counsel and is not
included in the record. If only one factual finding were
“permissible,” Owens Corning, 873 F.3d at 903, the major-
ity would not need to rely on an extra-record annotated
figure created by Hodges’s counsel to support that finding,
see Icon Health & Fitness, Inc. v. Strava, Inc., 849 F.3d
1034, 1043 (Fed. Cir. 2017) (“Attorney argument is not
evidence.”); Axiom Res. Mgmt., Inc. v. United States, 564
F.3d 1374 (Fed. Cir. 2009) (“The focus of judicial review of
agency action remains the administrative record . . . .”).
Moreover, relying upon Hodges’s counsel’s annotated
version of Rasmussen Figure 7, the majority finds that
“the unlabeled valve—and, therefore, the inlet seat there-
in—is not ‘an internal part’ of and ‘contained within’ the
outer casing of the drain valve.” Maj. Op. 7–8 (footnote

would be required to make factual findings as to anticipa-
tion in the first instance. Because the USPTO did not do
so here, the majority’s fact findings in its anticipation
analysis under the proper construction are inappropriate.
See Maj. Op. 13 (finding, in the first instance, that
“Frantz’s piston stem and head combination cannot fairly
be characterized as a sensor that generates such a signal,
much less one that is ‘reflective of a pressure downstream
of [the] inlet seat’” and that “the movement of Frantz’s
piston cannot be compared to a predetermined limit”).
IN RE: HODGES 5

omitted). However, the majority provides no support for
that finding. 2 In contrast, in its equally-unsupported
factual findings, the PTAB found that “[t]he unlabeled
valve is illustrated in [Rasmussen] Figure 7 as being
connected to . . . inlet port 17A,” such that “the seat of the
unlabeled valve would be ‘an internal part’ of and con-
tained within the ‘outer casing’ of drain valve 10.” Hodg-
es, 2016 WL 6441834, at *2. Therefore, the majority
assumes that the inlet seat is located within the unla-
beled valve and that the entire unlabeled valve is outside
the outer casing of the drain valve, whereas the PTAB
assumed that the inlet seat is located at inlet port 17A

2 In apparent reference to its summary of the
PTAB’s and examiner’s findings, see Maj. Op. 6, the
majority states that “the [PTAB] and [E]xaminer found
that Rasmussen’s inlet seat is contained within its unla-
beled valve, which, in turn, is located above and connected
to inlet port 17A” and that “the parties do not dispute that
finding.” Maj. Op. 7 n.4. However, the citations provided
by the majority do not support this conclusion. The
Examiner never stated that the inlet seat was “within”
the unlabeled valve, see J.A. 117, 200, 206, and the PTAB
only stated as much in inaccurately summarizing the
Examiner’s findings, see Ex Parte Hodges, No. 2014-
009710, 2016 WL 6441834, at *2 (P.T.A.B. Oct. 28, 2016)
(“The [E]xaminer finds that Rasmussen . . . includ[es],
inter alia, . . . an inlet seat (within the valve above con-
densate inlet port 17A) . . . .” (footnote omitted)), as the
majority concedes, see Maj. Op. 6 (stating that the PTAB
was “summarizing, and tacitly agreeing with” the Exam-
iner’s finding (emphasis added)). Moreover, even if the
majority is correct that the inlet seat is “within” the
unlabeled valve, the PTAB never found that the unlabeled
valve consists only of the portions of the annotated ver-
sion of Rasmussen Figure 7 that are depicted in red, as
opposed to including protrusions that extend therefrom.
6 IN RE: HODGES

and that the entirety of inlet port 17A is within the outer
casing of the drain valve. Because neither the majority
nor the PTAB supported their assumptions, and each is
mutually exclusive of the other, I cannot join either find-
ing. See Personal Web Techs., LLC v. Apple Inc., 848 F.3d
987, 992 (Fed. Cir. 2017) (stating that “effective judicial
review” is premised upon record-supported factual find-
ings and accompanying analysis (internal quotation
marks and citation omitted)). Indeed, if only one of those
findings were permissible, Supreme Court guidance
dictates that it would be the PTAB’s and, ipso facto,
affirmance would be appropriate.
III. The Majority Errs in Its Reliance upon 35 U.S.C.
§ 102
The majority’s conclusion that “[t]he [USPTO] shoul-
ders the burden during initial examination of establishing
that the examined claims are anticipated,” such that
reversal is appropriate where the USPTO “fail[s] to meet
that burden” in the first instance, also gives me great
pause. Maj. Op. 9. As its primary support for this propo-
sition, the majority relies upon the language in 35 U.S.C.
§ 102 (2006)3 that “[a] person shall be entitled to a patent
unless—.” See id. The majority reads too much into
§ 102.

3 Congress amended § 102 when it passed the
Leahy-Smith America Invents Act (“AIA”). Pub. L. No.
112-29, § 3(b)(1), 125 Stat. 284, 285–87 (2011). However,
because the ’222 application never contained a claim
having an effective filing date on or after March 16, 2013
(the effective date of the statutory changes enacted in
2011), or a reference under 35 U.S.C. §§ 120, 121, or
365(c) to any patent or application that ever contained
such a claim, the pre-AIA § 102 applies. See id. § 3(n)(1),
125 Stat. at 293.
IN RE: HODGES 7

First, over dissenting opinions, multiple panels of this
court have rejected the proposition that § 102 mandates
reversal without remand when an unpatentability deci-
sion by the PTAB lacks the requisite factual findings or
explanation. 4 Compare L.A. Biomed. Research Inst. at
Harbor-UCLA Med. Ctr. v. Eli Lilly & Co., 849 F.3d 1049,
1068 n.9 (Fed. Cir. 2017) (“Although the dissent urges
that we should resolve the obviousness question without a
remand, it would be improper for us to do so in the ab-
sence of the necessary factual findings by the [PTAB].”
(citations omitted)), Icon, 849 F.3d at 1049 (vacating and
remanding for failure to make the requisite factual find-
ings and attendant explanation), and In re Van Os, 844
F.3d at 1362 (vacating and remanding for further expla-
nation), with L.A. Biomed., 849 F.3d at 1072 (Newman, J.,
concurring-in-part and dissenting-in-part) (“It was ex-
pected that in the normal course questions of patentabil-
ity under [§§] 102 and 103 would be reliably and speedily
resolved. . . . I don’t say that remands are never appro-
priate, but remands should be rare.”), Icon, 849 F.3d at
1049 (O’Malley, J., concurring-in-part and dissenting-in-
part) (“I write separately because I believe that remand is
not the appropriate remedy in examination appeals in
which the [US]PTO has not carried its burden of estab-
lishing unpatentability.” (internal quotation marks and
citation omitted) (citing 35 U.S.C. § 102)), and In re Van
Os, 844 F.3d at 1362 (Newman, J., concurring-in-part and
dissenting-in-part) (similar). We are bound by prior
precedent. See Deckers Corp. v. United States, 752 F.3d
949, 959 (Fed. Cir. 2014) (“[A] later panel is bound by the
determinations of a prior panel, unless relieved of that

4 Although these opinions concerned obviousness
under 35 U.S.C. § 103 rather than anticipation, the dis-
senting judges appear to contend that § 102 places the
burden on the USPTO for both obviousness and anticipa-
tion determinations.
8 IN RE: HODGES

obligation by an en banc order of the court or a decision of
the Supreme Court.”).
Second, even if § 102 did place the burden on the
USPTO, nothing in § 102 suggests that the USPTO could
not satisfy the burden on remand. Therefore, before
concluding that § 102 prohibits the PTAB from satisfying
its burden on remand, the court would need to consider
the interplay between the relevant statutory schemes—in
light of the administrative law standards in Title 5, the
judicial review function in Title 28, and the patent princi-
ples of Title 35 of the U.S. Code—an inquiry that should
be undertaken with the assistance of briefing by the
parties. Otherwise, we risk ruling that the introductory
clause of § 102 supersedes the general rule that, when an
appellate court is tasked with reviewing deficient agency
decisions, “the proper course, except in rare circumstances,
is to remand to the agency for additional investigation or
explanation,” Fla. Power & Light, 470 U.S. at 744 (em-
phasis added), and the Supreme Court has repeatedly
cautioned against the creation of such patent-specific
rules, see, e.g., SCA Hygiene Prods. Aktiebolag v. First
Quality Baby Prods., LLC, 137 S. Ct. 954, 964 (2017)
(citing approvingly the proposition that “patent law is
governed by the same common-law principles, methods of
statutory interpretation, and procedural rules as other
areas of civil litigation” (internal quotation marks, altera-
tions, and citation omitted)); Halo Elecs., Inc. v. Pulse
Elecs., Inc., 136 S. Ct. 1923, 1934 (2016) (cautioning the
Federal Circuit against applying a “rigid formula” and
instructing that we should focus on “the longstanding
considerations [the Supreme Court] ha[s] identified as
having guided both Congress and the courts”).
CONCLUSION
By reversing the PTAB’s determination that Rasmus-
sen anticipates the asserted claims of the ’222 application,
the majority departs from the default rule that deficient
IN RE: HODGES 9

agency decisions should be vacated and remanded. In
doing so, the majority improperly acts as the fact-finder
and dramatically over-reads § 102. For these reasons, I
respectfully dissent from that portion of the majority’s
analysis.