iRobot Corporation

Patent Trials and Appeals Board

Oct 4, 2021

IPR2020-00733 (P.T.A.B. Oct. 4, 2021)

Trials@uspto.gov Paper 31
571-272-7822 Date: October 4, 2021

UNITED STATES PATENT AND TRADEMARK OFFICE

BEFORE THE PATENT TRIAL AND APPEAL BOARD

SHARKNINJA OPERATING LLC,
SHARKNINJA MANAGEMENT LLC, AND
SHARKNINJA SALES COMPANY,
Petitioner,

v.

IROBOT CORPORATION,
Patent Owner.

IPR2020-00733
Patent 9,550,294 B2

Before TERRENCE W. MCMILLIN, AMANDA F. WIEKER, and
JASON W. MELVIN, Administrative Patent Judges.
WIEKER, Administrative Patent Judge.

JUDGMENT
Final Written Decision
Determining All Challenged Claims Unpatentable
35 U.S.C. § 318(a)

IPR2020-00733
Patent 9,550,294 B2

2
I. INTRODUCTION
A. Background and Summary
SharkNinja Operating LLC, SharkNinja Management LLC, and
SharkNinja Sales Company (“Petitioner”) filed a Petition requesting inter
partes review of claims 1–13 (“challenged claims”) of U.S. Patent
No. 9,550,294 B2 (Ex. 1001, “the ’294 patent”). Paper 1 (“Pet.”). iRobot
Corporation (“Patent Owner”) filed a Preliminary Response. Paper 6
(“Prelim. Resp.”).1 We instituted an inter partes review of all challenged
claims 1–13 on all grounds of unpatentability, pursuant to 35 U.S.C. § 314.
Paper 11 (“Dec. on Inst.”).
After institution, Patent Owner filed a Response (Paper 19, “PO
Resp.”) to the Petition, Petitioner filed a Reply (Paper 21, “Reply”), and
Patent Owner filed a Sur-reply (Paper 25, “Sur-reply”). An oral hearing was
held on July 12, 2021, and a transcript of the hearing is included in the
record. Paper 30 (“Tr.”).
We issue this Final Written Decision pursuant to 35 U.S.C. § 318(a)
and 37 C.F.R. § 42.73. For the reasons set forth below, Petitioner has met
its burden of showing, by a preponderance of the evidence, that challenged
claims 1–13 of the ’294 patent are unpatentable.
B. Real Parties in Interest
Petitioner identifies SharkNinja Operating LLC, SharkNinja
Management LLC, SharkNinja Sales Company, and JS Global Lifestyle Co.

1 Prior to institution and with our authorization, Petitioner filed a
Preliminary Reply and Patent Owner filed a Preliminary Sur-reply, directed
to 35 U.S.C. §§ 312(a)(2), 325(d), and claim construction. Papers 7, 9.
IPR2020-00733
Patent 9,550,294 B2

3
Ltd. as real parties in interest. Pet. 73; Paper 13, 1; see also Dec. on
Inst. 43–47 (addressing real party in interest arguments).
Patent Owner identifies iRobot Corporation as the real party in
interest. Paper 4, 2.
C. Related Matters
The parties identify the following matters related to the ’294 patent:
SharkNinja Operating LLC v. iRobot Corporation,
No. 1:19-cv-01935 (D. Del.) (filed Oct. 11, 2019);
iRobot Corporation v. SharkNinja Operating LLC, et al.,
No. 1:19-cv-12125 (D. Mass.) (filed Oct. 15, 2019); and
SharkNinja Operating LLC v. iRobot Corporation,
No. 1:19-cv-12236 (D. Mass.) (filed Oct. 30, 2019).
Pet. 74; Paper 4, 2.
D. The ’294 Patent
The ’294 patent is titled “Autonomous Robot Auto-Docking and
Energy Management Systems and Methods,” and issued on January 24,
2017, from U.S. Patent Application No. 14/946,999, filed November 20,
2015. Ex. 1001, codes (21), (22), (45), (54).
The ’294 patent discloses an energy management system for a robotic
device. Id. at code (57). Figures 1 and 2A are reproduced below.

IPR2020-00733
Patent 9,550,294 B2

4
Figure 1 (above left) depicts a base station, and Figure 2A (above right)
depicts a perspective view of a robot. Id. at 4:64–67.
As shown in Figure 1, base station 10 includes electrical charging
contacts 16 that engage corresponding contacts on the bottom of the robot, to
charge the robot’s battery. Id. at 5:24, 5:35–38, 5:49–51. Base station 10
includes top signal emitter 18 and front signal emitter 20, which project, e.g.,
avoidance signals or homing signals to the robot. Id. at 6:4–36.
As shown in Figure 2A, autonomous robot 40 includes chassis 44,
cover 46, and bumper 48, and the robot is configured to mate with base
station 10 for charging. Id. at 6:37–39. When “bumper 48 encounters a
stationary object or obstacle of predetermined mass,” the bumper displaces
toward the chassis and the robot evades the object or obstacle. Id. at 7:1–14.
Robot 40 also includes detectors 50, 52 for receiving signals projected by the
base station’s emitters 18, 20. Id. at 7:15–18. Robot 40 includes, e.g.,
motors, wheels, drive shafts, or tracks, to move the robot. Id. at 8:5–11.
The robot also includes a battery (unlabeled). Id. at 13:60–61.
In operation, “robotic device 40 uses a variety of behavioral modes to
vacuum effectively a working area.” Id. at 8:26–27; see also id. at 8:26–
11:26 (providing detail of modes). During operation, the energy level of the
robot’s battery is measured. Id. at 15:37–43. When the robot has completed
vacuuming a room, or when its battery needs to be recharged, the robot will
seek the base station by detecting signals projected from the base station,
which allow the robot to dock accurately with the base station. Id. at 13:58–
65, 14:26–27. Once docked, the robot autonomously recharges its battery,
either fully or partially. Id. at 16:46–47, 18:59–19:3. Additionally, “the
IPR2020-00733
Patent 9,550,294 B2

5
robot may employ room-mapping programs, using the base station 10 and/or
walls and large objects as points of reference.” Id. at 16:41–43.
E. Illustrative Claim
Of the challenged claims, claim 1 is independent. Claim 1 is
illustrative and is reproduced below.
1. [a] An autonomous cleaning robot, comprising:
[b] an undercarriage;
[c] a motive system configured to propel the
undercarriage across a ground surface;
[d] an energy storage unit supported by the undercarriage
and configured to be charged while the robot is positioned at a
base charging station; and
[e] a navigational control system configured to
autonomously:
[f] control the motive system to direct the robot
about a room, while mapping the room with respect to
objects as points of reference;
[g] return the robot to the base charging station
before completing a cleaning task in the room, in
response to detecting a need to recharge the energy
storage unit;
[h] dock the robot to the base charging station
upon return;
[i] recharge the energy storage unit with the robot
docked; and then to
[j] control the motive system to direct the robot to
continue to clean the room.
Ex. 1001, 19:25–45 (alphabetical identifiers and emphasis added).
F. Evidence
Petitioner relies upon the following references:
Kopmann, U.S. Patent No. 4,677,363, filed Feb. 26,
1986, issued June 30, 1987 (Ex. 1010, “Kopmann”);
IPR2020-00733
Patent 9,550,294 B2

6
Kim, U.S. Patent No. 5,440,216, filed June 8, 1993,
issued Aug. 8, 1995 (Ex. 1003, “Kim”);2
Han, U.S. Patent No. 5,646,494, filed Mar. 3, 1995,
issued July 8, 1997 (Ex. 1005, “Han”);
Jeong, U.S. Patent No. 5,896,488, filed Nov. 27, 1996,
issued Apr. 20, 1999 (Ex. 1004, “Jeong”);
Fiegert et al., U.S. Patent No. 6,240,342 B1, filed Feb. 3,
1999, issued May 29, 2001 (Ex. 1006, “Fiegert”); and
Ohsawa et al., JP 2001-125641 A, filed Oct. 29, 1999,
published May 11, 2001 (Ex. 1008, “Ohsawa”).3
Pet. 2–3.
Petitioner relies upon, inter alia, the Declaration of Alonzo J. Kelly,
Ph.D. Ex. 1020. Patent Owner relies upon, inter alia, the Declaration of
Donald A. Coates, Ph.D (Ex. 2029), the November 4, 2019, deposition of
Dr. Kelly (Ex. 2028), and the January 12, 2021, deposition of Dr. Kelly
(Ex. 2027).

2 Petitioner identifies several references by a combination of Applicant-
Inventor name. See Pet. v (identifying short form names for Exs. 1003–1005
as “Samsung-Kim,” “Samsung-Jeong,” and “Samsung-Han”). In this
Decision, we identify all references by first named Inventor.
3 Exhibit 1007 is the Japanese-language publication of Ohsawa, and
Exhibit 1008 is its English-language translation. See also Ex. 1009
(certificate of translation). In this Decision, we cite to Exhibit 1008.
Petitioner identifies this reference by Applicant name, “Sony.” Pet. v.
IPR2020-00733
Patent 9,550,294 B2

7
G. Asserted Grounds of Unpatentability
Petitioner asserts that claims 1–13 are unpatentable based upon the
following grounds (Pet. 2–3):
Claim(s)
Challenged
35 U.S.C. § Reference(s)/Basis
1–5, 8, 9, 11–13 § 102 Kim
1–5, 8, 9, 11–13 § 103 Kim
1–5, 8, 9, 11–13 § 103 Kim, Jeong
2, 4, 5 § 103 Kim, Fiegert
2, 4, 5 § 103 Kim, Jeong, Fiegert
3 § 103 Kim, Han
3 § 103 Kim, Jeong, Han
6, 7 § 103 Kim, Ohsawa
6, 7 § 103 Kim, Jeong, Ohsawa
10 § 103 Kim, Kopmann
10 § 103 Kim, Jeong, Kopmann

II. ANALYSIS
A. Legal Standards
“Both anticipation under [35 U.S.C.] § 102 and obviousness under
[35 U.S.C.] § 103 are two-step inquiries. The first step in both analyses is a
proper construction of the claims. . . . The second step in the analyses
requires a comparison of the properly construed claim to the prior art.”
Medichem, S.A. v. Rolabo, S.L., 353 F.3d 928, 933 (Fed. Cir. 2003).
A claim is unpatentable under 35 U.S.C. § 102 if a prior art reference
discloses every limitation of the claimed invention, either explicitly or
inherently. Glaxo Inc. v. Novopharm Ltd., 52 F.3d 1043, 1047 (Fed. Cir.
1995). To establish an inherent disclosure, the extrinsic evidence “must
make clear that the missing descriptive matter is necessarily present” in the
IPR2020-00733
Patent 9,550,294 B2

8
single anticipating reference. Cont’l Can Co. v. Monsanto Co., 948 F.2d
1264, 1268 (Fed. Cir. 1991).
A claim is unpatentable under 35 U.S.C. § 103 if “the differences
between the subject matter sought to be patented and the prior art are such
that the subject matter as a whole would have been obvious at the time the
invention was made to a person having ordinary skill in the art to which said
subject matter pertains.” KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 406
(2007). The question of obviousness is resolved on the basis of underlying
factual determinations, including (1) the scope and content of the prior art;
(2) any differences between the claimed subject matter and the prior art; (3)
the level of skill in the art; and (4) objective evidence of non-obviousness.4
Graham v. John Deere Co., 383 U.S. 1, 17–18 (1966). When evaluating a
combination of teachings, we must also “determine whether there was an
apparent reason to combine the known elements in the fashion claimed by
the patent at issue.” KSR, 550 U.S. at 418 (citing In re Kahn, 441 F.3d 977,
988 (Fed. Cir. 2006)). Whether a combination of prior art elements would
have produced a predictable result weighs in the ultimate determination of
obviousness. Id. at 416–417.
In an inter partes review, the petitioner must show with particularity
why each challenged claim is unpatentable. Harmonic Inc. v. Avid Tech.,
Inc., 815 F.3d 1356, 1363 (Fed. Cir. 2016); 37 C.F.R. § 42.104(b). The
burden of persuasion never shifts to Patent Owner. Dynamic Drinkware,
LLC v. Nat’l Graphics, Inc., 800 F.3d 1375, 1378 (Fed. Cir. 2015).
We analyze the challenges presented in the Petition in accordance with the
above-stated principles.

4 Patent Owner has not presented objective evidence of non-obviousness.
IPR2020-00733
Patent 9,550,294 B2

9
B. Level of Ordinary Skill in the Art
Petitioner identifies the appropriate level of skill in the art as a person
having “at least a four year degree in mechanical engineering or electrical
engineering or a closely related field and at least one year of experience in
the design and implementation of robotics and embedded systems,” where
additional education could substitute for experience, or vice versa. Pet. 4
(citing Ex. 1020 ¶¶ 50–54). Patent Owner does not provide an assessment of
the appropriate skill level. See generally PO Resp.
We adopt Petitioner’s assessment, which appears consistent with the
level of skill reflected in the Specification and prior art.
C. Claim Construction
For petitions filed on or after November 13, 2018, a claim shall be
construed using the same claim construction standard that would be used to
construe the claim in a civil action under 35 U.S.C. § 282(b). 37 C.F.R.
§ 42.100(b) (2020). Accordingly, claim terms are generally given their
ordinary and customary meaning as would have been understood by one
with ordinary skill in the art in the context of the specification, the
prosecution history, other claims, and even extrinsic evidence including
expert and inventor testimony, dictionaries, and learned treatises, although
extrinsic evidence is less significant than the intrinsic record. Phillips v.
AWH Corp., 415 F.3d 1303, 1312–1317 (Fed. Cir. 2005). Usually, the
specification is dispositive, and it is the single best guide to the meaning of a
disputed term.
The parties address the claim limitation “mapping the room with
respect to objects as points of reference,” and dispute the meaning of
“objects” within that limitation. Ex. 1001, 19:35–36; Pet. 5–7; PO
IPR2020-00733
Patent 9,550,294 B2

10
Resp. 10–34; Reply 2–15; Sur-Reply 1–15. Petitioner contends that the
’294 patent uses “objects” consistent with its plain meaning—“something
material that may be perceived by the senses”—which includes walls.
Pet. 5–7 (citing, e.g., Exs. 1014, 1015); Ex. 1014, 4. By contrast, Patent
Owner contends that “the ’294 patent distinguishes between obstacles that
define a work area and obstacles within the defined work area, referring to
the former as ‘walls’ . . . and the latter as ‘objects,’ ‘obstacles,’ or
‘obstructions’ that are located within the mapped area.” PO Resp. 10–11.
Accordingly, Patent Owner argues that the claim “requires a map that
specifies objects within the perceived boundaries of the room as points of
reference.” Id. at 10, 11–12.
In our Decision on Institution, we preliminarily determined that “walls
are within the scope of the term [‘objects’].” Dec. on Inst. 13–14; see also
id. at 7–15. We also encouraged the parties to address several related issues,
including whether further explicit construction of “objects” is required;
whether disclosures directed to “obstacles” should inform our construction
of the claim language; whether the pertinent evidence is limited to room
mapping; and whether a distinction between objects that define a working
area as opposed to objects within a working area is supported. Id. at 14–15.
1. The Claimed Invention
Patent Owner notes that our claim construction inquiry seeks “the
correct construction—the construction that most accurately delineates the
scope of the claim[ed] invention.” PPC Broadband, Inc. v. Corning Optical
Comm’ns RF, LLC, 815 F.3d 734, 740–42 (Fed. Cir. 2016) (cited at PO
Resp. 10; Sur-reply 12). With this in mind, we assess the invention
disclosed and claimed in the ’294 patent. See supra § I.D.
IPR2020-00733
Patent 9,550,294 B2

11
The ’294 patent discloses a method for “energy management” of an
autonomous robot. Ex. 1001, codes (54), (57). Specifically, “[t]he present
invention relates generally to robotic systems and, more specifically, to auto-
docking and energy management systems for autonomous robots.” Id. at
1:27–29. The invention seeks to avoid known problems related to safe and
effective robot charging, namely, problems with “proper mating” and
inadvertent collisions between the robot and a charging base station. Id. at
1:52–2:25. The ’294 patent discloses various energy management features,
and the Specification’s focus on energy management also appears in the
claims, where independent claim 1 and dependent claims 2 and 4–13 are
largely drawn to energy management. See, e.g., id. at 2:29–4:62
(summarizing the invention as, e.g., determining battery quantity and
performing tasks based on battery level, returning to base station when
quantity is low), 19:25–48; see supra § I.E (limitations d, g, h, i, j).
Contrary to this facial focus on energy management, Patent Owner
describes the ’294 patent as providing
an improvement over prior art cleaning techniques by providing
a robot that maps a room defined by ‘walls’ (i.e., obstacles that
define a working area) using ‘objects,’ which include obstacles
distinct from ‘walls’ (i.e., obstacles within the defined working
area), as points of reference and uses the map to direct the robot
about the room to thereby enhance both the robot’s cleaning and
its recharging efficiency.
PO Resp. 4; see also id. at 5–8. Patent Owner’s support for this alleged
“improvement” in mapping rests primarily in the Background section of the
’294 patent, which purports to incorporate “[a]dditional robot control and
navigation systems,” although the Background does not mention “mapping”
at all. Ex. 1001, 1:47–51; see infra § II.C.1.c. To be sure, however,
IPR2020-00733
Patent 9,550,294 B2

12
mapping is recited in one limitation of independent claim 1 and in dependent
claim 3. Ex. 1001, 19:35–36, 20:1–4; see supra § I.E (limitation f).
Considering the ’294 patent as a whole, we discern that the disclosed
and claimed invention is primarily an energy management system that
optimizes the safe and effective charging of a robot. It is within this context
that we resolve “the correct construction” of “objects,” i.e., “the construction
that most accurately delineates the scope of the claim[ed] invention.” PPC
Broadband, 815 F.3d at 740–42.
a) ’294 Claims
Claim 1 recites “mapping the room with respect to objects as points of
reference,” and dependent claim 3 further recites that “a known position of
the base charging station” is used as a point of reference in mapping.
Ex. 1001, 19:35–36, 20:1–4. The claim language and context provide little
insight to the term “objects,” except to confirm that “objects” at least
encompasses the recited “base charging station” of claim 3.
b) ’294 Specification
The ’294 patent describes mapping only once: “the robot may employ
room-mapping programs, using the base station 10 and/or walls and large
objects as points of reference.” Id. at 16:41–43. Patent Owner argues that
“by separately calling out ‘objects’ in addition to ‘walls’ in the context of
mapping, this sentence is certainly consistent with an interpretation of the
‘objects’ in the recited ‘mapping’ feature being distinct from ‘walls.’” PO
Resp. 20. Patent Owner argues that when read in the context of the
remainder of the ’294 patent, a person of ordinary skill in the art would find
this sentence to support Patent Owner’s construction. Id. at 20–21.
IPR2020-00733
Patent 9,550,294 B2

13
In our view, this sentence simply indicates various structures—base
stations, walls, large objects—that may serve as points of reference when
mapping. This language does not speak to whether these items are inclusive
or mutually exclusive, i.e., whether “large objects” excludes walls. Thus,
this sentence provides little substantive evidence as to the meaning of
“objects.”
We also disagree that the broader “context” of the Specification
supports Patent Owner’s construction. Id. at 20–21. Patent Owner does not
identify any other portion of the ’294 Specification itself that describes
“mapping the room with respect to objects.” PO Resp. 10–34.5 Nor does
Patent Owner identify any portion of the ’294 Specification that
distinguishes between objects that define a working area (e.g., walls) and
objects within a working area (e.g., furniture). See, e.g., id. at 20–32. Given
the ’294 patent’s focus on energy management—not mapping—this is
unsurprising.
The ’294 patent utilizes “objects” in other instances apart from its
singular discussion of mapping. See, e.g., Ex. 1001, 7:8 (“encounters a
stationary object or obstacle”), 7:50 (“a solid overhanging object”), 9:10
(“evade a stationary object or obstacle”), 12:43 (“keep all objects to its
right”), 17:1 (“electrically conductive objects”).6 These usages of “objects,”
however, appear consistent with the term’s plain meaning as identified by
Petitioner.

5 Patent Owner’s reliance on purportedly incorporated material is discussed
below.
6 We disagree with Patent Owner’s assertion that these usages are irrelevant
because they concern “detecting” not “mapping.” See infra § II.C.1.c.
IPR2020-00733
Patent 9,550,294 B2

14
Indeed, Petitioner and Patent Owner appear to agree that the plain
meaning of “objects” includes walls, although Patent Owner disputes that
the plain meaning applies here, largely due to the prosecution history. See,
e.g., Pet. 5–7; Reply 4–5; PO Resp. 19 (“[T]he prosecution history would
have led a [person of ordinary skill in the art] to conclude that the recited
‘objects’ are distinct from ‘walls,’ despite the use of plain meaning of
‘objects’ suggesting that walls alone would suffice to satisfy the ‘mapping’
limitation.”) (emphasis added); Sur-reply 3 (“In the context of detecting, the
term ‘object’ may refer more generally to ‘something material that may be
perceived by the senses,’ such as walls and furniture, which appear the same
to the robot’s sensor.”), 6–10 (disputing plain meaning’s application); see
infra § II.C.2.a. We will return to the plain meaning of “objects” after
considering the additional evidence and Patent Owner’s arguments.
c) Incorporation by Reference – ’278 Patent
Patent Owner relies heavily upon the disclosure of U.S. Patent
No. 7,024,278 B2 (Ex. 2001, the “’278 patent”). PO Resp. 20–28; Sur-
reply 13–15. Patent Owner contends the ’278 patent is “incorporated by
reference in its entirety in the ’294 patent” and supports its construction of
“objects.” PO Resp. 6 (citing Ex. 1001, 1:42–51), 6–8. The ’278 patent is
the primary evidence Patent Owner cites to support the alleged distinction
between objects that define a working area (e.g., walls) and objects within a
working area (e.g., furniture). See, e.g., id. at 3, 8, 10–11, 20–28.
Patent Owner’s reliance on the ’278 patent fails, however, for three
reasons.
IPR2020-00733
Patent 9,550,294 B2

15
First, the ’278 patent does not meaningfully use the term “objects,”7
and Patent Owner fails to demonstrate that its disclosures related to
“obstacles” are pertinent to the construction of the claim term “objects.” We
invited the parties to address “whether support for the term ‘objects’ can be
found in disclosures directed to ‘obstacles.’” Dec. on Inst. 14. In its
Response, Patent Owner relies on Dr. Coates’ testimony to assert that it can:
“Disclosures directed to ‘obstacles’ in the context of room mapping can
provide support for the term ‘objects.’” PO Resp. 3, 11 n.1 (same) (citing
Ex. 2029 ¶ 39). But this is not supported by any persuasive evidence or
reasoning.
For example, Dr. Coates asserts that the challenged ’294 patent
produces maps by distinguishing between obstacles that define a work area
from those within the work area, which he refers to as “objects,” “obstacles,”
or “obstructions,” “consistent with how these words were used in the art.”
Ex. 2029 ¶ 39 (citing Ex. 1001, 16:41–43; Ex. 2001, 9:39–43, 18:1–16;
Ex. 2004, Figs. 3, 5). Thus, Dr. Coates appears to assert that these terms can
be used interchangeably. The evidence that Dr. Coates cites, however, does
not support this testimony.
Indeed, the ’294 patent does not use the term “obstruction” at all.
And although the ’294 patent utilizes the term “obstacle,” neither Patent
Owner nor Dr. Coates show where the ’294 patent equates it with “objects,”
or explains why a skilled artisan would have considered them to be
equivalent. Ex. 2029 ¶ 39; PO Resp. 3, 11. Moreover, the ’294 patent
utilizes the term “obstacles” in reference to, inter alia, walls; it does not use

7 The ’278 patent uses “objects” only to describe three goals, or purposes, of
the invention. Ex. 2001, 3:14, 21, 26.
IPR2020-00733
Patent 9,550,294 B2

16
that term to distinguish between obstacles that define a work area from those
within the work area, as Dr. Coates asserts. For example, the ’294 patent
plainly states that the robot “maintain[s] its position with respect to an
obstacle, such as a wall or a piece of furniture.” Ex. 1001, 9:31–33.8
The remainder of Dr. Coates’ citations also fail to equate “objects,”
“obstacles,” and “obstructions,” and fail to show the terms are used to
distinguish between obstacles that define a work area from those within a
work area. See Ex. 1001, 16:41–43 (“walls and large objects”); Ex. 2001,
18:1–16 (identifying hot spots that trigger, e.g., Obstacle-Following mode,
which cleans the perimeter of walls and furniture); Ex. 2004, Figs. 3, 5
(depicting running patterns for a robot around obstruction(s)).
It is not our position that a person of ordinary skill in the art would
necessarily understand “objects,” “obstacles,” and “obstructions” to refer to
different things. Rather, we simply determine that Patent Owner and its
declarant fail to support the contention that disclosures related to the
unclaimed term “obstacles” are pertinent to the proper construction of the
claimed term “objects,” despite having been invited to do so. Dec. on
Inst. 14. Accordingly, we give Dr. Coates’ testimony little weight in this
regard, and decline to speculate as to the import of the ’278 patent’s
disclosures regarding “obstacles.” 37 C.F.R. § 42.65(a).
Second, even if we were to consider the disclosure of the ’278 patent,
it does not support Patent Owner’s distinction between objects that define a
working area (like walls) and objects within a working area (like furniture).
To be sure, we agree with Patent Owner that the ’278 patent describes a

8 Indeed, Dr. Coates loosely admits that “walls” are sometimes referred to as
“obstacles.” Ex. 2029 ¶ 39 n.2.
IPR2020-00733
Patent 9,550,294 B2

17
“working area” bounded by walls. Ex. 2001, 19:41–51, Fig. 6A, 7; PO
Resp. 21–22. We also recognize that the ’278 patent discloses storing a
“grid map” of the “working area.” Ex. 2001, 12:12–30; PO Resp. 21. And,
as Patent Owner notes, “one or more cells in the stored grid map are pre-
identified . . . as ‘hot spots’ in the defined working area” and, when the robot
encounters such a hot spot, a predetermined cleaning behavior, like Spot
Coverage mode or Obstacle-Following mode, is triggered. Ex. 2001, 17:66–
8:16; PO Resp. 23–24.
Importantly, however, the cited disclosure does not explain how these
hot spots are mapped, or what these hot spots indicate, e.g., whether the hot
spots indicate an object within a working area as opposed to an object
defining a working area. Ex. 2001, 17:66–8:16. Given the triggered
cleaning behaviors, it appears that the presence of a hot spot may indicate a
condition warranting Spot Coverage mode—such as a high traffic area—or
may indicate a condition warranting Obstacle-Following mode—such as an
obstacle or a wall—although no specifics are disclosed expressly. Id.; see
also id. at 9:1–13 (Spot Coverage), 9:39–50 (Obstacle-Following Coverage).
Indeed, the ’278 patent’s disclosure of Obstacle-Following mode casts
doubt on Patent Owner’s distinction between objects that define a working
area versus those within a working area. The ‘278 patent discloses that
Obstacle-Following mode applies to obstacles (furniture) within a working
area and also applies to the walls defining the working area itself: “The
Obstacle-Following Coverage pattern causes the robotic device 100 to clean
the perimeter of the defined working area, e.g., a room bounded by walls,
and/or the perimeter of an obstacle (e.g., furniture) within the defined
working area.” Ex. 2001, 9:39–43 (cited at PO Resp. 11, 24). The ’278
IPR2020-00733
Patent 9,550,294 B2

18
patent further explains that the robot “continuously maintain[s] its position
with respect to an obstacle, e.g., wall, furniture.” Id. at 9:43–45 (emphasis
added) (cited at PO Resp. 24; Sur-reply 14). Thus, the ’278 patent does not
support Patent Owner’s distinction between objects (or obstacles) within a
working area, as opposed to those defining a working area; rather, it appears
that Obstacle-Following mode applies to both types of obstacles.9
To the extent that Patent Owner suggests in its Sur-reply that this
disclosure of Obstacle-Following mode is not relevant because it concerns
detection, not mapping, we disagree. Sur-reply 1–3. As an initial matter,
Patent Owner relies upon the disclosure of Obstacle-Following mode to
support its position. PO Resp. 11 (citing Ex. 2001, 9:39–43), 24 (citing
Ex. 2001, 9:38–50); Sur-reply 14 (citing Ex. 2001, 9:44–45). Moreover, we
do not agree that disclosures related to detection are irrelevant, when Patent
Owner and Dr. Coates admit that detection is a prerequisite for mapping.
PO Resp. 5 (“This is achieved by active searching for objects within the
defined room as points of reference to build a map of the room while also
using the map to direct the robot about the room.”) (emphasis added);
Ex. 2029 ¶ 29 (same); see infra § II.C.1.d.
In sum, we find little support for Patent Owner’s argument that the
’278 patent distinguishes between obstacles within the room (furniture) and
obstacles defining the room (walls). Patent Owner reads too much into the
’278 patent’s limited disclosure of “hot spots.”

9 We recognize that this mode may be applied at times other than when the
robot encounters a mapped hot spot. However, as discussed above, Patent
Owner does not identify any portion of the ’278 patent that speaks directly to
what the mapped hot spots indicate, other than by reference to the disclosure
of Obstacle-Following mode.
IPR2020-00733
Patent 9,550,294 B2

19
Third, we note that the ’278 patent is not properly incorporated by
reference. In its Background section, the ’294 patent identifies “[a]dditional
robot control and navigation systems,” including that disclosed in U.S.
Patent Application No. “10/661,835” (the “’835 application”), “which [is]
hereby incorporated by reference.” Ex. 1001, 1:47–51. Patent Owner is
correct that the ’835 application issued as the ’278 patent. See Ex. 2001,
code (21); PO Resp. 6. However, the ’294 patent incorporates only the
’835 application, not the ’278 patent. Patent Owner did not offer the
identified ’835 application into the record as evidence.
Moreover, Patent Owner relies upon the incorporated disclosure to
support its understanding of the “mapping” limitation of claim 1. Therefore,
this appears to be “essential material” necessary to satisfy 35 U.S.C. § 112.
See 37 C.F.R. § 1.57(d)(1)–(3). But, essential material may only be
incorporated by reference “by way of . . . a U.S. patent or U.S. patent
application publication,” not by incorporation of an unpublished application
like the ’835 application. See id. § 1.57(d) (emphasis added); but cf. id.
§ 1.57(e) (“Non-essential material may be incorporated from . . . prior and
concurrently filed commonly owned U.S. applications.”). As such, the
incorporation of the ’835 application is not effective as regards essential
material. Id. § 1.57(h). Although raised by Petitioner, see Reply 7 n.4,
Patent Owner declined to meaningfully address this issue, instead making
the conclusory assertion that the ’278 patent “is properly incorporated,”
without explanation or support. Sur-reply 14.
Nonetheless, regardless of whether it is properly incorporated by
reference, the ’278 patent is prior art identified on the face of the ’294 patent
IPR2020-00733
Patent 9,550,294 B2

20
and, as such, is considered as part of the intrinsic record. See infra
§ II.C.2.b; Ex. 1001, code (56).
d) Incorporation by Reference – ’844 Patent
The ’294 patent properly incorporates U.S. Patent No. 6,594,844
(Ex. 1019, the “’844 patent”). Ex. 1001, 1:42–46; 37 C.F.R. § 1.57(d).
Petitioner contends the following disclosure is pertinent: “If the object is the
surface, the output from the circuit causes the robot to be directed to avoid
an obstacle. If, on the other hand, the object is a wall, the output from the
circuit causes the robot to be directed back towards the wall.” Pet. 6
(quoting Ex. 1019, 3:11–14 (emphasis added)). Patent Owner argues this
statement is out of context and directed to object detection and avoidance,
not mapping. PO Resp. 28–32 (“[B]eing detected by a robot does not inform
the meaning of the term ‘objects,’ within the context of the recited
‘mapping’ feature, which specifically requires that the ‘objects’ not merely
be detected but also be accounted for as ‘points of reference’ in a room
map.”). Patent Owner relies on Dr. Coates’ testimony in support. Id. at 28–
32 (citing Ex. 2029 ¶¶ 63–67).
We agree with Petitioner that this evidence is pertinent. The
’844 patent is properly incorporated and forms part of the intrinsic record.
Additionally, the ’844 patent utilizes the claim term “objects.” The
’844 patent is directed to an autonomous cleaning robot that detects both
obstacles and walls, and seeks to improve its detection sensor systems.
Ex. 1019, code (57), 1:33–44. In describing the sensor subsystem, the ’844
patent explains that, when a detected object is a wall, the sensor output
causes the robot to turn toward the wall. Id. at 3:3–15.
IPR2020-00733
Patent 9,550,294 B2

21
We disagree that evidence directed to object detection is “out-of-
context.” PO Resp. 28. As discussed above, the ’294 patent does not
explain how its map is created, other than to say the map uses objects as
points of reference. Ex. 1001, 16:41–43. But Patent Owner states that the
map is created by first detecting objects and building a map from there. PO
Resp. 5 (“This is achieved by active searching for objects within the defined
room as points of reference to build a map of the room while also using the
map to direct the robot about the room.”) (emphasis added). Dr. Coates
agrees. Ex. 2029 ¶ 29 (same). Neither Patent Owner nor Dr. Coates explain
persuasively why we should disregard evidence directed to detecting objects,
when detecting objects is an admitted part of the “mapping” process.
Ex. 2029 ¶¶ 63–67; PO Resp. 28–32. Further, we recognize that not all
detected objects are mapped. PO Resp. 31. But this does not discredit the
’844 patent’s description of “objects” as including walls, because detection
is a necessary part of “mapping,” for whichever objects are built into that
map. PO Resp. 5; Ex. 2029 ¶ 29.
Accordingly, the ’844 patent’s use of “object,” consistent with its
plain and ordinary meaning as something material perceived by the senses,
including walls, is persuasive evidence as to the meaning of “objects.”
2. The Prosecution History
“In addition to consulting the specification, . . . a court ‘should also
consider the patent’s prosecution history, if it is in evidence.’” Phillips v.
AWH Corp., 415 F.3d 1303, 1317 (Fed. Cir. 2005). “Like the specification,
the prosecution history provides evidence of how the PTO and the inventor
understood the patent. . . . Yet because the prosecution history represents an
ongoing negotiation between the PTO and the applicant, rather than the final
IPR2020-00733
Patent 9,550,294 B2

22
product of that negotiation, it often lacks the clarity of the specification and
thus is less useful for claim construction purposes.” Id. at 1317.
a) Examination
The ’294 patent issued from Application Number 14/946,999 (the
“’999 application”). Ex. 1001, code (21). During prosecution, the Examiner
rejected pending claims 31–43 for non-statutory double patenting over U.S.
Patent No. 9,215,957 (the “’957 patent”)—the parent of the ’999 application.
Id. at 315–316; Ex. 1001, code (60). In response, the applicant filed a
Terminal Disclaimer (Ex. 1002, 94–95) and “request[ed] withdrawal of the
foregoing rejection[] of claims 31–43” (id. at 91).
Appellant also amended pending claims 31–33 and 36–39, without
further comment. Id. Pending independent claim 31 was amended as
follows:
31. (Currently Amended) An autonomous cleaning
robot, comprising:
an undercarriage;
a motive system configured to propel the undercarriage
across a ground surface;
an energy storage unit supported by the undercarriage
and configured to be charged while the robot is positioned at a
base charging station; and
a navigational control system configured to
autonomously:
control the motive system to direct the robot about a
room, while mapping the room with respect to walls and
objects as points of reference;
return the robot to a the base charging station before
completing a cleaning task in the room, in response to
detecting a need to recharge the energy storage unit;
dock the robot to the base charging station upon
return;
recharge the energy storage unit with the robot
docked; and then to
IPR2020-00733
Patent 9,550,294 B2

23
control the motive system to direct the robot to continue
to clean the room.
Ex. 1002, 88. The applicant also made minor changes to pending claims 32,
33, and 36–39. Id. at 88–89. Thereafter, the ’999 application was allowed,
with pending application claim 31 issuing as claim 1 of the ’294 patent. Id.
at 20–27.
Patent Owner makes several arguments regarding the prosecution
history, which we address in turn.
Patent Owner argues that the amendment to pending claim 31
“indicates an express intent by the inventors to distinguish between [‘walls’
and ‘objects’] i.e., that ‘objects’ are distinct from ‘walls’ within the context
of room mapping.” PO Resp. 12–13, 15–16. Patent Owner contends this
amendment “had actual substantive import, and was not simply for
‘streamlining’ the claim.” Id. at 15–16 (citing Dec. on Inst. 13–14).
We do not discern any clear “intent” in the prosecution history. In
response to the Examiner’s double patenting rejection, the applicant filed a
Terminal Disclaimer and, accordingly, requested that the rejection be
withdrawn. Ex. 1002, 91. The filing of the Terminal Disclaimer obviated
the outstanding double patenting rejection; therefore, we decline to speculate
as to any meaning to be attributed to the accompanying claim amendment,
which was made without comment. Id.
Although the amendment was made without comment, Patent Owner
argues that the conjunction “and” in the pending claims before amendment
indicates that “mapping with respect to ‘objects’ is not satisfied by mere
mapping with respect to ‘walls.’” PO Resp. 18–19. Patent Owner argues
that interpreting “objects” to “NOT be distinct from ‘walls’ . . . reads out the
‘and objects’ part of the [pre-amendment] limitation” because “any disclosed
IPR2020-00733
Patent 9,550,294 B2

24
mapping ‘with respect to walls’ would also be a disclosed mapping ‘with
respect to walls and objects’ because, under this interpretation, walls are also
objects.” PO Resp. 13–14 (also arguing that different claim terms are
presumed to have different meaning and scope), 16. Patent Owner notes that
the parent ’957 patent claims recite “walls and objects,” and argues that the
’294 patent’s prosecution history must be read in view of the other patents in
its family, including the ’957 patent. Id. at 17–18; see also Sur-Reply 8 (“If
‘objects’ alone included ‘walls,’ as Petitioner argues, there would be no need
for the additional ‘and objects’ claim language in the ’957 patent.”).
As an initial matter, we do not interpret “objects” to be indistinct from
“walls,” as Patent Owner’s argument presumes. Rather, we understand the
plain meaning of “objects” to be a broad term that includes “walls.” Second,
we do not agree that “any disclosed mapping ‘with respect to walls’10 would
also be a disclosed mapping ‘with respect to walls and objects.’” This is
incorrect: the first mapping (“with respect to walls”) requires mapping with
respect to only one thing, i.e., walls. The second mapping (“with respect to
walls and objects”) requires mapping with respect to two things, i.e., walls
and objects. Thus, a broad understanding of “objects” does not “read out”
anything. The amended claims simply have a different scope than the pre-
amendment claims and those in the ’957 patent, i.e., the amended claims
require only mapping with respect to one thing (objects) as opposed to two
things (walls and objects).

10 Importantly, the claims do not recite “with respect to walls”; they recite
“with respect to objects.” We address this argument because it is presented
by Patent Owner, even though it misrepresents the claim language.
IPR2020-00733
Patent 9,550,294 B2

25
According to Patent Owner, “the prosecution history would have led a
[person of ordinary skill in the art] to conclude that the recited ‘objects’ are
distinct from ‘walls,’ despite the use of plain meaning of ‘objects’
suggesting that walls alone would suffice to satisfy the ‘mapping’
limitation.” PO Resp. 19. Despite this admitted plain meaning, Patent
Owner contends that a person of ordinary skill in the art would not have
interpreted “objects” according to its plain meaning because
it would render the “with respect[] to objects” portion of the
“mapping” limitation meaningless. That is, all “mapping” of a
room in the context of mobile robotic cleaning systems would be
“with respect to objects as points of reference” under Petitioner’s
proposed plain meaning because all room mapping by such
systems would involve the mobile robotic cleaner’s sensors
perceiving materials in the robot’s environment.
Id. at 19–20 (citing Ex. 2029 ¶ 50); see also Sur-reply 6.
Again, we disagree. Mapping with “respect to objects as points of
reference” does not become meaningless if “objects” is interpreted according
to its admitted and broad plain meaning as something material that is
perceived by the senses, including walls. As discussed above, this limitation
requires mapping with respect to one specified thing (objects), and also
requires that the objects are reflected “as points of reference” in the map.
This is not meaningless and, in fact, is consistent with the extent of
disclosure of “mapping” in the ’294 patent Specification. See Ex. 1001,
16:41–43 (“Alternatively, the robot may employ room-mapping programs,
using the base station 10 and/or walls and large objects as points of
reference.”). Certainly, the scope of the ’294 patent claim language is
broader than the scope of the pre-amendment claims and those in the ’957
patent, but that does not render the limitation meaningless.
IPR2020-00733
Patent 9,550,294 B2

26
For the foregoing reasons, we do not find the examination history of
the ’999 application to support Patent Owner’s proposed construction.
b) Prior Art Cited in the ’294 Patent
“The prosecution history, which we have designated as part of the
‘intrinsic evidence,’ consists of the complete record of the proceedings
before the PTO and includes the prior art cited during the examination of the
patent.” Phillips, 415 F.3d at 1317.
Petitioner identifies several references cited on the face of the
’294 patent that it argues “uniformly confirm[] that ‘objects’ include
‘walls.’” Reply 15 (citing Ex. 1005, 4:40–42, Ex. 1027, 16:46–47;
Ex. 1028, 1:48–49); Ex. 1001, code (56). Patent Owner does not dispute
that these references describe “objects” as including “walls,” but instead
argues that these references describe only object detection, not mapping.
Han (one of the cited references), like the ’294 patent, is directed to an
energy management system for a robotic cleaner that aims to “accurately
induce the robot to a power source supplier” when its battery level drops.
Ex. 1005, code (57); see also infra § II.F. As the parties note, Han detects
“a distance to a fixed object . . . more precisely, a distance to a wall surface.”
Ex. 1005, 4:40–42. Likewise, the cited reference to Allen also discloses a
cleaning robot that detects “most large, immobile objects in the operating
environment including walls, posts, furniture.” Ex. 1027, 16:46–47.
As discussed above, we disagree that evidence directed to detecting
objects is irrelevant. PO Resp. 5; Ex. 2029 ¶ 29 (same). Accordingly, Han’s
and Allen’s use of “object,” consistent with its plain and ordinary meaning
IPR2020-00733
Patent 9,550,294 B2

27
as something that may be perceived by the senses, including walls, is
persuasive evidence as to the meaning of “objects.”11
c) Prior Art Cited in the Parent ’957 Patent
Patent Owner identifies U.S. Patent No. 4,674,048 (Ex. 2004,
“Okamura”), which is cited on the face of the ’957 patent—the parent to the
challenged ’294 patent. PO Resp. 32; Ex. 2001, code (56). Patent Owner
alleges that Okamura’s robot “studies the boundary of the” area to be
cleaned, stores the boundary as a map, and then traverses the map area and
updates the map to reflect obstructions internal to the boundaries of the
mapped area. Id. at 32–33 (citing Ex. 2004, 4:33–5:18). Patent Owner
alleges that “Okumura expressly distinguishes between boundaries, i.e.,
‘walls,’ of the mapped work area and obstructions, i.e., ‘objects,’ that are
mapped as points of reference internal to the mapped work area.” Id. at 33.
Okamura does not meaningfully use the term “objects.” Ex. 2004,
1:58–2:53 (identifying “objects” for purposes of the invention). In the
portion cited by Patent Owner, Okamura explains that the robot proceeds
through an area “to see if any obstruction is present” and, if so, “it is stored
in the storage 10.” Id. at 5:2–9. As discussed above, however, Patent
Owner and Dr. Coates fail to support the contention that disclosures related
to “obstructions” are pertinent to the proper construction of “objects,”
despite having been invited to do so. Dec. on Inst. 14; Ex. 2029 ¶ 39; see
supra § II.C.1.c. Accordingly, we decline to speculate as to the import of
Okamura’s disclosures regarding “obstructions.”

11 We find the cited reference to Chiappetta less instructive, as the parties do
not address whether its list of “walls, obstacles, or other objects” is inclusive
or exclusive, i.e., whether the listed “walls” are included in, or excluded
from, “objects.” Compare Ex. 1028, 1:48–49, with Ex. 1001, 16:41–43.
IPR2020-00733
Patent 9,550,294 B2

28
3. Extrinsic Evidence
“[W]hile extrinsic evidence ‘can shed useful light on the relevant
art,’” it is generally considered to be “less significant than the intrinsic
record in determining ‘the legally operative meaning of claim language.’”
Phillips v. AWH Corp., 415 F.3d 1303, 1317–18 (Fed. Cir. 2005).
a) Dictionary Definitions
“Within the class of extrinsic evidence, the court has observed that
dictionaries and treatises can be useful in claim construction. . . . Such
evidence, we have held, may be considered if the court deems it helpful in
determining ‘the true meaning of language used in the patent claims.’” Id. at
1318.
Petitioner provides several dictionaries and robotics textbooks that it
contends confirm that the term “objects” is used in the ’294 patent claims in
accordance with its plain meaning. See, e.g., Ex. 1014, 4 (“something
material that may be perceived by the senses”); Ex. 1015, 4 (“[s]omething
perceptible by one or more of the senses”); Ex. 1016, 293 (describing
mapping permanent objects such as walls, desks, filing cabinets); Ex. 1017,
631 (describing mapping walls and furniture). Patent Owner and its
declarant accept this plain meaning of “objects” and, as discussed above,
present no persuasive reasoning to explain why this meaning does not apply
here. PO Resp. 19 (arguing an artisan would “conclude that the recited
‘objects’ are distinct from ‘walls,’ despite the use of plain meaning of
‘objects’ suggesting that walls alone would suffice to satisfy the ‘mapping’
limitation”); Sur-reply 3 (“In the context of detecting, the term ‘object’ may
refer more generally to ‘something material that may be perceived by the
senses,’ such as walls and furniture, which appear the same to the robot’s
IPR2020-00733
Patent 9,550,294 B2

29
sensor.”); Ex. 2029 ¶ 49; see supra § II.C.2.a (addressing arguments that the
prosecution history compels rejection of the plain meaning).
These dictionaries and textbooks support Petitioner’s contentions.
4. Conclusion
For the foregoing reasons, we maintain our preliminary determination
that we need not provide an express construction of “objects,” other than to
determine that walls are included within the scope of the term, consistent
with the plain and ordinary meaning of “objects.” Dec. on Inst. 13–14; see,
e.g., Ex. 1014, 4. This determination is supported by the ’294 patent
Specification, the ’844 patent incorporated therein, prior art cited in the ’294
patent, and several dictionaries and robotic textbooks. See, e.g., Ex. 1001,
7:8, 7:50, 9:31–33, 16:41–43; Ex. 1005, 4:40–42; Ex. 1014, 4; Ex. 1015, 4;
Ex. 1016, 293; Ex. 1017, 631; Ex. 1019, 3:3–15; Ex. 1027, 16:46–47. For
the reasons explained above, Patent Owner does not show persuasively that
this admitted plain meaning should not apply.
D. Anticipation or Obviousness over Kim, or
Obviousness over the Combined Teachings of Kim and Jeong
Petitioner contends that claims 1–5, 8, 9, and 11–13 of the ’294 patent
are unpatentable as (1) anticipated by Kim, (2) obvious over Kim, or
(3) obvious over the combined teachings of Kim and Jeong. Pet. 8–55.
Petitioner relies on Kim for all claim limitations. Id. With respect to the
“mapping” limitation of claim 1, Petitioner also provides alternative
contentions relying on Jeong. Id. at 22–34.
Patent Owner disputes Petitioner’s contentions regarding only the
“mapping” limitation of claim 1. PO Resp. 34–57. Specifically, Patent
Owner argues that Petitioner advances an improper claim construction (id. at
34–36), that Kim does not map objects within a defined working area (id. at
IPR2020-00733
Patent 9,550,294 B2

30
36–41), that it would not have been obvious for Kim’s system to map
obstacles within a defined working area (id. at 41–49), and that Petitioner’s
contentions based on the combination of Kim and Jeong fail (id. at 49–57).
1. Kim (Ex. 1003)
Kim is a U.S. patent that discloses a robot cleaner. Ex. 1003,
code (57). Kim’s robot 1 includes microprocessor 2, memory 3, driving
means 5, roller 6, vacuum motor 8, and dirt collecting chamber 16. Id. at
3:13–31. The robot also includes navigation sensor 23, magnet 32, magnet
sensing means 34, and charging receptacle 31 for guiding the robot to
automatic charging means 30 to recharge its battery. Id. at 3:31–50.
During its initial circulation prior to actually cleaning a room, the
robot “moves along and parallel to walls of the room” to “memorize[] a
structure and a size of the room in the microprocessor 2 by detection of the
navigation sensor 23.” Id. at 10:4–13. “[M]icroprocessor 2 compares the
[detected] structure with its cleaning programs previously stored in the
memory 3 of the microprocessor 2 to select that program which is most
similar to the structure” of the room. Id. at 9:55–64. “Therefore, the robot
cleaner 1 can carry out a cleaning operation according to the selected
program.” Id. at 10:13–14; see also id. at 9:55–11:65, Figs. 13–15. In this
way, the robot “select[s] the most efficient cleaning program” for the room
to be cleaned. Id. at 10:1–2.
Kim also discloses that the charge level of the battery is continually
monitored during cleaning. Id. at 14:3–6. If the battery voltage decreases
below a certain level, the robot is “moved to the position of the charging
means 30” and the battery is “charged with direct current.” Id. at 14:7–62.
“When the battery 26 is completely charged . . . the robot cleaner 1 moves to
IPR2020-00733
Patent 9,550,294 B2

31
the position at which the cleaning operation is interrupted . . . and then again
carries out the cleaning operation at that position.” Id. at 14:63–15:2.
2. Jeong (Ex. 1004)
Jeong is a U.S. patent that discloses a self-propelled robot capable of
cleaning and monitoring a work area. Ex. 1004, code (57), 1:8–11. During
an initial operation, Jeong’s robot “makes an environmental map for the
work area [A],” which includes a plurality of cells C with obstacles H (e.g.,
furniture) and walls W. Id. at 5:4–10, Fig. 3. To create the map, Jeong’s
robot utilizes ultrasonic sensors 411, 421, 431 to determine the distance
between the robot and various wall surfaces, to move the robot along the
wall surfaces, and to segment and assign cell numbers to portions of the
room. Id. at 5:37–7:1, 7:6–64. When the robot has returned to its initial
position, travel is stopped and the data associated with the cells is stored. Id.
at 6:65–7:5.
Jeong explains that use of the environmental map allows the robot to
properly perform tasks and to travel accurately. Id. at 8:27–29. Moreover,
Jeong explains that
the present invention enables the apparatus to create the
environmental map requiring relatively smaller memory capacity
without separate external devices for storage during the initial
tasks, recognize the current location of the activated robot based
upon the created environmental map and update the associated
position, and thus perform the given tasks of the robot while the
robot accurately travels to the target point.
Id. at 8:30–38.
IPR2020-00733
Patent 9,550,294 B2

32
3. Analysis of Claim 1 - Anticipation
Petitioner contends that Kim anticipates claim 1. Pet. 8–55. Patent
Owner disputes Petitioner’s contentions regarding only the “mapping”
limitation. PO Resp. 34–41.
a) “An autonomous cleaning robot”
Petitioner contends that Kim discloses the subject matter of the
preamble, including a “self-moving robot cleaner” that is autonomous.12
Pet. 14–15 (citing, e.g., Ex. 1003, code (57)).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses a “self moving robot cleaner.” Ex. 1003, code (57), 1:5–7
(“capable of cleaning a floor while moving on a floor by itself”).
b) “an undercarriage”
Petitioner contends that Kim discloses an undercarriage as claimed,
e.g., the underside of the robot cleaner and bottom plate 4, on which various
structures are mounted. Pet. 16–17 (citing, e.g., Ex. 1003, 3:17–18, 3:35–39,
4:4–40, 5:48–6:2, Fig. 1).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses an “underside of the cleaner” and bottom plate 4 on which
various structures are mounted. Ex. 1003, 3:17–18, 3:35–39, Fig. 1.
c) “a motive system configured to propel the undercarriage across a
ground surface”
Petitioner contends that Kim discloses a motive system as claimed,
e.g., driving means 5. Pet. 17–18 (citing, e.g., Ex. 1003, 3:13–21, 3:66–
4:52).

12 We need not resolve whether the preamble is limiting because Petitioner
shows sufficiently that the recitation in the preamble is satisfied by the prior
art, regardless of whether the preamble is limiting.
IPR2020-00733
Patent 9,550,294 B2

33
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses “driving means 5 mounted on an underside of the cleaner 1
and adapted to drive a driving roller 6 . . . to cause the robot cleaner 1 to be
moved forward and backward and to turn left and right.” Ex. 1003, 3:17–21.
d) “an energy storage unit supported by the undercarriage and configured
to be charged while the robot is positioned at a base charging station”
Petitioner contends that Kim discloses an energy storage unit as
claimed, e.g., battery 26, which is supported by the underside and bottom
plate of the robot cleaner. Pet. 18–21 (citing, e.g., Ex. 1003, 3:42–46, 3:57–
59, 8:18–21, 9:21–35, Figs. 1–2).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses battery 26, which is charged when the robot is positioned at
automatic charging means 30. See, e.g., Ex. 1003, 3:42–46, 3:57–59, 9:21–
35 (“When the robot cleaner 1 is precisely located with respect to the
automatic charging means 30 . . . the battery 26 [may] be completely
charged.”). As seen in Figures 1 and 2, the battery is supported by the
underside and bottom plate of the robot via intermediate structures. Id. at
Fig. 1; see also Ex. 1020 ¶¶ 82–83.
e) “a navigational control system configured to autonomously . . .”
Petitioner contends that Kim discloses a navigational control system
that operates autonomously, as claimed. Pet. 21–22 (citing, e.g., Ex. 1002,
1:7–8, 3:13–18, 11:33–39).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses the robot cleaner includes “microprocessor 2 for controlling
the overall operation of the cleaner” and “memory 3 for storing driving
programs for the cleaner . . . and driving means 5 . . . to cause the robot
cleaner 1 to be moved forward and backward and to turn left and right,”
IPR2020-00733
Patent 9,550,294 B2

34
wherein the robot is capable of moving by itself, e.g., autonomously. See,
e.g., Ex. 1003, 1:5–8, 3:13–21; see also Ex. 1020 ¶ 88.
f) “[the] navigational control system configured to autonomously: control
the motive system to direct the robot about a room, while mapping the
room with respect to objects as points of reference”
Petitioner contends that Kim’s control system directs the self-moving
robot about a room while mapping the room with respect to (at least) walls
as points of reference, as claimed, during its “initial circulation.” Pet. 22–27
(citing, e.g., Ex. 1003, 2:62–66, 3:13–21, 9:55–10:14, 10:18–11:65).
Patent Owner argues only that Petitioner relies upon “an improper
interpretation of ‘objects’ as being fully satisfied by obstacles that define the
room (i.e., the ‘walls’)” and, accordingly, “Kim’s memorization of room
structure and size using only walls . . . does not meet this requirement” of
claim 1. PO Resp. 34–35; see also id. at 34–36. As discussed above,
however, we determine that walls fall within the scope of “objects.” See
supra § II.C.
The evidence of record supports Petitioner’s contention. Kim’s robot
is directed about a room during an “initial circulation” in which the robot
cleaner “does not clean a floor but memorizes only the structure of the
floor.” See, e.g., Ex. 1003, 9:55–59. Specifically, the robot “moves along
and parallel to walls of the room” to “memoriz[e] a structure and a size of
the room.” Id. at 10:4–8, 9:65–68. Kim’s robot memorizes the size and
structure of the room so that it can be compared against cleaning programs
previously stored in the robot’s memory, and the program most closely
matching the room’s structure may be selected as the cleaning operation
program. Id. at 9:59–64.
IPR2020-00733
Patent 9,550,294 B2

35
Figure 13, reproduced below, shows the initial circulation of the robot
cleaner.

Figure 13 depicts robot cleaner 1 performing an initial circulation in which
the robot travels around the periphery of the room, parallel to the room’s
walls. Id. at 9:55–56, 10:4–8.
Thus, we are persuaded by Petitioner’s contention that Kim’s initial
circulation satisfies this limitation of claim 1 because Kim’s robot is directed
about the room while mapping the room with respect to the walls of the
room as points of reference. Id. at 9:55–59, 10:4–8, Fig. 13; see also
Ex. 1020 ¶¶ 90–97; § II.D.3.e.
IPR2020-00733
Patent 9,550,294 B2

36
g) “[the] navigational control system configured to autonomously . . .
return the robot to the base charging station before completing a
cleaning task in the room, in response to detecting a need to recharge
the energy storage unit”
Petitioner contends that Kim’s self-moving robot returns to charging
means 30, as claimed. Pet. 35–36 (citing, e.g., Ex. 1002, 1:6–13, 8:44–62,
14:3–33).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses that the self-moving robot cleaner “mov[es] to an automatic
charging means to charge its battery when a charging voltage of the battery
is decreased to below a predetermined level during a cleaning operation and
then returning to the interrupted cleaning position to again carry out the
cleaning operation.” See, e.g., Ex. 1003, 1:6–13; see also id. at 8:44–46;
§ II.D.3.e.
h) “[the] navigational control system configured to autonomously . . . dock
the robot to the base charging station upon return”
Petitioner contends that Kim’s self-moving robot cleaner docks at
charging means 30, as claimed. Pet. 37–40 (citing, e.g., Ex. 1003, 2:58–59,
7:64–9:35, 8:31–41, 14:47–62, Figs. 10, 11a–b, 12).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses that charging means 30 emits an ultrasonic wave that the
robot uses to locate the charging means, detects the position of the robot
by detecting a magnetic force of the robot, and “cause[s] the robot cleaner
to be precisely moved to the charging position” for charging, by
generation of a line of magnetic force. See, e.g., Ex. 1003, 8:6–25,
Fig. 12; § II.D.3.e.
IPR2020-00733
Patent 9,550,294 B2

37
i) “[the] navigational control system configured to autonomously . . .
recharge the energy storage unit with the robot docked”
Petitioner contends that Kim’s self-moving robot cleaner is recharged
by charging means 30, as claimed, when docked. Pet. 40–41 (citing, e.g.,
Ex. 1003, 9:21–35, 14:37–62, Fig. 16).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses that “[w]hen the robot cleaner 1 is precisely located with
respect to the automatic charging means 30, . . . the direct current applying
plug 156 of the charging means 30 is positively connected to conductor 31a
of the charging receptacle 31, thereby permitting the battery 26 to be
completely charged.” See, e.g., Ex. 1003, 9:21–35, Fig. 16 (S11); § II.D.3.e.
j) “[the] navigational control system configured to autonomously . . . then
to control the motive system to direct the robot to continue to clean the
room”
Petitioner contends that Kim’s self-moving robot cleaner continues to
clean the room after being recharged. Pet. 40–41 (citing, e.g., Ex. 1003,
1:12–14, 3:13–21, 9:42–54, 14:63–15:2, Fig. 16).
The evidence of record supports Petitioner’s undisputed contention.
Kim discloses that “after the battery is completely charged, the robot cleaner
1 can return to the discontinued cleaning location of the floor and then again
carry out its cleaning operation.” See, e.g., Ex. 1003, 9:42–54; § II.D.3.e.
k) Summary
Accordingly, we determine that Petitioner has met its burden of
demonstrating by a preponderance of the evidence that claim 1 is
unpatentable as anticipated by Kim.
IPR2020-00733
Patent 9,550,294 B2

38
4. Analysis of Claim 1 – Obviousness over Kim
To account for Patent Owner’s construction of “objects,” see supra
§ II.C, Petitioner alternatively contends it would have been obvious to
modify Kim to map the room with respect to non-wall obstacles as points of
reference.13 Id. at 28–30. Petitioner relies upon Kim for all remaining
limitations, and we incorporate our findings set forth above. Pet. 14–42; see
supra § II.D.3. Patent Owner disputes the “mapping” contentions only. PO
Resp. 41–49.
Petitioner contends that, during Kim’s room mapping process,
navigation sensor 23 “detect[s] a presence of an environmental obstacle,” as
well as its distance and direction. Pet. 28 (quoting Ex. 1003, 3:32–35, 7:59–
60; citing Ex. 1003, 10:18–11:16, Fig. 15). If an obstacle is detected,
Petitioner contends the robot turns to avoid it. Id. at 29 (citing Ex. 1003,
11:1–7). Additionally, Kim’s microprocessor “inputs the data of the moved
distance [and] the positions and turning directions at which the robot
cleaner 1 has been turned.” Id. (quoting Ex. 1003, 11:57–65).
According to Petitioner, it would have been obvious to map this
obstacle-avoiding turn information because Kim already maps turns that are
made to avoid walls during the initial circulation. Id. at 30 (citing Ex. 1020
¶ 99). Petitioner contends this would have been the simple substitution of
one known element—“data related to the position of non-wall obstacles”—
for another—“data related to the position of walls”—“to achieve the
predictable and desired result of increasing the granularity with which the

13 Although we reject Patent Owner’s claim construction, see supra § II.C,
we apply Patent Owner’s construction here for completeness in evaluating
Petitioner’s asserted grounds.
IPR2020-00733
Patent 9,550,294 B2

39
room was mapped by including not only the outer boundaries of the room
but also the internal obstacles within those boundaries,” which “would
improve overall robot performance and coverage of the area.” Id.
Patent Owner disputes these contentions. Patent Owner contends that
Kim does not disclose an algorithm to allow the robot to traverse the interior
of the room when mapping, but simply follows along and parallel to the
walls of the room. PO Resp. 42–43. Patent Owner describes two potential
scenarios that Kim’s wall-traversing robot might encounter, which it depicts
in two modified versions of Kim’s Figure 13, reproduced below. Id. at 45–
46. We added identifiers A, B, and C to the second (right) modified figure.

Patent Owner’s modified versions of Kim’s Figure 13 depict added
“objects,” shown in blue. In the above left figure, one object is added to the
left corner of the room (the “First Scenario”). In the above right figure,
three objects are added, with Object A near the left wall and with Objects B
and C in the interior of the room (the “Second Scenario”).
In the First Scenario, Patent Owner admits that if Kim’s robot detects
the object when moving parallel to a wall, it turns to avoid it. PO Resp. 43
IPR2020-00733
Patent 9,550,294 B2

40
(citing Ex. 1003, 11:1–7). Patent Owner does not dispute Petitioner’s
contention, or Dr. Kelly’s testimony, that a person of ordinary skill in the art
would have found it obvious to save turn information for this obstacle-
avoiding turn into the map; indeed, Patent Owner admits as much.14 Id. at
44 (explaining that Kim’s robot “maps the turn performed in step S12 in
response to detecting the obstacle as a boundary of the room (i.e., as a
wall)”). But, according to Patent Owner, Kim’s robot still only maps the
room with respect to “walls” because these wall-adjacent objects are seen by
the robot as a “wall” that effectively defines the boundary of the working
area. Id. at 43–44.
In the Second Scenario, Patent Owner contends that Object A is
placed away from the wall but is still detectable by Kim’s robot as it moves
along the wall. Id. at 45. Patent Owner argues that the robot recognizes that
the wall is on its left and, therefore, that Object A is not a wall, so will
maneuver to pass between the wall and Object A by making a slight left turn
toward the wall. Id. In this circumstance, Patent Owner argues that Kim
does not save this slight object-avoiding turn. Id. at 45–46. Patent Owner
contends there is no reason to save this information because this would
capture only the corner of Object A and would not capture the entire
perimeter, which “would have no value in terms of improving cleaning . . .
[or] selecting a cleaning pattern.” Id. Additionally, Patent Owner argues
that because Kim’s robot only moves along and parallel to the walls, the

14 Patent Owner describes such objects as “against or close enough to the
actual wall” to be seen as a wall by the robot. PO Resp. 44. For simplicity,
we refer to them as “wall-adjacent.”
IPR2020-00733
Patent 9,550,294 B2

41
robot does not detect objects that are fully interior to the room, like
Objects B and C in the Second Scenario depicted above. Id. at 47.
We are persuaded by Petitioner’s contentions regarding Kim’s
teachings and what they would have suggested to a person of ordinary skill
in the art. Specifically, Kim explains that during its “initial operation,” the
robot determines the distance to and direction of an environmental obstacle,
and turns to avoid it. Ex. 1003, 10:18–20, 10:34–38, 11:1–7. Additionally,
the robot “inputs data of . . . the turned angle.” Id. at 11:30–31. We credit
Dr. Kelly’s opinion that, in view of this disclosure, a skilled artisan would
have found it obvious to save into the map this information regarding turns
based on obstacles, in the same manner as turns based on walls, to increase
the granularity of the map and improve robot coverage of the area. Ex. 1020
¶¶ 98–99.
We disagree with Patent Owner’s arguments against these
contentions. As an initial matter, the claims do not require that the robot
traverse the interior of the room. The claims simply require that the robot
map the room “with respect to objects as points of reference,” without
exclusion of objects that are placed adjacent to a wall.
Patent Owner’s claim construction position is that this limitation
requires mapping more than “walls,” i.e., more than obstacles that define the
boundaries of a mapped work area but instead requires mapping “objects
within the perceived boundaries of the room as points of reference.” PO
Resp. 10, 11–12. Despite asserting this construction, Patent Owner’s
briefing does not explain how a robot perceives the difference between an
object that defines the boundary of the room versus one within the boundary
of the room. Based on the First and Second Scenarios depicted above,
IPR2020-00733
Patent 9,550,294 B2

42
Patent Owner’s implicit argument appears to be that the distinction turns on
whether the robot can traverse the space between a wall and the potential
object; if it cannot (First Scenario), it is perceived as a wall and, if it can
(Second Scenario), it is perceived as an object within the room. But Patent
Owner identifies no support for such a distinction, where the driving factor
in distinguishing between items perceived as “walls” and “objects” is the
size of the robot itself. Such a distinction is not warranted.15
As such, we understand a wall-adjacent object located within a room,
such as that depicted in Patent Owner’s First Scenario above, to be an
“object” within the boundary of the room, in accordance with Patent
Owner’s construction. Here, Patent Owner admits that, in the First Scenario,
Kim’s robot turns to avoid wall-adjacent objects within a room, and would
have included them in the map. PO Resp. 44 (“In one scenario, the obstacle
is against or close enough to the actual wall along which the robot is moving
such that the robot sees the obstacle as a ‘wall’ in step S12 and maps the
turn performed in step S12 in response to detecting the obstacle as a
boundary of the room (i.e., as a wall).”) (emphasis added). This satisfies the
“mapping” limitation of claim 1.

15 During the hearing, we asked Patent Owner’s counsel about an
intermediate scenario, in which an item is located between the position of
the object in the First Scenario and the position of Object A in the Second
Scenario, i.e., where the item is closer to the wall than Object A such that the
robot cannot fit between it and the wall, but is nonetheless not adjacent the
wall. Tr. 33:8–23. Patent Owner’s counsel stated this would be understood
as an “object” and not a “wall.” Id. Given this statement, we can discern no
standard for evaluating whether a given item is perceived to be a wall or an
object.
IPR2020-00733
Patent 9,550,294 B2

43
Despite admitting that it would have been obvious for Kim’s robot to
map wall-adjacent objects located within a room, Patent Owner also argues
that it makes little sense to map “a few obstacles that the robot just so
happened to have encountered as it was moving parallel to and along a wall”
because it would “require development of cleaning programs for largely
incomplete room maps.” Id. at 48. Moreover, if Kim were modified to map
objects within the room, Patent Owner argues that Kim would have to be
modified “to store a preposterously large number of cleaning programs”
against which to match the map “to account for every possible room
structure and configuration of internal obstacles in a room.” Id.
We disagree. Kim discloses matching the size and structure of the
room to pre-stored cleaning programs to select the stored program that most
closely matches the room map generated during the initial circulation.
Ex. 1003, 9:62. Thus, Kim already stores cleaning programs “for largely
incomplete room maps” that are based solely on the size and structure of the
room, as dictated by its walls. Also including wall-adjacent objects within
the map, as proposed by Petitioner, only further “completes” the map.
Additionally, we disagree that a “preposterously large number of
cleaning programs” must be stored. Patent Owner misinterprets Kim’s
disclosure. Kim seeks to select the pre-stored cleaning program that is
“most similar” to the structure of the room, not to store all the programs
necessary to ensure a full and complete match between every mapped room
configuration and its exact pre-stored cleaning program equivalent. Id. at
9:62; see also id. at 10:2 (“most efficient”), 10:12 (“most similar”), 12:7
(“most effective”). Moreover, Patent Owner does not allege (or show, for
that matter) that it would have been beyond the capability of an ordinarily
IPR2020-00733
Patent 9,550,294 B2

44
skilled artisan to implement Kim’s disclosed cleaning technique consistent
with a modified map, i.e., by storing, selecting, and implementing cleaning
patterns that include both walls and obstacles, even if such were required.
See KSR, 550 U.S. at 421 (“A person of ordinary skill is also a person of
ordinary creativity, not an automaton.”).
Accordingly, for the foregoing reasons, we determine that Petitioner
has met its burden of demonstrating by a preponderance of the evidence that
claim 1 is unpatentable as obvious over Kim.
5. Analysis of Claim 1 – Obviousness over Kim and Jeong
To account for Patent Owner’s construction of “objects,” see supra
§ II.C, Petitioner alternatively contends it would have been obvious, in light
of the teachings of Jeong, to modify Kim to map the room with respect to
non-wall obstacles as points of reference.16 Id. at 31–34. Petitioner relies
upon Kim for all remaining limitations, and we incorporate our findings set
forth above. Pet. 14–42; see supra § II.D.3–4. Patent Owner disputes the
“mapping” contentions only. PO Resp. 49–57.
Petitioner contends that Jeong discloses an initial operation in which
the robot makes an environmental map of the work area, which includes
both walls (W) and obstacles (H). Pet. 31 (citing, e.g., Ex. 1004, 5:4–6, 5:8–
12). Petitioner contends that Jeong’s map reduces memory requirements by
storing data as cells. Id. at 32 (citing Ex. 1004, 8:30–38). Therefore,
according to Petitioner, it would have been obvious to combine Jeong’s
mapping with Kim, for two reasons. Id. First, this would have been the

16 Although we reject Patent Owner’s claim construction, see supra § II.C,
we apply Patent Owner’s construction here for completeness in evaluating
Petitioner’s asserted grounds.
IPR2020-00733
Patent 9,550,294 B2

45
simple substitution of one known element (Jeong’s mapping with respect to
walls and obstacles) for another (Kim’s mapping with respect to walls) to
achieve the predictable and desired result of “increasing the granularity with
which the room is mapped,” which “would improve robot performance and
coverage because the map would allow the robot to navigate with respect to
both the walls and the internal non-wall obstacles.” Id. at 33 (citing
Ex. 1020 ¶ 103). Second, incorporating Jeong’s mapping technique based
on stored cells would have required smaller memory and been cheaper to
produce. Id. at 33–34 (citing, e.g. Ex. 1020 ¶ 104).
Patent Owner disputes these contentions. First, Patent Owner
contends that Petitioner fails to explain how a person of ordinary skill in the
art would have combined Kim’s cleaning patterns with Jeong’s map. PO
Resp. 50. For example, Patent Owner alleges that Petitioner does not
provide details about how the modification would be implemented, or how
Kim’s disclosed cleaning technique would be modified to use Jeong’s
environmental map, e.g., “how a POSITA would marry Jeong’s mapping
with Kim’s cleaning techniques given that billions of Kim’s cleaning
programs would be needed to account for every possible room structure and
configuration of internal obstacles in a room.” Id. at 51–52, 53 (comparing
Jeong’s flexible navigation techniques to Kim’s library of static cleaning
programs). Petitioner responds to this argument by arguing that the claims
do not require use of the map, therefore, Patent Owner’s arguments are
irrelevant. Reply 21. Additionally, Petitioner contends that “the Petition
explained that the Kim/Jeong combination uses the Jeong map and
navigation techniques.” Id. at 21–22 (citing Pet. 14, 33–34).
IPR2020-00733
Patent 9,550,294 B2

46
Second, Patent Owner argues that Petitioner fails to explain how
Jeong’s sensor configuration would be implemented in Kim’s robot, which
employs different sensors. Id. at 53. For example, Patent Owner compares
Jeong’s front, left, and right ultrasonic sensors that calculate distance and
angles to calculate X-Span and Y-Span data to identify cell boundaries for
three directions at the same time, with Kim’s single 360-degree sensor that
provides distance data in one direction at a time. Id. at 54–56. Patent
Owner alleges that Petitioner fails to address how Jeong’s sensor
configuration would be migrated to Kim’s robot, or the modifications
necessary to make the proposed combination. Id. at 56.
We are persuaded by Petitioner’s contentions. Petitioner
demonstrates that Jeong maps a room using walls and obstacles, “such as a
piece of furniture,” as points of reference in its map. Pet. 31; Ex. 1004, 5:4–
12. Jeong explains that its environmental map stores data as cells, which
reduces memory requirements. Ex. 1004, 8:30–38, Figs. 3, 8. Petitioner’s
contention that it would have been obvious to combine Jeong’s mapping
with Kim is supported. First, Jeong itself supports the contention that such a
combination would have reduced memory requirements in storing map data.
Ex. 1004, 8:30–38; Ex. 1020 ¶¶ 102, 104, 105 (also supported by Dr. Kelly’s
testimony). Second, we credit Dr. Kelly’s testimony that a person of
ordinary skill in the art would have found the proposed combination to
increase the granularity of the map. Ex. 1020 ¶¶ 103–104. Although Patent
Owner and Dr. Coates make several arguments, addressed below, neither
disputes that the proposed combination would have reduced memory usage
and increased map granularity, as proposed by Petitioner and Dr. Kelly.
IPR2020-00733
Patent 9,550,294 B2

47
We do not agree with Patent Owner’s first argument. To be clear,
however, we also do not agree with Petitioner’s assertion that its proposed
combination “uses the Jeong map and navigation techniques.” Reply 21–22
(citing Pet. 14, 33, 34). The Petition plainly relies upon Jeong for its
mapping technique only, and not for use of that map in Jeong’s navigation or
cleaning process. See, e.g., Pet. 32 (“It would have been obvious to combine
[Jeong’s] obstacle mapping with [Kim] for at least two reasons.”), 33
(“Doing so would have been the simple substitution of one known element
([Jeong’s] mapping with respect to both walls and non-wall obstacles) for
another ([Kim’s] mapping with respect to walls, according to iRobot).”), 33
(explaining that it would have been “obvious to include [Jeong’s] mapping
and map storage techniques into [Kim]”), 33–34 (“A skilled artisan would
have found it obvious to modify [Kim] to include [Jeong’s] grid-based
mapping that includes mapping with respect to walls and obstacles as
reference points.”). That Petitioner recognizes Jeong’s use of its own map in
its own processes does not amount to a contention that it would have been
obvious to incorporate Jeong’s cleaning process into Kim. Contra
Reply 21–22; Pet. 14 (quoting Ex. 1004, 1:10–11).
Nonetheless, we do not agree that Petitioner fails to explain how
Kim’s disclosed navigation and cleaning technique would be modified to use
Jeong’s environmental map. Importantly, Petitioner is correct that the
claims do not require using the map. Reply 21. Claim 1 simply requires
“mapping the room with respect to objects as points of reference,” but does
not specify any further action taken in association with the map.
As noted above, Petitioner presents the supported and unchallenged
assertion that modifying Kim to map both wall and non-wall objects, as
IPR2020-00733
Patent 9,550,294 B2

48
taught by Jeong, would realize memory savings by storing map data that
depicts walls and obstacles as cells. Pet. 32 (citing Ex. 1004, 8:30–38). This
holds true even if the map is never used by Kim’s robot, as Patent Owner
contends. Ex. 2029 ¶¶ 101–103. We are persuaded that a person of ordinary
skill in the art would have found it obvious to modify Kim to create a cell-
based map of the room, including its walls and obstacles, as taught by Jeong,
to reduce the memory storage associated with the map. Ex. 1020 ¶¶ 102,
104, 105.
Petitioner presents an additional motivation for the combination: to
increase granularity of the map by including both walls and obstacles,
which, it contends, would “improve overall robot performance and coverage
of the area.” Pet. 33; Ex. 1020 ¶ 103. Although the claims themselves do
not require “use” of the map, this alternative reasoning proposed by
Petitioner does at least imply such use.
We do not agree, however, that using Jeong’s map with Kim’s
cleaning technique is insufficiently explained in the Petition. PO Resp. 50.
As discussed above in Section II.D.4, Patent Owner misinterprets Kim’s
disclosure. Kim seeks to select the pre-stored cleaning program that is
“most similar” to the structure of the room, not to store all the programs
necessary to ensure a full and complete match between every mapped room
configuration and its exact pre-stored cleaning program equivalent. See,
e.g., id. at 9:62. Therefore, we do not agree that “billions” of cleaning
programs must be stored, against which the detailed map suggested by Jeong
would be compared. PO Resp. 52. Moreover, we discern that a person of
ordinary skill in the art would have been capable of weighing such tradeoffs
(storage of more cleaning programs) against the identified benefits (more
IPR2020-00733
Patent 9,550,294 B2

49
granular mapping). See Winner Int’l Royalty Corp. v. Wang, 202 F.3d 1340,
1349 n.8 (Fed. Cir. 2000) (“The fact that the motivating benefit comes at the
expense of another benefit, however, should not nullify its use as a basis to
modify the disclosure of one reference with the teachings of another.
Instead, the benefits, both lost and gained, should be weighed against one
another.”); see also In re Urbanski, 809 F.3d 1237, 1244 (Fed. Cir. 2016).
We also do not agree with Patent Owner’s second argument, which
primarily focuses on whether components of Kim’s and Jeong’s structures
can be physically combined together.17 “The test for obviousness is not
whether the features of a secondary reference may be bodily incorporated
into the structure of the primary reference . . . Rather, the test is what the
combined teachings of those references would have suggested to those of
ordinary skill in the art.” In re Keller, 642 F.2d 413, 425 (CCPA 1981); see
also In re Sneed, 710 F.2d 1544, 1550 (Fed. Cir. 1983) (“[I]t is not necessary
that the inventions of the references be physically combinable to render
obvious the invention under review.”).
Here, our inquiry concerns what would have been suggested to a
skilled artisan from Jeong’s teaching of including obstacles in the map
generated by a robot, not whether Jeong’s sensors could be included in
Kim’s robot, or vice versa. The record before us demonstrates that both Kim

17 At the end of the section addressing the alleged incompatibility between
Kim’s and Jeong’s sensors, Patent Owner alleges that “Petitioner has failed
to recognize that the initial circulation and cleaning algorithm of Kim is
fundamentally different than Jeong’s techniques, irrespective of their
physical combinability.” PO Resp. 56. This argument has already been
addressed in considering Patent Owner’s first argument. Despite this
statement, Patent Owner’s second argument really is drawn to “physical
combinability.” Id.
IPR2020-00733
Patent 9,550,294 B2

50
and Jeong use sensors to generate environmental maps of the room to be
cleaned, wherein Kim’s map includes walls and Jeong’s map includes both
walls and obstacles. See, e.g., Ex. 1003, 10:4–11; Ex. 1004, 5:4–12. We
credit Dr. Kelly’s testimony that it would have been a simple substitution of
Jeong’s mapping for Kim’s mapping. Ex. 1020 ¶ 103. Petitioner has
identified persuasive reasons explaining why an ordinarily skilled artisan
would have modified Kim to include obstacles in the map in the grid-based
manner taught by Jeong, e.g., to create a more detailed map and to reduce
memory space, as proposed. Ex. 1020 ¶¶ 103–105.
We do not agree that Kim’s sensor arrangement necessarily would
require modification in order to detect and store obstacles in the map, in the
same way walls are already detected and stored. Patent Owner does not
make a persuasive showing in this regard.18 See, e.g., PO Resp. 53–57;
Ex. 2029 ¶¶ 104–106. But, even if modification of Kim’s sensors were
required, we do not discern that this would have been beyond the skill level
of a person of ordinary skill in the art. See, e.g., ClassCo, Inc. v. Apple, Inc.,
838 F.3d 1214, 1219 (Fed. Cir. 2016) (“KSR does not require that a
combination only unite old elements without changing their respective
functions. Instead, KSR teaches that ‘[a] person of ordinary skill is also a
person of ordinary creativity, not an automaton.’ And it explains that the
ordinary artisan recognizes ‘that familiar items may have obvious uses
beyond their primary purposes, and in many cases a person of ordinary skill
will be able to fit the teachings of multiple patents together like pieces of a

18 Indeed, although Patent Owner details ways in which the sensors differ, as
discussed above, these differences do not preclude Kim from detecting the
presence of obstacles. See § II.D.4; Ex. 1003, 10:34–38 (determining “a
distance to an obstacle . . . [and] a direction of the obstacle”).
IPR2020-00733
Patent 9,550,294 B2

51
puzzle.’ The rationale of KSR does not support [Patent Owner’s] theory that
a person of ordinary skill can only perform combinations of a puzzle
element A with a perfectly fitting puzzle element B. To the contrary, KSR
instructs that the obviousness inquiry requires a flexible approach.”).
Accordingly, for the foregoing reasons, we determine that Petitioner
has met its burden of demonstrating by a preponderance of the evidence that
claim 1 is unpatentable as obvious over the combined teachings of Kim and
Jeong.
6. Analysis of Dependent Claims 2–5, 8, 9, and 11–13
Petitioner contends that dependent claims 2–5, 8, 9, and 11–13 are
anticipated and/or obvious over Kim, or over the combined teachings of Kim
and Jeong. Pet. 43–55. Petitioner relies upon Kim’s teachings to satisfy the
limitations of these dependent claims. Id. Patent Owner does not dispute
these contentions. PO Resp. 34–57. The evidence of record supports
Petitioner’s undisputed contentions.
Claim 2 depends from claim 1 and further recites “the control system
is configured to track a cleaning route of the robot and monitor an amount of
the room remaining to be cleaned.” Kim discloses that the cleaning robot
operates according to its selected program, recharges its battery during
operation, and returns to the point at which cleaning was discontinued in
order to resume cleaning from that position. See, e.g., Ex. 1003, 12:6–15,
9:51–55, Fig. 14; Pet. 43–46. We credit Dr. Kelly’s unrebutted testimony
that a person of ordinary skill in the art would have understood that the robot
monitors the amount of the room remaining to be cleaned based on, inter
alia, these teachings. See, e.g., Ex. 1020 ¶¶ 124–125; Pet. 43–46.
IPR2020-00733
Patent 9,550,294 B2

52
Claim 3 depends from claim 1 and further recites “the control system
is configured to use a known position of the base charging station as a point
of reference in mapping the room.” Kim discloses initializing the initial
circulation at a particular location and finishing the initial circulation when
the robot returns to “the initial start location.” See, e.g., Ex. 1003, 10:21–24,
11:57–65; Pet. 46–47. We credit Dr. Kelly’s unrebutted testimony that a
person of ordinary skill in the art would have understood this location to be
that of the charging means 30, because that is the location to which robot
returns when cleaning is complete and, therefore that this position is used as
a point of reference when mapping. Ex. 1020 ¶ 128.
Claim 4 depends from claim 1 and further recites “the control system
is configured to determine, through use of route-tracking subroutines while
controlling the motive system to direct the robot about the room, which
portion of the room still requires cleaning.” Kim discloses that “the
microprocessor 2 outputs data regarding that portion of the floor which has
been cleaned, that is, the data of a distance and a direction of cleaned floor,”
and that the robot returns to a discontinued cleaning location after the battery
is recharged. See, e.g., Ex. 1003, 9:43–46, 9:51–55; Pet. 47–49. We credit
Dr. Kelly’s unrebutted testimony that a person of ordinary skill in the art
would have found it obvious that these functions are implemented with
routine-tracking subroutines. And that the device would determine the
portion of the floor that continues to require cleaning. Ex. 1020 ¶¶ 131–132.
Claim 5 depends from claim 1 and further recites “the control system
is configured to direct the robot to the portion of the room still requiring
cleaning, following recharging the energy storage unit.” Kim discloses that
the robot cleaner “mov[es] to an automatic charging means to charge its
IPR2020-00733
Patent 9,550,294 B2

53
battery . . . and then return[s] to the interrupted cleaning position to again
carry out the cleaning operation.” See, e.g., Ex. 1003, 1:8–14; Pet. 49–50.
Claim 8 depends from claim 1 and further recites “the control system
is configured to return the robot to the base charging station before
completing the cleaning task in the room, in response to a quantity of energy
remaining in the energy storage unit falling to a predetermined threshold.”
Kim discloses that the robot cleaner “mov[es] to an automatic charging
device by itself and repeatedly charg[es] its battery when power of the
battery is discharged to below a predetermined level during a cleaning
operation” and then resumes the cleaning operation from its position prior to
recharging. See, e.g., Ex. 1003, 1:58–62, 14:3–62; Pet. 50–51.
Claim 9 depends from claim 8 and further recites “the navigational
control system is configured to direct the robot to the base charging station
by seeking a base station homing signal.” Kim discloses that the robot
cleaner seeks an “ultrasonic wave generat[ed] from the ultrasonic wave
oscillator 150 of the automatic charging means 30.” See, e.g., Ex. 1003,
14:6–23; Pet. 52–53; see also Ex. 1001, 7:37–38 (“ultrasonic sensors and
transducers may be employed”), 15:27–36 (“sound waves”).
Claim 11 depends from claim 1 and further recites “the robot is
configured to disable a cleaning function while returning to the base
charging station.” Kim discloses that while the robot is returning to the
charging station, “the vacuum motor is not operated.” See, e.g., Ex. 1003,
14:33–35; Pet. 53.
Claim 12 depends from claim 1 and further recites “the energy storage
unit comprises a battery.” Kim discloses battery 26. See, e.g., Ex. 1003,
3:57–59; Pet. 53.
IPR2020-00733
Patent 9,550,294 B2

54
Claim 13 depends from claim 1 and further recites “the energy storage
unit comprises electrical contacts arranged to automatically make electrical
contact with the base charging station upon docking of the robot.” Kim
discloses plug 156 of charging means 30, which makes electrical contact
with the robot cleaner’s conductor 31a, which is electrically connected to
battery 26. See, e.g., Ex. 1003, 8:31–41, 14:56–62, Fig. 12; Pet. 53–55.
Accordingly, we determine that Petitioner has met its burden of
demonstrating by a preponderance of the evidence that claims 2–5, 8, 9, and
11–13 are unpatentable as anticipated or obvious over Kim, or over the
combined teachings of Kim and Jeong. See supra § II.D.3–5.
E. Obviousness over the Combined Teachings of Kim and Fiegert,
or Kim, Jeong, and Fiegert
Petitioner relies upon its contentions regarding Kim (see supra
§ II.D.3–4) and the combination of Kim and Jeong (see supra § II.D.5), as
discussed above, and further contends that dependent claims 2, 4, and 5 also
would have been obvious over the combined teachings of Kim, with or
without Jeong, and Fiegert. Pet. 55–60. Other than its arguments set forth
above regarding claim 1, Patent Owner does not dispute these contentions.
PO Resp. 57.
1. Fiegert (Ex. 1006)
Fiegert is a U.S. patent that discloses an autonomous mobile unit for
surface processing that seeks to cover completely a floor surface. Ex. 1006,
1:5–8, 1:46–47, 3:19–20. Fiegert discloses that the system tracks, and
computes the percentage of, areas that have been cleaned. Id. at 6:54–67.
2. Analysis
Claim 2 depends from claim 1 and further recites, “wherein the
control system is configured to track a cleaning route of the robot and
IPR2020-00733
Patent 9,550,294 B2

55
monitor an amount of the room remaining to be cleaned.” Ex. 1001, 19:46–
49. Claim 4 depends from claim 1 and further recites, “wherein the control
system is configured to determine, through use of route-tracking subroutines
while controlling the motive system to direct the robot about the room,
which portion of the room still requires cleaning.” Id. at 20:5–9. And
claim 5 depends from claim 4 and further recites, “wherein the control
system is configured to direct the robot to the portion of the room still
requiring cleaning, following recharging the energy storage unit.” Id. at
20:10–13.
Petitioner contends that Fiegert teaches “dividing an area to be
cleaned into grid areas that ‘are at first initialized as uncleaned.’ . . . The
traversed regions ‘then can be calculated’ and ‘corresponding grid cells are
entered as cleaned in a memory field for this purpose.’” Pet. 56 (quoting
Ex. 1006, 6:55–56, 6:58–61). According to Petitioner, “this process yields
‘two numbers, the one thereby denoting an overall number of all grid cells
and the second indicating the plurality of cells that have not yet been
cleaned,’” i.e., a percentage of the cleaned area, from which a skilled artisan
would discern the percentage of the area that remains uncleaned. Id. at 56–
57 (quoting Ex. 1006, 6:61–645; citing Ex. 1020 ¶¶ 148–150).
Petitioner contends that a person of ordinary skill in the art would
have found it obvious to combine Fiegert’s teachings with Kim, or with Kim
and Jeong, because, “by adding Fiegert’s method of calculating which grid
cells remain to be cleaned, the robot cleaner described in [Kim] (and in its
combination with [Jeong]) would have determined, with greater accuracy,
which portions of the room had been cleaned and which portions remained
to be cleaned.” Pet. 57–58 (citing Ex. 1020 ¶ 151). Petitioner contends this
IPR2020-00733
Patent 9,550,294 B2

56
“would have improved [Kim’s and Kim/Jeong’s] tracking and return/resume
method” and “would further improve the autonomous nature” of Kim’s
robot. Id. at 58. Petitioner asserts this would have been the simple
combination of known elements to achieve the predictable result of
“improved tracking, increased granularity into which parts of the rooms still
required cleaning, and more efficient battery charging.” Id. (citing Ex. 1020
¶ 152).
We have considered Petitioner’s undisputed contentions and
determine that Petitioner has met its burden with respect to claims 2, 4, and
5. Petitioner shows that Fiegert discloses dividing the area “into grid areas”
where “all grid areas are at first initialized as uncleaned.” Ex. 1006, 6:54–
56. As each grid area is cleaned, “[t]he corresponding grid cells are entered
as cleaned in a memory field,” and “two numbers” are obtained: “one
thereby denoting an overall number of all grid cells and the second
indicating the plurality of cells that have not yet been cleaned. The
percentage of the area already cleaned can be determined from these two
numbers.” Id. at 6:58–65. We credit Dr. Kelly’s unrebutted testimony that,
from this information, an ordinarily skilled artisan would have calculated the
portion of the area that remains uncleaned, including through the use of
subroutines. Ex. 1020 ¶¶ 150, 157. Petitioner provides adequate reasoning
explaining why an ordinarily skilled artisan would have been motivated to
combine the teachings of Fiegert with Kim, or with Kim and Jeong, e.g., to
improve accuracy when determining the remaining areas to be cleaned. Id.
¶¶ 151–153.
Accordingly, we determine that Petitioner has met its burden of
demonstrating by a preponderance of the evidence that claims 2, 4, and 5 are
IPR2020-00733
Patent 9,550,294 B2

57
unpatentable over the combined teachings of Kim (both with and without
Jeong) and Fiegert.
F. Obviousness over the Combined Teachings of Kim and Han,
or Kim, Jeong, and Han
Petitioner relies upon its contentions regarding Kim (see supra
§ II.D.3–4) and the combination of Kim and Jeong (see supra § II.D.5), as
discussed above, and further contends that dependent claim 3 also would
have been obvious over the combined teachings of Kim, with or without
Jeong, and Han. Pet. 61–67. Patent Owner does not dispute these
contentions. PO Resp. 57.
1. Han (Ex. 1005)
Han is a U.S. patent that discloses “a charging apparatus for a robot
cleaner and a method for charging the robot cleaner by which the robot
cleaner can be accurately induced to a power source supplier to receive a
battery charge when a battery power source drops below a predetermined
level.” Ex. 1005, code (57). Han explains that the robot cleaner moves
toward the charging apparatus based on “initial position information” that
has been stored. Id. at 3:41–47, Fig. 10 (power source supplier 110).
2. Analysis
Claim 3 depends from claim 1 and further recites, “wherein the
control system is configured to use a known position of the base charging
station as a point of reference in mapping the room.” Ex. 1001, 20:1–4.
Petitioner contends that Han teaches storing initial position
information of robot cleaner 1 and power source supplier 110, prior to
beginning an initial circulation of the room. Pet. 63–64 (citing Ex. 1005,
7:57–60, Fig. 8A). Petitioner contends that “this stored position is used in
[Han’s] mapping because [Han] discloses ‘moving the cleaner to the electric
IPR2020-00733
Patent 9,550,294 B2

58
power source supplier according to the initial position information stored at
the initial position information storage step in order to charge a battery.’”
Id. at 64–65 (quoting Ex. 1005, 3:42–44; citing Ex. 1005, 3:30–35, 7:30–61,
12:39–48, Figs. 9–10; Ex. 1020 ¶ 166).
Petitioner contends that a person of ordinary skill in the art would
have found it obvious to combine Han’s teachings with Kim, or Kim and
Jeong, because this would have (1) “allow[ed] the robot to know where the
charging station was, which would enable the cleaning robot to return to the
base station as quickly as possible when it sensed a need to recharge,” and
(2) “allow[ed] the cleaning robot to find the charging station without
needing to rely solely on [Kim’s] homing signals.” Id. at 66 (citing Ex. 1020
¶ 167). Petitioner asserts this would have been the simple addition of one
known element for another to achieve the predictable result of “increasing
the likelihood that the robot cleaner would reach the base station before
depleting the battery.” Id. at 67 (citing Ex. 1020 ¶ 168).
We have considered Petitioner’s undisputed contentions and
determine that Petitioner has met its burden with respect to claim 3.
Petitioner shows that Han discloses storing “initial position information
(X,Y) in a state where the robot cleaner 1 is connected to the power source
supplier 110.” Ex. 1005, 7:55–60. Han explains that this stored position is
used as a point of reference in the room, such that the cleaner moves to that
position in order to charge the battery. Id. at 3:42–44, Figs. 8A, 9, 10.
Petitioner provides adequate reasoning explaining why an ordinarily skilled
artisan would have been motivated to combine the teachings of Han with
Kim, or with Kim and Jeong, e.g., to allow the robot cleaner to more
IPR2020-00733
Patent 9,550,294 B2

59
quickly return to the power source supplier for recharging, without
homing signals. See, e.g., Pet. 65–67; see, e.g., Ex. 1020 ¶¶ 167–168.
Accordingly, we determine that Petitioner has met its burden of
demonstrating by a preponderance of the evidence that claim 3 is
unpatentable over the combined teachings of Kim (both with and without
Jeong) and Han.
G. Obviousness over the Combined Teachings of Kim and Ohsawa,
or Kim, Jeong, and Ohsawa
Petitioner relies upon its contentions regarding Kim (see supra
§ II.D.3–4) and the combination of Kim and Jeong (see supra § II.D.5), as
discussed above, and further contends that dependent claims 6 and 7 would
have been obvious over the combined teachings of Kim, with or without
Jeong, and Ohsawa. Pet. 67–71. Patent Owner disputes Petitioner’s
contentions regarding claim 7. PO Resp. 57–61.
1. Ohsawa (Ex. 1008)
Ohsawa is a Japanese-language patent publication that discloses a
charging system for charging a mobile robot. Ex. 1008, code (57). Ohsawa
explains that when the robot detects a decline in battery charge, it
automatically stops work and docks at a charging station to recharge the
battery. Id. ¶ 15. Once the battery is partially or fully recharged, the robot
leaves the charging station and resumes the work that had been interrupted
to recharge the battery. Id.
2. Analysis of Claim 6
Claim 6 depends from claim 1 and further recites, “wherein the robot
is configured to control charging at the base charging station by only
partially charging the energy storage unit before leaving the base charging
station to continue cleaning the room.” Ex. 1001, 20:15–19.
IPR2020-00733
Patent 9,550,294 B2

60
Petitioner contends that Ohsawa teaches “recharging the battery of its
mobile robot either ‘fully’ or ‘to a certain capacity’”—i.e., partially
charging, as recited in claim 6—“before undocking to resume its work.”
Pet. 68 (citing Ex. 1008 ¶¶15, 88; Ex. 1020 ¶¶ 171–172).
Petitioner contends that a person of ordinary skill in the art would
have found it obvious to combine Ohsawa’s teachings with Kim, or with
Kim and Jeong, because this would have allowed the robot to “charge[]
more quickly (because a partial charge does not take as long as a full
charge),” which would have “reduc[ed] charging time when a small portion
of the room needs to be cleaned and thus a full charge is unnecessary.” Id.
at 68–69 (citing Ex. 1020 ¶¶ 174, 176). According to Petitioner, “[b]ecause
the [Kim] robot knows how much of the room remains to be cleaned, it
would have been obvious to a skilled artisan to use this data in combination
with [Ohsawa’s] disclosed ability to partially charge the battery, to only
charge the battery as needed to finish cleaning the room.” Id. at 69.
Petitioner contends this would have been a simple substitution of a known
element into a known system to achieve “the predictable result of a cleaning
robot that can partially charge its battery to complete the interrupted
cleaning task more quickly.” Id.
We have considered Petitioner’s undisputed contentions and
determine that Petitioner has met its burden with respect to claim 6.
Petitioner shows that Ohsawa’s robot resumes cleaning the room “[o]nce the
battery is fully charged or restored to a certain capacity.” Ex. 1008 ¶ 15.
We credit Dr. Kelly’s unrebutted testimony that a skilled artisan would have
understood this teaching to include partially charging the battery. Ex. 1020
¶ 172. Petitioner provides adequate reasoning explaining why an ordinarily
IPR2020-00733
Patent 9,550,294 B2

61
skilled artisan would have been motivated to combine the teachings of
Ohsawa with Kim, or with Kim and Jeong, e.g., to partially charge the
battery to complete an interrupted cleaning task more quickly. Pet. 69;
Ex. 1020 ¶ 174.
Accordingly, we determine that Petitioner has met its burden of
demonstrating by a preponderance of the evidence that claim 6 is
unpatentable over the combined teachings of Kim (both with and without
Jeong) and Ohsawa.
3. Analysis of Claim 7
Claim 7 depends from claim 6 and further recites, “wherein the robot
is configured to leave the base charging station and continue to clean the
room upon determining that the energy storage unit has reached a minimum
amount of stored energy necessary to complete the cleaning task in the
room.” Id. at 20:20–24.
Petitioner relies upon its contentions made regarding claim 6 and
further contends that
[Kim] prepares a cleaning path, tracks the progress of the
cleaning robot along the cleaning path, identifies a location
where the robot stopped to charge, determines the portion of the
room that has been cleaned and thus that still requires cleaning,
and resumes cleaning where it stopped to charge. Because
[Kim’s] cleaning robot knows how much of the room remains to
be cleaned, it would have been obvious to a skilled artisan to
combine this data with [Ohsawa’s] ability to partially charge the
battery, and only charge the battery as needed to finish cleaning.
Pet. 70 (citing Ex. 1020 ¶ 176). According to Petitioner, this would have
been the substitution of a known element—Ohsawa’s partial charging—into
a known system—Kim’s cleaning robot, which tracks its progress within a
room to be cleaned—“to achieve the predictable result of a cleaning robot
IPR2020-00733
Patent 9,550,294 B2

62
that can partially charge its battery to complete the interrupted cleaning task
more quickly.” Id. at 71.
Patent Owner disagrees. PO Resp. 57–61. Patent Owner argues that
“the Petition does not point to any disclosure in Kim or [Ohsawa] that
describes ‘determining that the energy storage unit has reached a minimum
amount of stored energy necessary to complete the cleaning task in the
room.’” Id. at 58. Specifically, Patent Owner argues that Kim fails to
describe “any mechanism to relate the remaining portion of a cleaning task
to an amount of energy stored in an energy storage unit.” Id. (citing
Ex. 2029 ¶ 108). Likewise, Patent Owner contends that Ohsawa fails to cure
this deficiency because it “is silent as to any relationship between the
‘certain capacity’ that the robot’s battery is partially charged to and the
current operation or task that the robot is performing.” Id. at 60 (citing
Ex. 2029 ¶¶ 110–111). Patent Owner also relies on Dr. Coates’ testimony
that “it is well known that batteries typically degrade faster when repeatedly
recharged to 100% or with excessive temperature increases” and, because of
this, Ohsawa may recharge to less than full capacity to avoid degradation—
not because of any relationship between the recharged capacity and the
portion of the room remaining to be cleaned. Id. at 60–61 (citing Ex. 2029
¶¶ 112–113).
We are persuaded by Petitioner’s contentions and evidence. First, it is
not determinative that neither Kim nor Ohsawa explicitly disclose the
precise claim language. Contra PO Resp. 58–60. This is an obviousness
ground and, therefore, “the question under 35 [U.S.C. §] 103 is not merely
what the references expressly teach but what they would have suggested to
one of ordinary skill in the art at the time the invention was made.” Merck
IPR2020-00733
Patent 9,550,294 B2

63
& Co. v. Biocraft Labs., Inc., 874 F.2d 804, 807–08 (Fed. Cir. 1989)
(quoting In re Lamberti, 545 F.2d 747, 750 (CCPA 1976) (citation
omitted)).
We are persuaded that Kim discloses tracking a robot’s cleaning
progress within a room and resuming cleaning at a location in the room at
which cleaning was paused in order to charge the robot’s battery, and that
Ohsawa teaches resuming a self-propelled, autonomous task when a robot’s
battery is “fully charged or restored to a certain capacity.” Ex. 1003, 9:51–
55, 14:63–15:2; Ex. 1008 ¶ 15. Patent Owner does not dispute these
disclosures. See, e.g., PO Resp. 58 (discussing “Kim’s disclosure of
determining the portion of the room that has been cleaned and thus that still
requires cleaning, and resuming cleaning where it stopped to charge”), 59
(quoting Ex. 1008 ¶ 15). We recognize that neither reference provides an
express disclosure sufficient for anticipation under 35 U.S.C. § 102.
However, we credit Dr. Kelly’s unrebutted testimony that “a partial
charge does not take as long as a full charge.” Ex. 1020 ¶ 174. Dr. Kelly
testifies that, in light of the teachings of Kim and Ohsawa, “it would have
been obvious to one skilled in the art . . . to only charge the battery to the
extent necessary to finish cleaning the room,” which would have been
“nothing more than the substitution of a known element ([Ohsawa’s] partial
charging) into a known system ([Kim’s] cleaning robot that tracks its
progress along a cleaning path, including how much of the room remains to
be cleaned) to achieve the predictable and desired result of a cleaning robot
that can partially charge its battery to complete the interrupted cleaning task
more quickly.” Ex. 1020 ¶ 176; see also id. ¶ 174 (explaining the
modification would have “reduce[d] charging time when only a small
IPR2020-00733
Patent 9,550,294 B2

64
portion of the room needs to be cleaned and thus a full charge is not
needed”).
Neither Patent Owner nor Dr. Coates dispute Dr. Kelly’s testimony in
this regard but instead focus on each reference’s purported deficiencies
individually. PO Resp. 57–61; Ex. 2029 ¶¶ 107–113. However, the
Supreme Court’s decision in KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398,
406 (2007) “does not support [the] theory that a person of ordinary skill can
only perform combinations of a puzzle element A with a perfectly fitting
puzzle element B. To the contrary, KSR instructs that the obviousness
inquiry requires a flexible approach,” recognizing that “‘[a] person of
ordinary skill is also a person of ordinary creativity, not an automaton.’”
ClassCo, Inc. v. Apple, Inc., 838 F.3d 1214, 1219 (Fed. Cir. 2016). In this
case, we are persuaded by Dr. Kelly’s testimony that a person of ordinary
skill in the art, recognizing that Kim already tracks the robot’s cleaning
progress through a room, would have found it obvious to charge the battery
only to the extent necessary to finish cleaning the room, i.e., to determine
that the battery has reached a sufficient charge to complete the cleaning task,
such that the cleaning task could be resumed and completed more quickly.
Ex. 1020 ¶¶ 174, 176.
We recognize Dr. Coates’ testimony that there may be other reasons
to partially recharge a battery, e.g., to avoid degradation (Ex. 2029 ¶¶ 112–
113), however, that alone does not discredit the rationale advanced by
Dr. Kelly in support of the proposed modification. Petitioner advances an
obviousness ground, not an inherency ground. Cf. Akamai Techs., Inc. v.
Cable & Wireless Internet Servs., Inc., 344 F.3d 1186, 1192 (Fed. Cir. 2003)
(“A claim limitation is inherent in the prior art if it is necessarily present in
IPR2020-00733
Patent 9,550,294 B2

65
the prior art, not merely probably or possibly present.”). In other words, that
a skilled artisan might partially recharge a battery to avoid degradation is not
at odds with Dr. Kelly’s testimony that such an artisan would have found it
obvious to recharge the robot’s battery only to the point necessary to
complete the cleaning task in the room, in light of the cited prior art
teachings. Ex. 1020 ¶¶ 174, 176.
Accordingly, we determine that Petitioner has met its burden of
demonstrating by a preponderance of the evidence that claim 7 is
unpatentable over the combined teachings of Kim (both with and without
Jeong) and Ohsawa.
H. Obviousness over the Combined Teachings of Kim and Kopmann,
or Kim, Jeong, and Kopmann
Petitioner relies upon its contentions regarding Kim (see supra
§ II.D.3–4) and the combination of Kim and Jeong (see supra § II.D.5), as
discussed above, and further contends that dependent claim 10 would have
been obvious over the combined teachings of Kim, with or without Jeong,
and Kopmann. Pet. 71–73. Other than its arguments set forth above with
respect to claim 1, Patent Owner does not dispute these contentions. PO
Resp. 57.
1. Kopmann (Ex. 1010)
Kopmann is a U.S. patent that discloses a method and apparatus for
monitoring the state of charge of a battery, including through use of
coulometry and time. Ex. 1010, code (57). Specifically, Kopmann teaches:
A method of and an apparatus for monitoring the state of
charge of a rechargeable battery . . ., which use a reference value
corresponding to a defined state of charge stored in a memory.
During each battery charging and discharging cycle, the terminal
voltage, the current and time are measured and are converted
IPR2020-00733
Patent 9,550,294 B2

66
into energy delivered to the battery or by the battery depending
on the direction of current flow. The actual state of charge is
then calculated from the reference value stored. The
characteristic terminal voltage curve during charging and
discharging is used for minimizing the differences between the
quantities measured and the actual state of charge of the battery
so monitored by monitoring the rate of change in the terminal
voltage during the charging and the discharging cycles, by
determining, after a phase of a relatively steady and small rate of
change in voltage, at least one point at which the rate of change
in voltage is higher and by adjusting the actual state of charge
determined at this point by coulometry to the predetermined
reference value.
Id. (emphasis added).
2. Analysis
Claim 10 depends from claim 8 and further recites, “wherein the robot
is configured to determine the quantity of energy in the energy storage unit
using at least one of coulometry and operating time.” Ex. 1001, 20:35–38.
Petitioner contends that Kim teaches determining the quantity of
energy in the battery (Pet. 72 (citing Ex. 1003, 1:7–11)) and that a person of
ordinary skill in the art would have understood that using coulometry and
operating time were well-known techniques to do so. Id. (citing Ex. 1020
¶¶ 179–181; Ex. 1010, code (57), 2:55–3:52, 5:36–37). Petitioner contends
that a person of ordinary skill in the art would have found it obvious to
combine Kopmann with Kim, or with Kim and Jeong, because Kopmann
provides details for measuring the energy remaining in the battery, as
suggested by Kim. Id. at 73 (citing Ex. 1020 ¶ 182). Petitioner asserts this
would have been the application of a known technique to a known system to
achieve the predictable results of “a battery monitoring functionality that
improves upon prior art coulometric methods” and “‘allows the reliable
IPR2020-00733
Patent 9,550,294 B2

67
interruption of discharging [circuits] prior to passing minimum or maximum
voltage limits, thereby warranting long life of the battery.’” Id. (citing
Ex. 1010, 7:17–22; Ex. 1020 ¶¶ 182–183).
We have considered Petitioner’s undisputed contentions and
determine that Petitioner has met its burden with respect to claim 10.
Petitioner shows that Kopmann determines the state of battery charge
utilizing a combination of operating time and coulometry. See, e.g.,
Ex. 1010, code (57), 2:28–3:52; Ex. 1020 ¶¶ 179, 181. Petitioner provides
adequate reasoning explaining why an ordinarily skilled artisan would have
been motivated to combine the teachings of Kopmann with Kim, or with
Kim and Jeong, e.g., to monitor battery charge levels. Pet. 73; Ex. 1020
¶¶ 182–183 (providing details regarding battery measurement to ensure long
battery life).
Accordingly, we determine that Petitioner has met its burden of
demonstrating by a preponderance of the evidence that claim 10 is
unpatentable over the combined teachings of Kim (both with and without
Jeong) and Kopmann.

IPR2020-00733
Patent 9,550,294 B2

68
III. CONCLUSION
In summary:19

19 Should Patent Owner wish to pursue amendment of the challenged claims
in a reissue or reexamination proceeding subsequent to the issuance of this
decision, we draw Patent Owner’s attention to the April 2019 Notice
Regarding Options for Amendments by Patent Owner Through Reissue or
Reexamination During a Pending AIA Trial Proceeding. See 84 Fed. Reg.
16,654 (Apr. 22, 2019). If Patent Owner chooses to file a reissue application
or a request for reexamination of the challenged patent, we remind Patent
Owner of its continuing obligation to notify the Board of any such related
matters in updated mandatory notices. See 37 C.F.R. § 42.8(a)(3), (b)(2).
Claims

35
U.S.C. §
Reference(s)/Basis Claims
Shown
Unpatentable
Claims
Not shown
Unpatentable
1–5, 8, 9,
11–13
§ 102 Kim 1–5, 8, 9, 11–
13

1–5, 8, 9,
11–13
§ 103 Kim 1–5, 8, 9, 11–
13

1–5, 8, 9,
11–13
§ 103 Kim, Jeong 1–5, 8, 9, 11–
13

2, 4, 5 § 103 Kim, Fiegert 2, 4, 5
2, 4, 5 § 103 Kim, Jeong,
Fiegert
2, 4, 5
3 § 103 Kim, Han 3
3 § 103 Kim, Jeong, Han 3
6, 7 § 103 Kim, Ohsawa 6, 7
6, 7 § 103 Kim, Jeong,
Ohsawa
6, 7
10 § 103 Kim, Kopmann 10
10 § 103 Kim, Jeong,
Kopmann
10
Overall
Outcome
1–13
IPR2020-00733
Patent 9,550,294 B2

69
IV. ORDER
In consideration of the foregoing, it is hereby:
ORDERED that claims 1–13 of the ’294 patent have been shown to be
unpatentable; and
FURTHER ORDERED that, because this is a final written decision,
parties to the proceeding seeking judicial review of the decision must
comply with the notice and service requirements of 37 C.F.R. § 90.2.

IPR2020-00733
Patent 9,550,294 B2

70
FOR PETITIONER:
Erika Arner
Daniel Tucker
Michael Young
Courtney Kasuboski
erika.arner@finnegan.com
daniel.tucker@finnegan.com
michael.young@finnegan.com
courtney.kasuboski@finnegan.com

FOR PATENT OWNER:
Walter Renner
Roberto Devoto
Kim Leung
axf-ptab@fr.com
devoto@fr.com
leung@fr.com