Trials@uspto.gov Paper 44
571-272-7822 Entered: October 15, 2020
UNITED STATES PATENT AND TRADEMARK OFFICE
____________
BEFORE THE PATENT TRIAL AND APPEAL BOARD
____________
MICROSOFT CORPORATION,
Petitioner,
v.
IPA TECHNOLOGIES INC.,
Patent Owner.
____________
Case IPR2019-00811
Patent 6,851,115 B1
_____________
Before LYNNE E. PETTIGREW, MINN CHUNG, and KEVIN C. TROCK,
Administrative Patent Judges.
CHUNG, Administrative Patent Judge.
JUDGMENT
Final Written Decision
Determining Some Challenged Claims Unpatentable
35 U.S.C. § 318(a)
IPR2019-00811
Patent 6,851,115 B1
2
I. INTRODUCTION
In this inter partes review, instituted pursuant to 35 U.S.C. § 314,
Microsoft Corporation (“Petitioner”) challenges the patentability of claims 1,
12–14, 26–29, 33, 35–37, 45–47, and 86–89 (the “challenged claims”) of
U.S. Patent No. 6,851,115 B1 (Ex. 1001, “the ’115 patent”), owned by IPA
Technologies Inc. (“Patent Owner”). This Final Written Decision is entered
pursuant to 35 U.S.C. § 318(a) and 37 C.F.R. § 42.73. For the reasons
discussed below, we determine Petitioner has shown by a preponderance of
the evidence that claims 1, 12–14, and 26–28 of the ’115 patent are
unpatentable, but has not proven by a preponderance of the evidence that
claims 29, 33, 35–37, 45–47, and 86–89 of the ’115 patent are unpatentable.
II. BACKGROUND
A. Procedural History
On March 19, 2019, Petitioner filed a Petition (Paper 1, “Pet.”)
requesting an inter partes review of the challenged claims of the ’115 patent.
Patent Owner filed a Preliminary Response (Paper 6, “Prelim. Resp.”).
On October 16, 2019, applying the standard set forth in 35 U.S.C.
§ 314(a), which requires demonstration of a reasonable likelihood that
Petitioner would prevail with respect to at least one challenged claim, we
instituted an inter partes review of the challenged claims. Paper 12
(“Inst. Dec.”). In the Institution Decision, we determined Petitioner
demonstrated a reasonable likelihood that it would prevail as to at least one
challenged claim, and we instituted trial on all claims and all grounds in the
Petition. Inst. Dec. 87–88.
IPR2019-00811
Patent 6,851,115 B1
3
After institution, pursuant to our order consolidating briefing
(Paper 19), the parties filed consolidated papers in in IPR2019-00810,
IPR2019-00811, IPR2019-00812, IPR2019-00813, and IPR2019-00814.
Patent Owner filed a Patent Owner Response (Paper 22, “PO Resp.”),
Petitioner filed a Reply to Patent Owner Response (Paper 30, “Pet. Reply”),
and Patent Owner filed a Corrected Sur-reply (Paper 40, “PO Sur-reply”).
In addition, in IPR2019-00810, Patent Owner moved to exclude (Paper 35) a
portion of Exhibit 1129, as well as related testimony relied on by Petitioner;
Petitioner opposed (Paper 36); and Patent Owner replied (Paper 38).
A consolidated oral hearing in IPR2019-00810, IPR2019-00811,
IPR2019-00812, IPR2019-00813, IPR2019-00814, IPR2019-00835,
IPR2019-00836, and IPR2019-00837 was held on July 27, 2020 and July 28,
2020, and a copy of the hearing transcript has been entered into the record.
Paper 43 (“Tr.”).
B. Related Matters
According to the parties, the ’115 patent is the subject of the
following district court litigation: IPA Techs. Inc. v. Google LLC, No. 1:18-
cv-00318 (D. Del.); IPA Techs. Inc. v. Microsoft Corp., No. 1:18-cv-00001
(D. Del.); and IPA Techs. Inc. v. Amazon.com, Inc., No. 1:16-cv-01266
(D. Del.). Pet. 2–3; Paper 3, 2. The ’115 patent is also the subject of
co-pending petitions for inter partes review filed by Petitioner in Cases
IPR2019-00810, IPR2019-00812, IPR2019-00813, and IPR2019-00814.
Paper 3, 2.
IPR2019-00811
Patent 6,851,115 B1
4
C. The ’115 Patent
The ’115 patent describes “software-based architectures for
communication and cooperation among distributed electronic agents.”
Ex. 1001, 1:27–29. Figure 4 of the ’115 patent is reproduced below.
Figure 4 depicts the structure of an exemplary distributed agent system of
the ’115 patent. Id. at 6:25–32. As shown in Figure 4, system 400 includes
facilitator agent 402, user interface agents 408, application agents 404, and
meta-agents 406. Id. at 6:25–28. The ’115 patent describes that system 400
is organized “as a community of peers by their common relationship” to
facilitator agent 402 (id. at 6:28–30), which is “a specialized server agent
IPR2019-00811
Patent 6,851,115 B1
5
that is responsible for coordinating agent communications and cooperative
problem-solving” (id. at 6:32–35).
According to the ’115 patent, cooperation among agents is structured
around a three-part approach as follows: (1) providers of services register
their capabilities specifications with a facilitator; (2) requesters of services
construct goals and relay them to the facilitator; and (3) the facilitator
coordinates the efforts of the appropriate service providers in satisfying
these goals. Id. at 10:42–47. The ’115 patent describes that cooperation
among agents is achieved via messages expressed in a common language,
called the Interagent Communication Language (“ICL”). Id. at 10:40–48.
Referencing Figure 3 (not reproduced herein) and Figure 4, the ’115
patent describes the operation of a distributed agent system in a preferred
embodiment. Id. at 7:13–39. The ’115 patent describes that, when invoked,
a client agent makes a connection to a facilitator, e.g., facilitator agent 402,
and registers with the facilitator a specification of the capabilities and
services it can provide. Id. at 7:15–21. For example, a natural language
agent may register the characteristics of its available natural language
vocabulary. Id. at 7:21–23. When facilitator agent 402 receives a service
request and determines that registered services 416 of one of its client agents
will help satisfy a goal of the request, the facilitator sends that client a
request expressed in ICL 418. Id. at 7:25–29. The client agent parses this
request, processes it, and returns answers or status reports to the facilitator.
Id. at 7:30–32.
Referencing Figures 5 and 6 (not reproduced herein), the ’115 patent
describes an exemplary embodiment where user interface agent 408 runs on
IPR2019-00811
Patent 6,851,115 B1
6
a user’s laptop, accepting user input, sending requests to facilitator agent 402
for delegation to appropriate agents, and displaying the results of the
distributed computation. Id. at 7:53–63. As illustration, the ’115 patent
describes that, when the question “What is my schedule?” is entered on user
interface (UI) 408, UI 408 sends the request to facilitator agent 402, which
in turn asks natural language (NL) agent 426 to translate the query into ICL.
Id. at 8:4–8. The translated ICL expression is then routed by facilitator
agent 402 to appropriate agents, e.g., calendar agent 434, to execute the
request. Id. at 8:13–15. Finally, results are sent back to UI agent 408 for
display. Id. at 8:15–16.
D. Illustrative Claim
Of the challenged claims, claims 1, 29, and 86 are independent.
Claim 1 is illustrative of the challenged claims and is reproduced below.
1. A computer-implemented method for communication and
cooperative task completion among a plurality of distributed
electronic agents, comprising the acts of:
registering a description of each active client agent’s functional
capabilities as corresponding registered functional
capabilities, using an expandable, platform-independent,
inter-agent language, wherein the inter-agent language
includes:
a layer of conversational protocol defined by event types and
parameter lists associated with one or more of the events,
wherein the parameter lists further refine the one or more
events;
a content layer comprising one or more of goals, triggers and
data elements associated with the events;
receiving a request for service as a base goal in the inter-agent
language, in the form of an arbitrarily complex goal
expression; and
IPR2019-00811
Patent 6,851,115 B1
7
dynamically interpreting the arbitrarily complex goal expression,
said act of interpreting further comprising:
generating one or more sub-goals expressed in the inter-agent
language;
constructing a goal satisfaction plan wherein the goal satisfaction
plan includes:
a suitable delegation of sub-goal requests to best complete the
requested service request—by using reasoning that includes
one or more of domain-independent coordination strategies,
domain-specific reasoning, and application-specific
reasoning comprising rules and learning algorithms; and
dispatching each of the sub-goals to a selected client agent for
performance, based on a match between the sub-goal being
dispatched and the registered functional capabilities of the
selected client agent.
Ex. 1001, 29:10–44.
E. Applied References and Declarations
Petitioner relies upon the following references in its challenges to
patentability.
Reference and Relevant Date(s) Designation Exhibit No.
U.S. Patent No. 6,484,155 B1 (filed July 21,
1999; provisional filed July 21, 1998)
Kiss1 Ex. 1005
1 For clarity and ease of reference, we only list the first named inventor.
IPR2019-00811
Patent 6,851,115 B1
8
Reference and Relevant Date(s) Designation Exhibit No.
FIPA 97 Version 1.0 Specification (Oct. 10,
1997)2
FIPA97
Exs. 1006–
10123
Douglas B. Moran, et al., Multimodal User
Interfaces in the Open Agent Architecture,
PROCEEDINGS OF 1997 INTERNATIONAL
CONFERENCE ON INTELLIGENT USER
INTERFACES 61 (1997)4
Moran Ex. 1013
Petitioner also cites the Declaration of Dr. Henry Lieberman, HDR
(Ex. 1003, “Lieberman Declaration” or “Lieberman Decl.”) in support of its
unpatentability contentions. In addition, Petitioner relies on the Declaration
of Dr. Timothy Finin, Ph.D. (Ex. 1049, “Finin Declaration” or “Finin
2 Petitioner asserts that FIPA97 was accessible to the public by October
1997 and submits various evidence in support of its contention. See Pet. 23–
25. Patent Owner contends that Petitioner has not shown sufficiently that
FIPA97 was accessible to the public before the priority date of the ’115
patent. PO Resp. 45–65. As discussed below in Section III.D., we
determine that Petitioner has shown by a preponderance of the evidence that
FIPA97 was accessible to the public by October 1997.
3 Petitioner asserts that FIPA97 was made publicly available as a single
document. Pet. 23 (citing Ex. 1055). Petitioner explains that, due to its size,
the reference was broken up into 7 parts and filed as Exhibits 1006 through
1012. Id. Patent Owner does not dispute that FIPA97 is one reference for
purposes of §§ 102 and 103. See PO Resp. 117–120. As discussed below in
Section III.C.2., we agree with Petitioner that FIPA97 constitutes a single
prior art reference, and we treat it as such for purposes of our unpatentability
analysis.
4 Petitioner contends that Moran was published in 1997. Pet. 27. Patent
Owner does not dispute the publication date of Moran. For the reasons
explained below, we determine that Petitioner has shown sufficiently that
Moran was accessible to the public in 1997.
IPR2019-00811
Patent 6,851,115 B1
9
Decl.”). Patent Owner relies on the Declaration of Nenad Medvidovic,
Ph.D. (Ex. 2032, “Medvidovic Declaration” or “Medvidovic Decl.”) and the
Declaration of Philip R. Cohen, Ph.D. (Ex. 2033, “Cohen Declaration” or
“Cohen Decl.”) in support of its Patent Owner Response.
F. Instituted Grounds of Unpatentability
Petitioner asserts the following grounds of unpatentability:
Claims Challenged 35 U.S.C. § Basis
1, 26–29, 33, 45–47, 86–89 § 103(a)5 Kiss and FIPA97
12–14, 35–37 § 103(a) Kiss, FIPA97, and Moran
Pet. 3–4.
III. ANALYSIS
A. Level of Ordinary Skill in the Art
We begin our analysis by addressing the level of ordinary skill in the
art. Petitioner’s declarant, Dr. Lieberman, opines that a person of ordinary
skill in the art at the time of the invention of the ’115 patent would have
been someone familiar with the principles and conventions of computer
science and computer networking, and also with multi-agent systems and
5 The Leahy-Smith America Invents Act, Pub. L. No. 112-29, 125 Stat. 284
(2011) (“AIA”), amended 35 U.S.C. §§ 102 and 103. Because the
’115 patent has an effective filing date prior to the effective date of the
applicable AIA amendments, we refer to the pre-AIA versions of §§ 102 and
103.
IPR2019-00811
Patent 6,851,115 B1
10
inter-agent communication languages as documented in agent-centered
literature by 1999. Ex. 1003 ¶ 34; see Pet. 4–5 (citing Ex. 1003 ¶ 34).
Dr. Lieberman also testifies that a person of ordinary skill in the art may
have been a graduate student in mathematics, engineering, or computer
science, and may have had an advanced degree in one of these disciplines,
and would also have had at least two years of experience working in the
field of computer science, or a related field, and may have worked in
academia, either as a professor or a graduate student, for a technology
company, or for a government. Id.
Citing the testimony of its declarant, Dr. Medvidovic, Patent Owner
asserts that a person of ordinary skill in the art at the time of the ’115
patent’s filing date would have had “a Bachelor’s degree in Computer
Science or equivalent field and at least two years of work experience in
design and development of distributed systems, software specification
languages, or a related area.” PO Resp. 23 (citing Ex. 2032 ¶ 37). Patent
Owner criticizes Petitioner’s articulation of the level of ordinary skill in the
art as being “vague and uncertain” due to Petitioner’s and Dr. Lieberman’s
use of the word “may” in their proposed definition. Id. at 23–24. Patent
Owner contends that Petitioner’s proposed definition “leav[es] every aspect
of the level of skill ‘fluid,’” which is “particularly problematic here, where
Petitioner claims that it would have been obvious to create an agent system
that is very complex and advanced in its facilitation, goal processing, and
inter-agent communication capacities.” Id. at 24 (citing Ex. 2032 ¶ 40).
Petitioner responds that its proposal “simply (and realistically)
expresses alternatives—i.e., that a Skilled Artisan would have been at least a
IPR2019-00811
Patent 6,851,115 B1
11
graduate student in several relevant fields (e.g, ‘mathematics, engineering, or
computer science’), and would have had at least two years of work
experience in those fields.” Pet. Reply 13–14 (citing Pet. 4–5). Petitioner
also argues that Patent Owner does not articulate how any differences
between the parties’ proposals that would alter or even affect the outcome of
this proceeding. Id. at 14.
Patent Owner asserts that it has “identified how the difference
between the proposals would affect this proceeding” because Patent Owner
has pointed out that “it is Petitioner’s burden under the law to address each
Graham factor and Petitioner has failed to proffer a coherent level of
ordinary skill in the art, Petitioner’s obviousness assertion fails for that
reason alone.” PO Sur-reply 15 (citing PO Resp. 24).
But Patent Owner does not identify any “meaningful differences”
between the parties’ proposed definitions or how “the outcome of [this
proceeding] would have been different” if we adopted Patent Owner’s
proposed definition, as opposed Petitioner’s proposal. See ESIP Series 2,
LLC v. Puzhen Life USA, LLC, 958 F.3d 1378, 1385 (Fed. Cir. 2020) (citing
Genzyme Therapeutic Prod. Ltd. P’ship v. Biomarin Pharm. Inc., 825 F.3d
1360, 1371–72 (Fed. Cir. 2016)). For example, Patent Owner does not argue
the challenged claims are not unpatentable under Patent Owner’s proposed
definition of the level of ordinary skill in the art, nor does it identify any
claim limitation that would have been taught by the asserted prior art under
Petitioner’s proposed definition, but not under Patent Owner’s proposal. See
PO Resp. 23–24; PO Sur-reply 14–16.
IPR2019-00811
Patent 6,851,115 B1
12
In our view, the parties’ proposals are not materially different despite
the differences in wording between them (e.g., the education level of a
Bachelor’s degree in Computer Science or equivalent field under Patent
Owner’s proposal, as opposed to a graduate student’s level of education with
or without a Bachelor’s degree under Petitioner’s proposed definition). See
Pet. 4–5; Pet. Reply 13–14; PO Resp. 23. For purposes of this Final Written
Decision, we find no meaningful differences between the parties’ respective
definitions that would materially alter the outcome of this Decision. These
proposals are similar for all purposes relevant to this Final Written Decision,
and both are consistent with the level of ordinary skill in the art reflected in
the disclosure of the ’115 patent and the prior art of record. See Okajima v.
Bourdeau, 261 F.3d 1350, 1355 (Fed. Cir. 2001); In re GPAC Inc., 57 F.3d
1573, 1579 (Fed. Cir. 1995).
Nonetheless, for this Final Written Decision, we adopt Patent Owner’s
definition of a person of ordinary skill in the art at the time of the claimed
invention. That is, a person of ordinary skill in the art at the time of the
claimed invention would have had “a Bachelor’s degree in Computer
Science or equivalent field and at least two years of work experience in
design and development of distributed systems, software specification
languages, or a related area.” Our analysis and conclusions in this Final
Written Decision would be the same regardless of whether Petitioner’s or
Patent Owner’s definition of the level of ordinary skill in the art is adopted.
B. Claim Construction
In an inter partes review, we apply the same claim construction
standard that would be used in a civil action under 35 U.S.C. § 282(b),
IPR2019-00811
Patent 6,851,115 B1
13
following the standard articulated in Phillips v. AWH Corp., 415 F.3d 1303
(Fed. Cir. 2005) (en banc). 37 C.F.R. § 42.100(b) (2019). In applying such
standard, claim terms are generally given their ordinary and customary
meaning, as would be understood by a person of ordinary skill in the art, at
the time of the invention and in the context of the entire patent disclosure.
Phillips, 415 F.3d at 1312–13. “In determining the meaning of the disputed
claim limitation, we look principally to the intrinsic evidence of record,
examining the claim language itself, the written description, and the
prosecution history, if in evidence.” DePuy Spine, Inc. v. Medtronic
Sofamor Danek, Inc., 469 F.3d 1005, 1014 (Fed. Cir. 2006) (citing Phillips,
415 F.3d at 1312–17).
In our Decision on Institution, we preliminarily interpreted five claim
terms as follows.
Term Construction
“event” “a message or goal communicated between agents”
“event type” “a type of an event”
“goal” “a request for service”
“arbitrarily complex
goal expression”
“a goal expressed in a language or syntax that
allows an expression, when appropriate or when
desired, that expresses multiple sub-goals and can
potentially include more than one type of logical
connector (e.g., AND, OR, NOT), and/or more
than one level of logical nesting (e.g., use of
parentheses), or the substantive equivalent,
although not every goal is itself necessarily
complex”
IPR2019-00811
Patent 6,851,115 B1
14
Term Construction
“compound goal” “a single goal expression that specifies multiple
sub-goals to be performed”
Inst. Dec. 29–38.
The parties do not dispute the constructions of these five terms in their
Patent Owner Response or Petitioner Reply. See PO Resp. 25–27;
Pet. Reply 1. Upon considering the complete record, we discern no reason
to deviate from our preliminary constructions and, therefore, adopt the
constructions of the claim terms “event,” “event type,” “goal,” “arbitrarily
complex goal expression,” and “compound goal” as set forth above for this
Final Written Decision.
In the Petition, Petitioner discusses constructions for six additional
claim terms, including “goal satisfaction plan,” “layer of conversational
protocol,” “content layer,” “trigger,” “symbolic name,” and “parameter lists
further refine the one or more events.” Pet. 8–16. In addition, in the
Petitioner Reply, Petitioner discusses construction of the claim term
“process characteristics.” Pet. Reply 12–13. Patent Owner disputes
constructions for two terms, “goal satisfaction plan” and “process
characteristics.” See PO Resp. 24–38; PO Sur-reply 13–14. For this Final
Written Decision, we need only to construe the term “goal satisfaction plan.”
Patent Owner does not discuss Petitioner’s proposed construction for
the claim term “parameter lists further refine the one or more events” or
propose its own construction for the term. See PO Resp. 24–38.
Nonetheless, Patent Owner discusses meaning of this term in the context of
Patent Owner’s arguments that FIAP97 does not teach an “expandable . . .
IPR2019-00811
Patent 6,851,115 B1
15
inter-agent language” and “parameter lists” that “further refine the one or
more events,” as recited in claim 1. See id. at 97–106. Although Patent
Owner’s arguments raise an issue of claim construction, Patent Owner’s
arguments are closely related to and interspersed with Patent Owner’s
arguments that FIPA97 does not teach an “inter-agent language” recited in
claim 1. Thus, for efficiency and completeness, we address this issue in the
context of the patentability discussion below.
No other claim terms need to be construed expressly for purposes of
this Final Written Decision. See Vivid Techs., Inc. v. Am. Sci. & Eng’g, Inc.,
200 F.3d 795, 803 (Fed. Cir. 1999) (holding that only terms that are in
controversy need to be construed, and “only to the extent necessary to
resolve the controversy”); see also Nidec Motor Corp. v. Zhongshan Broad
Ocean Motor Co., 868 F.3d 1013, 1017 (Fed. Cir. 2017) (applying Vivid
Techs. in the context of an inter partes review).
1. “goal satisfaction plan”
The claim term “goal satisfaction plan” is recited in all but one of
independent claims of the ’115 patent—claims 1, 29, 61, 71, and 86, the one
exception being claim 48. See Ex. 1001, 29:33–40 (claim 1), 32:9–21
(claim 29), 35:24–29 (claim 61), 36:17–23 (claim 71), 37:22–38:5
(claim 86). “[W]e presume, unless otherwise compelled, that the same claim
term in the same patent or related patents carries the same construed
meaning.” Paice LLC v. Ford Motor Co., 881 F.3d 894, 904 (Fed. Cir.
2018) (quoting Omega Eng’g, Inc. v. Raytek Corp., 334 F.3d 1314, 1334
(Fed. Cir. 2003). The parties do not contend that the term “goal satisfaction
IPR2019-00811
Patent 6,851,115 B1
16
plan” has different meanings in different claims of the ’115 patent. See Pet.
10; PO Resp. 27–38; Pet. Reply 1–12; PO Sur-reply 2–13.
Petitioner contends that a “goal satisfaction plan” is a “procedure for
sending one or more requests for service to one or more agents in order to
satisfy a goal.” Pet. 10. Patent Owner disagrees and asserts that the claim
term “goal satisfaction plan” should be construed to mean “a plan for the
satisfaction of a complex goal expression in an optimal or near-optimal
manner that is consistent with any advice parameters or constraints.” PO
Resp. 28.
a. Claim Language
We begin our claim construction analysis by considering the language
of the claims themselves. Phillips, 415 F.3d at 1314. Addressing the claim
language of the disputed term, we note that the claim language “a goal
satisfaction plan” indicates that the term’s plain meaning is “a plan for
satisfying a goal.”
In the challenged independent claims 1 and 29, the term “goal
satisfaction plan” is further limited (in different ways) by specific limitations
recited in the claims. For example, claim 1 recites:
constructing a goal satisfaction plan wherein the goal
satisfaction plan includes:
a suitable delegation of sub-goal requests to best complete the
requested service request—by using reasoning that includes
one or more of domain-independent coordination strategies,
domain-specific reasoning, and application-specific
reasoning comprising rules and learning algorithms;
Ex. 1001, 23:33–40 (emphases added). In contrast, claim 29 recites:
IPR2019-00811
Patent 6,851,115 B1
17
constructing a base goal satisfaction plan including the sub-acts
of:
determining whether the request service is available,
determining sub-goals required in completing the base goal
by using reasoning that includes one or more of domain-
independent coordination strategies, domain-specific
reasoning, and application-specific reasoning comprising
rules and learning algorithms,
selecting service-providing electronic agents from the agent
registry suitable for performing the determined sub-goals,
and
ordering a delegation of sub-goal requests complete the
requested service;
Id. at 32:9–21 (emphasis added).
As shown above, the “optimal or near-optimal”6 limiting feature
argued by Patent Owner is recited in claim 1 (“a suitable delegation of
sub-goal requests to best complete the requested service request”), but not in
claim 29. This shows that the patentee knew how to restrict the “goal
satisfaction plan” to “best complete the requested service request,” i.e.,
satisfy a goal “in an optimal or near-optimal manner.” If the patentee had
intended to similarly restrict the “goal satisfaction plan” in claim 29, it could
have done so using the language of claim 1, but did not. See Intellectual
6 Patent Owner asserts that the term “optimal” and “near optimal” has a
well-established meaning in the context of computer engineering, citing a
technical dictionary, which defines optimization as “[t]he process of finding
the best solution to some problem, where ‘best’ accords to prestated
criteria.” PO Sur-reply 6 (emphasis added) (citing Ex. 2078, 3).
IPR2019-00811
Patent 6,851,115 B1
18
Ventures I LLC v. T-Mobile USA, Inc., 902 F.3d 1372, 1379 (Fed. Cir. 2018)
(citing Unwired Planet, LLC v. Apple Inc., 829 F.3d 1353, 1359 (Fed. Cir.
2016)).
Thus, Patent Owner’s proposed construction that generally requires a
“goal satisfaction plan” to satisfy a goal “in an optimal or near-optimal
manner” would have the effect of obfuscating the material differences in the
claim language of claim 1 and claim 29 discussed above.
In addition, for claim 1, Patent Owner’s proposed construction would
render the limitation “suitable delegation of sub-goal requests to best
complete the requested service request” recited in claim 1 superfluous. Such
a construction is presumed improper. See Digital-Vending Servs. Int’l, LLC
v. Univ. of Phoenix, Inc., 672 F.3d 1270, 1274–75 (Fed. Cir. 2012) (rejecting
the district court’s construction narrowing a term by a superfluous limitation
when the claims explicitly recited the narrowing limitation, and discussing
the “well-established rule that claims are interpreted with an eye toward
giving effect to all terms in the claim”) (quoting Bicon, Inc. v. Straumann
Co., 441 F.3d 945, 950 (Fed. Cir. 2006)) (internal quotation marks omitted);
LSI Indus., Inc. v. ImagePoint, Inc., 279 F. App’x 964, 972 (Fed. Cir. 2008)
(nonprecedential) (rejecting the district court’s construction of “display
device” as necessarily including the superfluous limitation of “internal
illumination” because other claim terms specifically recited an “illuminated
display device”); but cf. ERBE Elektromedizin GmbH v. Canady Tech. LLC,
629 F.3d 1278, 1286 (Fed. Cir. 2010) (“no canon of [claim] construction is
absolute in its application”) (citation omitted).
IPR2019-00811
Patent 6,851,115 B1
19
Considering next the Patent Owner-argued limiting feature of
satisfying a goal “consistent with any advice parameters or constraints,” this
feature is recited in claim 29, but not in claim 1. Claim 29 recites “the act of
interpreting including the sub-acts of: determining any task completion
advice provided by the base goal, and determining any task completion
constraints provided by the base goal” (Ex. 1001, 32:4–8 (emphases
added)), whereas claim 1 does not recite “advice” or “constraint” (see id. at
29:10–44). Thus, Patent Owner’s proposed construction is disfavored
because it would blur the material differences in the claim language of
claim 1 and claim 29 discussed above.
Finally, Patent Owner’s proposed construction replaces “goal” with
“complex goal expression.” Again, this feature is recited in claim 1, but not
in claim 29. Claim 1 recites “receiving a request for service as a base goal in
the inter-agent language, in the form of an arbitrarily complex goal
expression; and dynamically interpreting the arbitrarily complex goal
expression” (Ex. 1001, 29:25–30), whereas claim 29 recites “interpreting a
service request in order to determine a base goal that may be a compound,
arbitrarily complex base goal” (id. at 31:59–61) (emphasis added). Because
of the permissive language “may be” used in claim 29, a “goal” is not
necessarily required to be in the form of an arbitrarily complex goal
expression in claim 29. Again, Patent Owner’s proposed construction is
disfavored because it would blur the material differences in the claim
language of claim 1 and claim 29 discussed above.
Thus, at least based on the claim language, it is not appropriate to
limit “goal satisfaction plan” as Patent Owner contends. As discussed
IPR2019-00811
Patent 6,851,115 B1
20
above, the meaning and the scope of “goal satisfaction plan” is clear on the
face of each claim because claims 1 and 29 each recite specifically what a
“goal satisfaction plan” includes or comprises in different claims.
Therefore, absent compelling evidence to the contrary, it is neither necessary
nor appropriate to limit the meaning of the term “goal satisfaction plan”
beyond the plain meaning indicated by the claim language—i.e., “a plan for
satisfying a goal.”
b. Written Description
Turning to the Specification, Patent Owner cites the following
statement in the “Summary of the Invention” section:
[e]xtreme flexibility is achieved through an architecture
organized around the declaration of capabilities by service-
providing agents, the construction of arbitrarily complex goals
by users and service-requesting agents, and the role of
facilitators in delegating and coordinating the satisfaction of
these goals, subject to advice and constraints that may
accompany them.
PO Resp. 31 (quoting Ex. 1001, 5:3–9). The very first sentence of the
paragraph cited by Patent Owner states, however, “[a] first embodiment of
the present invention discloses a highly flexible, Software-based architecture
for constructing distributed systems.” Ex. 1001, 4:58–60 (emphasis added).
Indeed, all of the portions of the Specification cited by Patent Owner,
including the passage reproduced above, describe various embodiments of
the ’115 patent. See PO Resp. 30–33 (citing Ex. 1001, 5:3–9, 5:48–49,
15:37–39, 15:66–16:1, 16:11–21, 16:38–46, 18:35–48, 28:47–54, Fig. 11).
In general, “a particular embodiment appearing in the written
description may not be read into a claim when the claim language is broader
IPR2019-00811
Patent 6,851,115 B1
21
than the embodiment.” SuperGuide Corp. v. DirecTV Enters., Inc., 358 F.3d
870, 875 (Fed. Cir. 2004) (citation omitted). As discussed above, the
challenged independent claims 1 and 29 do not recite all of the limiting
features required by Patent Owner’s proposed construction of “goal
satisfaction plan.” Thus, unless one of the established exceptions, such as
lexicography or disavowal, applies, the challenged claims are not restricted
as Patent Owner contends. See, e.g., GE Lighting Sols., LLC v. AgiLight,
Inc., 750 F.3d 1304, 1309 (Fed. Cir. 2014) (“the specification and
prosecution history only compel departure from the plain meaning in two
instances: lexicography and disavowal”) (citation omitted). Here, Patent
Owner does not argue lexicography or disavowal. Nor does Patent Owner
explain any other reason why the Specification limits “goal satisfaction
plan” as it contends. Thus, claim construction in this case is governed by the
general principle that “a particular embodiment appearing in the written
description may not be read into a claim when the claim language is broader
than the embodiment.” SuperGuide, 358 F.3d at 875; see also WesternGeco
LLC v. ION Geophysical Corp., 889 F.3d 1308, 1323–24 (Fed. Cir. 2018)
(“It is well established that claims are not limited to preferred embodiments,
unless the specification clearly indicates otherwise.” (citing Comaper Corp.
v. Antec, Inc., 596 F.3d 1343, 1348 (Fed. Cir. 2010) (“[T]his court has
repeatedly cautioned against limiting claims to a preferred embodiment.”))).
Accordingly, none of the embodiments in the Specification cited by Patent
Owner limits “goal satisfaction plan” as Patent Owner contends.
IPR2019-00811
Patent 6,851,115 B1
22
c. Prosecution History
Turning next to the prosecution history, Patent Owner asserts that
“examples in the file history confirm the elements of ‘goal satisfaction plan’
reflected in Patent Owner’s proposed construction.” PO Resp. 35–36.
Patent Owner argues, during the prosecution of the application for the ’115
patent,
Applicant further stated that “the facilitating engine is able to use
reasoning to delegate the sub-goals to service providing agents
in such a way as ‘to best complete the requested service
request.’…The facilitating engine is able to use reasoning to
delegate the sub-goal task of roasting coffee to the service-
providing agent that can roast beans in the least amount of time
because the facilitating engine has reasoned that the least amount
of time taken to make coffee is the best way to accomplish the
base goal of making coffee.” Applicant concluded that “the base
goal is carried out not by merely parsing the request into sub-
goals” but rather “the facilitating engine used reasoning to decide
upon using competing message transfer agents to reminding Bob
of lunch, in lieu of delegating the task to just one message
transfer agent.”
Id. at 35 (citing Ex. 1002, 172–173). As discussed above, however, the
limiting feature argued by Patent Owner is recited in claim 1 (“a suitable
delegation of sub-goal requests to best complete the requested service
request”), but not in claim 29. Thus, Patent Owner’s argument is
unpersuasive for the same reasons discussed above.
d. Extrinsic Evidence
Patent Owner also cites the testimony of its declarant,
Dr. Medvidovic, in support of Patent Owner’s proposed construction of
“goal satisfaction plan.” PO Resp. 36–37 (citing Ex. 2032 ¶¶ 51–52, 54–59,
IPR2019-00811
Patent 6,851,115 B1
23
129). We have reviewed the cited testimony from Dr. Medvidovic, but do
not find anything in his testimony that would change our analysis based on
intrinsic record discussed above. See Phillips, 415 F.3d at 1318 (authorizing
the consideration of extrinsic evidence in determining the meaning of claims
but noting that it is “in general . . . less reliable than the patent and its
prosecution history in determining how to read claim terms”).
e. Other Arguments
The parties dispute whether Petitioner’s use of “procedure” (in place
of a “plan”) and “sending one or more requests for service” in Petitioner’s
proposed construction is appropriate. See PO Resp. 37–38; Pet. Reply 2–4,
11–12; PO Sur-reply 3, 13. We need not address these arguments in detail
because, for the reasons discussed above, we see no reason to depart from
the plain meaning of the term “goal satisfaction plan.”
f. Conclusion
Based on the complete record and after examining the claims as a
whole, the Specification, and the prosecution history, we construe the term
“a goal satisfaction plan” according to its plain meaning—“a plan for
satisfying a goal.”
C. Prior Art Overview
Before discussing the obviousness challenges presented in the
Petition, we provide an overview of Kiss and FIPA97, the main prior art
references asserted against the challenged claims, to provide context for the
discussion that follows.
IPR2019-00811
Patent 6,851,115 B1
24
1. Overview of Kiss (Ex. 1005)
Kiss describes “a knowledge management system that supports
inquiries of distributed knowledge resources,” as illustrated in Figure 1.
Ex. 1005, 2:44–45.
Figure 1 of Kiss is reproduced below.
Kiss’s Figure 1 shows a conceptual overview of one embodiment of a
knowledge management system. Id. at 5:1–2. Kiss’s knowledge
management system 100 includes a presentation layer 103, a user interface
layer 105, a meta agent layer 107, a knowledge agent layer 109, a
knowledge module layer 111, and an agent service layer 113. Id. at 5:3–7.
The meta agent layer 107 includes one or more meta agents 119 that
are responsible for analyzing queries or problem formulations provided by
the user interface layer 105 and constructing a plan for finding a solution to
the problem. Id. at 5:20–24. More specifically, “meta agent 119 is
responsible for formulating a dynamic ‘solution plan’ for the distributed
IPR2019-00811
Patent 6,851,115 B1
25
inferencing to be performed by the system 100, and allocates tasks to the
knowledge agent layer 109 in furtherance of the solution plan.” Ex. 1005,
5:33–37. “The meta agent solution plan attempts to employ all elements of
information contained in the parsed inquiry in the development of search-
space constraints.” Id. at 8:34–37.
The knowledge agent layer 109 includes multiple knowledge agents
121, each of which may be associated with one or more knowledge modules
123 in the knowledge module layer 111. Ex. 1005, 6:31–35. The
knowledge agent 121 is configured to accept from a meta agent 119 a
problem statement and convert that problem statement into a format
appropriate for the knowledge module 123 associated with the knowledge
agent 121. Id. at 6:35–38.
2. Overview of FIPA97 (Exs. 1006–1012)
FIPA97 (Exs. 1006–1012) is a specification created by the Foundation
for Intelligent Physical Agents (FIPA) beginning in 1996 with an official
release date in October 1997. Ex. 1006, Cover, iv (identifying specification
as FIPA 97 Version 1.0 issued on October 10, 1997). FIPA97 is a
“specification of basic agent technologies that can be integrated by agent
systems developers to make complex systems with a high degree of
interoperability.” Id. at v. The specification comprises seven parts: three
parts for basic agent technologies (Parts 1–3 relating to agent management,
agent communication language, and agent/software integration) and four
informative applications providing examples of how the technologies can be
applied (Parts 4–7 describing personal travel assistance, personal assistant,
audio-visual entertainment, and broadcasting and network management and
IPR2019-00811
Patent 6,851,115 B1
26
provisioning applications). Id. at v–vii; see generally Exs. 1006–1008
(Parts 1–3); Exs. 1009–1012 (Parts 4–7).
Part 1 of FIPA97 “provides a normative framework within which
FIPA compliant agents can exist, operate and be managed.” Ex. 1006, vi.
Part 2 of FIPA97 describes an “Agent Communication Language (ACL)”
that “is based on speech act theory: messages are actions, or communicative
acts, as they are intended to perform some action by virtue of being sent.”
Id. The ACL specification “consists of a set of message types and the
description of their pragmatics, [i.e.,] the effects on the mental attitudes of
the sender and receiver agents.” Id. The ACL specification “also provides
the normative description of a set of high-level interaction protocols,
including requesting an action, contract net and several kinds of auctions.”
Id.
Part 2 of FIPA97 discloses a “simple abstract model of inter-agent
communication” including ACL messages having components as illustrated
in Figure 1 below. Ex. 1007, 7, 12.
Figure 1, above, illustrates the main structural elements of an ACL message
in FIPA97. Id. at 12. “The first element of the message is a word which
identifies the communicative act being communicated, which defines the
IPR2019-00811
Patent 6,851,115 B1
27
principal meaning of the message.” Id. Each ACL message contains a set of
one or more parameters, including a mandatory “:receiver” parameter that
identifies the intended recipient of the message. Id. at 13.
Petitioner asserts that FIPA97 was made publicly available as a single
document and relies on FIPA97 as a single printed publication for purposes
of 35 U.S.C. § 102(a). See Pet. 23; Pet. Reply 34–35. As discussed below
in the section addressing the status of FIPA97 as a printed publication, FIPA
approved and released FIPA 97 Version 1.0 as a single specification. See
infra § III.D.; Ex. 1026, 5; Ex. 1050 ¶ 165. Thus, although sometimes the
seven parts are referred to as “documents,” a person of ordinary skill in the
art would have considered FIPA97 to be a single reference. See Ex. 1003
¶¶ 143–148. Patent Owner does not dispute that FIPA97 is one reference for
purposes of §§ 102 and 103. See PO Resp. 117–120.7 Because FIPA and
skilled artisans considered FIPA97 to be a single specification, we agree that
FIPA97 constitutes a single prior art reference, and we treat it as such for
purposes of our unpatentability analysis.
7 Contrary to Patent Owner’s argument (PO Resp. 117–120), Petitioner did
not argue in another case, and the Board did not find, that FIPA97 is seven
separate prior art references. See Microsoft Corp. v. IPA Techs. Inc.,
IPR2019-00838, Paper 13 at 3–5 (PTAB Apr. 3, 2020) (Decision Denying
Petitioner Request for Rehearing), Paper 10 at 4–9 (Petitioner’s Request for
Rehearing, filed Dec. 6, 2019).
IPR2019-00811
Patent 6,851,115 B1
28
D. Status of FIPA97 as a Prior Art Printed Publication
Before reaching the merits of Petitioner’s obviousness contentions, all
of which are based in part on FIPA97 (Exhibits 1006–1012), we must
determine as a threshold matter whether FIPA97 is a prior art printed
publication under 35 U.S.C. § 102(a) or § 102(b). See Pet. 23–25.
Petitioner bears the burden of establishing by a preponderance of the
evidence that FIPA97 is a printed publication. See 35 U.S.C. § 316(e);
Nobel Biocare Servs. AG v. Instradent USA, Inc., 903 F.3d 1365, 1375 (Fed.
Cir. 2018); Hulu, LLC v. Sound View Innovations, LLC, IPR2018-01039,
Paper 29 at 11 (PTAB Dec. 20, 2019) (precedential) (citing Nobel Biocare,
903 F.3d at 1375). For purposes of instituting an inter partes review we
accepted Petitioner’s contention, unchallenged in Patent Owner’s
Preliminary Response, that FIPA97 was available as prior art as of
October 1997, more than one year before January 5, 1999, the filing date of
the ’115 patent. Inst. Dec. 43–45 (citing Pet. 23–25 (citing Ex. 1049 ¶¶ 3, 6,
11, 12, 20–34; Ex. 1003 ¶¶ 121, 122, 125–128, 143–148)). Patent Owner,
however, challenges that contention in its Patent Owner Response, and the
parties further address the issue in Petitioner’s Reply and Patent Owner’s
Sur-reply. See PO Resp. 45–65; Pet. Reply 17–29; PO Sur-reply 20–35.
The determination of whether a document is a “printed publication”
under 35 U.S.C. § 102 “involves a case-by-case inquiry into the facts and
circumstances surrounding the reference’s disclosure to members of the
public.” In re Klopfenstein, 380 F.3d 1345, 1350 (Fed. Cir. 2004).
“Because there are many ways in which a reference may be disseminated to
the interested public, ‘public accessibility’ has been called the touchstone in
IPR2019-00811
Patent 6,851,115 B1
29
determining whether a reference constitutes a ‘printed publication’ . . . .”
Jazz Pharm., Inc. v. Amneal Pharm., LLC, 895 F.3d 1347, 1355 (Fed. Cir.
2018) (quoting In re Hall, 781 F.2d 897, 898–99 (Fed. Cir. 1986)). A
reference is considered publicly accessible if it was “disseminated or
otherwise made available to the extent that persons interested and ordinarily
skilled in the subject matter or art, exercising reasonable diligence, can
locate it.’” Id. at 1355–56 (quoting In re Wyer, 655 F.2d 221, 226 (CCPA
1981)).
In the discussion below, we begin with background information
regarding FIPA and an overview of the evidence submitted by Petitioner in
support of its contention that FIPA97 was publicly accessible in October
1997. The evidence includes extensive testimony from Dr. Finin, a
professor who has personal knowledge of FIPA’s activities and the
development process of FIPA97. See Ex. 1049 ¶¶ 1–55 (Finin Decl.).
Dr. Finin was one of the first two FIPA Fellows invited to provide “high
quality and independent advice to FIPA.” Id. ¶ 18 (quoting Ex. 1076, 1);
Ex. 1025, 1; Ex. 1062, 1. The evidence also includes testimony from
Petitioner’s expert witness, Dr. Lieberman, and more than sixty
documentary exhibits on which Petitioner relies to show that FIPA97 was
publicly accessible in late 1997 and early 1998. See Ex. 1003 ¶¶ 121–128,
143–148; Exs. 1021–33, 1047, 1051–1105.
After providing an overview of the evidence, we consider the parties’
arguments as to whether FIPA97 was publicly accessible before the
January 5, 1999, critical date. For the reasons explained below, we
IPR2019-00811
Patent 6,851,115 B1
30
determine that Petitioner has met its burden to show that FIPA97 was
publicly accessible before the critical date.
1. FIPA
FIPA was a non-profit association based in Geneva, Switzerland,
created to promote agent-based technology and develop open standards.
Ex. 1006, iv (FIPA 97 Specification Version 1.0, Part 1, Foreword);
Ex. 1049 ¶ 3. Founded in 1996, FIPA’s membership included numerous
technology companies, educational institutions, and governmental entities.
Ex. 1049 ¶ 3. As of October 1997, FIPA had thirty-five corporate members
representing twelve countries.8 Ex. 1006, iv. According to FIPA
documentation, membership was “open to any corporation and individual
firm, partnership, governmental body or international organi[z]ation without
restriction.” Id. FIPA stated its intent “to make the results of its activities
available to all interested parties.” Id. Throughout its tenure, FIPA
produced standards designed to promote and advance agent-based
technology. Ex. 1049 ¶ 3. In 2005, FIPA was incorporated into the Institute
of Electrical and Electronics Engineers (IEEE) as one of its standards
committees. Id.
2. Development of FIPA97
The process of drafting the FIPA 97 specification began with FIPA’s
first meeting in London in April 1996. Ex. 1049 ¶ 6; Ex. 1058 (Main results
8 The corporate members included Alcatel, British Telecommunications,
Deutsche Telekom, France Telecom, Hitachi, Hewlett-Packard, IBM, NEC,
NTT, Nortel, Siemens, and Toshiba. See Ex. 1095, 22.
IPR2019-00811
Patent 6,851,115 B1
31
of London meeting); Ex. 1066 (Resolutions of London meeting).
Representatives from twenty-six companies and organizations attended.
Ex. 1066, 1. At the meeting, the members agreed on FIPA’s mission,
drafted statutes and operational principles, and produced a work plan for a
specification that would become FIPA 97. Ex. 1058, 1; Ex. 1066, 1–3;
Ex. 1049 ¶ 6. The work plan set a December 1997 target for producing a
first completed specification. Ex. 1066, 2–3; Ex. 1049 ¶ 7.
FIPA subsequently held several meetings in 1996 and 1997. At the
second meeting in June 1996 in Yorktown Heights, NY, FIPA approved
several documents, including a framework for FIPA activity and a list of
requirements for FIPA-specified agent capabilities. Ex. 1067, 1
(Resolutions of Yorktown meeting); Ex. 1060, 1 (Results of Yorktown
meeting); Ex. 1049 ¶ 9. At the third meeting in October 1996 in Tokyo,
FIPA approved a final work plan calling for production of a FIPA
specification in October 1997. Ex. 1049 ¶ 11; Ex. 1070, 1 (work plan);
Ex. 1023 (Results of Tokyo meeting). Also at the Tokyo meeting, FIPA
produced its First Call for Proposals, which outlined in detail three
technology parts (corresponding to Parts 1–3 of FIPA 97) and four
application parts (corresponding to Parts 4–7 of FIPA 97). Ex. 1069, 1–27
(First Call for Proposals); Ex. 1049 ¶ 12.
In 1997, FIPA began to produce draft specifications. At the fourth
meeting in January 1997 in Turin, Italy, FIPA publicly released drafts of
Parts 1–4 and posted them on the FIPA home page
(http://drogo.cselt.stet.it/fipa). Ex. 1021, 1 (Results of Turin meeting);
Ex. 1049 ¶ 13. Additionally, FIPA released a Second Call for Proposals at
IPR2019-00811
Patent 6,851,115 B1
32
the Turin meeting. Ex. 1071, 1–3 (Second Call for Proposals); Ex. 1049
¶ 13. At the fifth meeting in April 1997 in Reston, VA, FIPA produced
drafts of Parts 1–7, which were publicly released and posted on the FIPA
home page. Ex. 1024 (Results of Reston meeting); Ex. 1061 (Resolutions of
Reston meeting); Ex. 1049 ¶ 17. At the sixth meeting in June 1997 in Cheju
Island, South Korea, FIPA publicly released revised drafts of Parts 1–7 and
posted them on the FIPA home page. Ex. 1025, 1 (Results of Cheju
meeting); Ex. 1062, 5 (Resolutions of Cheju meeting); Ex. 1049 ¶ 19.
According to Dr. Finin, the Reston Draft and Cheju Draft were substantially
complete versions of what would become FIPA 97. Ex. 1049 ¶¶ 17, 19.
Dr. Finin testifies that news of FIPA and its standardization efforts
“spread quickly throughout the relatively small community of software agent
researchers.” Ex. 1049 ¶ 8. For instance, the AgentWeb website, “a
meeting place for researchers in agent-based technology from 1995–2000,”
was home to a “Software Agents” mailing list and a newsletter that had more
than 1,300 subscribers in 1996. Id.; Ex. 1089, 5. Dr. Finin was “the
founder, author, and a subscriber” of the AgentWeb newsletter. Ex. 1049
¶ 8. The widely distributed AgentWeb newsletters announced upcoming
FIPA meetings, reported results of previous meetings, including calls for
proposals, and provided links to the specification drafts posted on the FIPA
home page. See id. ¶¶ 8, 10, 12, 14, 17, 19.
For example, the AgentWeb newsletter dated May 11, 1996,
announced the upcoming FIPA June 1996 Yorktown meeting, described as
the “FIPA Opening Forum.” Ex. 1090, 5; Ex. 1049 ¶ 8. The purpose of the
meeting, as stated in the newsletter, was to “refine the list of basic agent
IPR2019-00811
Patent 6,851,115 B1
33
capabilities candidate for FIPA specification; compile a first list of agent
capabilities intended for specification by end 1997; and establish the first
working groups.” Ex. 1090, 5. Likewise, the AgentWeb newsletter dated
August 12, 1996, advertised the FIPA October 1996 Tokyo meeting.
Ex. 1088, 1; Ex. 1049 ¶ 10. Thereafter, the AgentWeb newsletter dated
October 14, 1996, publicized the Call for Proposals issued at the
October 1996 Tokyo meeting and provided the deadline for submission of
proposals to be considered at the Turin meeting in January 1997. Ex. 1087,
1; Ex. 1049 ¶ 12. FIPA had decided that respondents who submitted
proposals would be invited to the January 1997 meeting to present their
proposals even if they were not FIPA members. Ex. 1023, 1; see Ex. 1069,
27.
Once FIPA began to produce drafts of the specification, AgentWeb
newsletters announced those drafts and provided links to access them. The
AgentWeb newsletter dated February 2, 1997, reported that “[i]nitial
specifications for three technology parts (Agent Management, Agent
Communication and Agent/Software Interaction) and one application part
(Personal Travel Assistance) have been produced” at the recent Turin
meeting. Ex. 1091, 1; Ex. 1049 ¶ 14. The newsletter included a freely
accessible and public link to access the draft. Ex. 1049 ¶ 14 (citing
Ex. 1091, 1–2). It also indicated that drafts of the remaining three
specification parts would be generated at the FIPA meeting to be held in
April 1997 in Reston. Ex. 1091, 1–2. A subsequent AgentWeb newsletter
dated April 13, 1997, advertised the upcoming Reston meeting and
IPR2019-00811
Patent 6,851,115 B1
34
contained links to information about the meeting on the FIPA website.
Ex. 1092, 3; Ex. 1049 ¶ 17.
On May 5, 1997, following the Reston meeting, an AgentWeb
newsletter reported that attendees of that meeting produced a revised draft
including all seven parts of the specification. Ex. 1093, 2; Ex. 1049 ¶ 17.
The newsletter contained individual public links to the seven parts and
invited “both members and non-members” to comment, providing an email
address for submission. Id.; Ex. 1049 ¶ 17. Additionally, the newsletter
announced the dates and locations of the next two meetings—the June 1997
Cheju meeting and the seventh meeting to take place in Munich on
October 6–10, 1997. Id. Similarly, the AgentWeb newsletter dated June 30,
1997, announced that the Cheju meeting resulted in revised versions of the
FIPA specification documents and provided direct links to the seven parts.
Ex. 1094, 1–2; Ex. 1049 ¶ 19. Again, the newsletter stated that FIPA invited
comments and that non-members making substantial comments would be
invited to attend the October 1997 Munich meeting. Ex. 1094, 2. Dr. Finin
was among those who submitted comments on the Cheju draft. Ex. 1073;
Ex. 1049 ¶ 20.
Meanwhile, the AgentWeb Software Agents mailing list also provided
information regarding FIPA’s efforts to the community of software agent
researchers. See Ex. 1049 ¶ 8. Dr. Finin was an organizer of and
contributor to the Software Agents mailing list. Id. On May 2, 1996, a
posting to the mailing list referenced the upcoming June 1996 FIPA meeting
in Yorktown Heights, New York. Ex. 1079, 1; Ex. 1049 ¶ 8. Members of
the Software Agents mailing list posted comments regarding the Turin draft
IPR2019-00811
Patent 6,851,115 B1
35
after it was released publicly. Exs. 1078, 1081, 1082; see Ex. 1049 ¶¶ 14–
15. Dr. Finin was among those who shared their thoughts. See Ex. 1082;
Ex. 1049 ¶ 15. A posting after the Reston meeting included a public and
open link to access the Reston draft and invited comments from the mailing
list. Ex. 1083, 1; Ex. 1049 ¶ 17.
3. Public Release of FIPA97
FIPA’s seventh meeting took place in Munich on October 6–10, 1997.
Ex. 1026 (Resolution of Munich meeting); Ex. 1049 ¶ 20. At this meeting,
FIPA incorporated final edits and comments and approved the FIPA 97
specification for publication. Ex. 1026, 1, 5; Ex. 1049 ¶ 20. The
specification as approved at the meeting was called “FIPA 97 ver. 1.0,” or
FIPA 97 Version 1.0. Ex. 1026, 5. Working groups for various parts of the
specification were given about one month to check for consistency and make
minor edits. Id. at 2; Ex. 1049 ¶ 20.
An “initial” version of FIPA 97 Version 1.0, approved at the Munich
meeting and dated October 10, 1997, was posted to the FIPA website,
housed at the time at http://drogo.cselt.stet.it/fipa. Ex. 1049 ¶¶ 20–21.
According to Dr. Finin, the specification documents were available online
and free for anybody to access by sometime in October 1997. Id. ¶ 21. On
November 18, 1997, an AgentWeb newsletter announced completion of
FIPA 97 Version 1.0 and its publication on the FIPA website. Ex. 1086, 1–
2; Ex. 1049 ¶ 21. The newsletter provided individual, direct links to the
seven parts of the specification on the FIPA website. Ex. 1086, 1–2. The
links were not password protected and were open to the public without a
requirement for secrecy or confidentiality. Ex. 1049 ¶ 21. As with the
IPR2019-00811
Patent 6,851,115 B1
36
earlier AgentWeb newsletters, the newsletter containing links to FIPA 97
Version 1.0 was sent to more than 1,300 subscribers. Id.
Dr. Finin testifies that in late 1997, “a few weeks after the Munich
meeting,” he accessed the approved FIPA 97 Version 1.0 posted on the
FIPA website. Id. ¶¶ 20–21. He specifically recalls accessing the
documents in late 1997 to review the first major work product of FIPA and
to prepare for FIPA’s upcoming January 1998 meeting in Palo Alto, which
he attended. Id. ¶ 21 (citing Ex. 1074). Dr. Lieberman also testifies that he
recalls FIPA 97 Version 1.0 being publicly available on the FIPA website
throughout late 1997 and early 1998 and that he accessed it in late 1997 as a
member of the interested public. Ex. 1003 ¶ 125.
The release of FIPA 97 Version 1.0 was well publicized in the weeks
after the Munich meeting. Ex. 1049 ¶ 22. For example, on October 20,
1997, the EETimes featured an article reporting FIPA’s approval of the
FIPA 97 specification. Ex. 1095, 22; see Ex. 1049 ¶ 22. The article was
based on an interview with Leonardo Chiariglione, the president of FIPA.
Ex. 1095, 22. The article described the seven parts of the specification:
“The first three parts cover different aspects of agent behavior: agent
management; agent communication; and agent-software interaction. The
four remaining parts, which cover application areas, . . . are: personal travel
assistance; personal assistant; audio-visual entertainment and broadcast; and
network provision and management.” Id. After describing FIPA’s plans for
1998, including testing of FIPA 97 Version 1.0 to provide input for a further
round of standardization, the article referred readers to the FIPA website for
IPR2019-00811
Patent 6,851,115 B1
37
information regarding the released specification: “CSELT[9] maintains
extensive information on FIPA’s activities at www.cselt.stet.it/fipa.”10 Id.
(emphasis omitted). It concluded with a quote from Dr. Chiariglione
regarding the specification’s public availability: “‘It is our policy to make
the standard freely available over the Internet,’ Chiariglione said. ‘FIPA
will retain the copyright but it will be free for others to use.’” Id.
The November-December 1997 issue of IEEE Internet Computing
also announced the release of the FIPA 97 specification: “The Foundation
for Intelligent Physical Agents, a non-profit organization established to
promote emerging agent-based applications, has released its first
specification. Named FIPA 97, the specification will provide a benchmark
for interoperable products.” Ex. 1096, 93; Ex. 1049 ¶ 22. The news brief
directed readers to the FIPA website for the text of the specification.
Ex. 1096, 93 (“The text is available at http://drogo.cselt.stet.it.fipa/.”11);
Ex. 1049 ¶ 22.
Sometime after the Munich meeting, FIPA also produced a hardcopy
book version of the FIPA 97 Version 1.0 specification to be sent to FIPA
members and various institutions and standards groups. Ex. 1049 ¶ 24; see
9 Centro Studi e Laboratori Telecommunicazioni, Dr. Chiariglione’s
employer at the time. See Ex. 1095, 22.
10 The record suggests this website (http://www.cselt.stet.it/fipa) was a
mirror of the FIPA website (http://drogo.cselt.stet.it/fipa) in 1997. See
Ex. 1049 ¶ 22; Ex. 1095, 22; Tr. 94:11–14.
11 The URL provided in the article contains a typographical error. See
Ex. 1049 ¶ 22 (correcting the URL to http://drogo.cselt.stet.it[/]fipa/).
IPR2019-00811
Patent 6,851,115 B1
38
Ex. 1026, 1; Ex. 1063, 1 (Resolution of the Palo Alto meeting, Jan. 29,
1998). At least two copies of the book were available in libraries in
January 2019. Ex. 1049 ¶ 24; Ex. 1057, 1 (WorldCat catalog entry for “Fipa
specification : Version 1.0”); Ex. 1055 (cover page and table of contents of
each part in copy at library in Italy); Ex. 1056 (cover page and table of
contents of Part 1 in copy at library in Switzerland). Dr. Finin testifies that
the books contain the November 28, 1997, edited version of FIPA 97
Version 1.0, which “only made minor edits (mostly spelling and formatting)
over the October 10, 1997 version.” Ex. 1049 ¶ 25 (citing Ex. 1055, 5
(Part 2 “publication date” of November 28, 1997); Ex. 1056).
4. Publications and Patents Referring to FIPA97 and FIPA Activities
In late 1997 and 1998, several papers reviewing and outlining the
FIPA 97 specification were published and presented at meetings. See
Ex. 1049 ¶¶ 26–30. A paper dated December 1997 memorializes a
presentation made to a meeting on Intelligent Agent Technology organized
by the EPSRC12 Community Club in Advanced Computing Techniques.
Ex. 1097, Cover (“Dickinson”);13 see Ex. 1049 ¶ 26. This paper outlines the
background and rationale for the creation of an agent interoperability
standard. Ex. 1097, 1–3. It also describes the development history of the
FIPA 97 specification, noting that FIPA’s work program “culminated in the
publication of an initial draft standard, FIPA 97, at the Munich meeting on
12 “The Engineering and Physical Sciences Research Council (EPSRC) is the
main funding body for engineering and physical sciences research in the
UK.” Engineering and Physical Sciences Research Council,
epsrc.ukri.org/about (last visited Sept. 28, 2020).
13 Ian J. Dickinson, Agent Standards, HP Laboratories Bristol, HPL-97-156
(Dec. 1997).
IPR2019-00811
Patent 6,851,115 B1
39
October 1997.” Id. at 3. Dickinson further provides that “[a]ll FIPA
documents and meeting outputs are publicly available on the web” at
“http://drogo.cselt.stet.it/fipa/,” and “comments and review have been sought
at each stage of the process.” Id. at 3, 7 n.9. The paper then describes the
features of the specification’s three substantive technical sections, Parts 1–3.
Id. at 3–6.
Similarly, a 1998 paper titled “Industrial Applications of Multi-Agent
Technology” summarizes FIPA’s background and presents an overview of
the three technical parts (i.e., Parts 1–3) of FIPA 97 Version 1.0 released in
October 1997. Ex. 1099, 12–13 (“Steiner”);14 see Ex. 1049 ¶ 28. The
Steiner paper provides the FIPA website address and explains that the four
application parts (i.e., Parts 4–7) will be used in field trials to gather
information to be incorporated into a revised version of FIPA 97. Ex. 1099,
13 & n.1.
Likewise, a paper titled “FIPA — towards a standard for software
agents” appearing in the July 1998 issue of the BT Technology Journal
explains the motivation for an agent standard and provides an overview of
FIPA’s background and activities. Ex. 1100, 51–52 (“O’Brien”);15 see
Ex. 1049 ¶ 29. It then describes the different parts of the FIPA 97
Version 1.0 specification issued in October 1997. Ex. 1100, 51–52.16
14 Donald Steiner, Siemens AG, Industrial Applications of Multi-Agent
Technology (IEEE 1998).
15 P.D. O’Brien & R.C. Nicol, FIPA — towards a standard for software
agents, BT Tech. J. Vol. 16, No. 3 (July 1998).
16 Dr. Finin testifies that O’Brien provides public links to FIPA97. Ex. 1049
¶ 29. The copy of O’Brien entered into the record, however, appears to be
missing several pages, including one containing endnotes 10–12 that might
IPR2019-00811
Patent 6,851,115 B1
40
Dr. Finin cites several other publications that reference FIPA’s
development of an agent communication standard in the 1997–98 timeframe.
Ex. 1049 ¶ 31 (citing Exs. 1027–29). Dr. Finin also identifies several patent
applications in the same timeframe that refer to FIPA’s agent
communication language. Id. ¶ 32 (citing Exs. 1030–33).
5. The FIPA Website and Exhibits 1006–1012
Sometime in 1998, the FIPA website migrated from its original site
(http://drogo.cselt.stet.it/fipa) to fipa.org. Ex. 1049 ¶ 23; see also Ex. 1026
(Resolution of Munich meeting indicating FIPA’s plans for the “redesign of
a new FIPA home page”); Ex. 1065, 4 (Resolutions of the Durham meeting
in October 1998 indicating the website will be transferred to www.fipa.org).
The fipa.org website still exists today even though FIPA is no longer active
as a standalone organization. See Ex. 1049 ¶¶ 23, 48–49. The current FIPA
website maintains a repository of past FIPA documents, including meeting
notes, press releases, FIPA resolutions, and different versions of the FIPA
specification. Id. ¶ 49.
Dr. Finin testifies that the “initial” version of FIPA 97 Version 1.0
(i.e., the specification posted on the FIPA website on October 10, 1997)
“remained publicly accessible during all of FIPA’s future activities and is
still available on FIPA’s home page today.” Id. ¶ 21. According to
Dr. Finin, Exhibits 1006–1012 (collectively referred to as FIPA97) are
Parts 1–7 of FIPA 97 Version 1.0 as found on the current fipa.org website.
Id. Based on his personal knowledge, Dr. Finin further testifies that
provide links to the FIPA website. See Ex. 1100, 52 (“FIPA97 (issued in
October 1997) [10–12] is the first output from FIPA covering part of the
requirements for an agent standard.”).
IPR2019-00811
Patent 6,851,115 B1
41
Exhibits 1006–1012 are the “same version” and contain “the same
disclosures, content and information” as Parts 1–7 of FIPA 97 Version 1.0
that were publicly available on the FIPA website
(http://drogo.cselt.stet.it/fipa) on October 10, 1997, and which Dr. Finin
himself accessed in late 1997. Id. ¶¶ 21, 34. He also notes that although
Exhibit 1006 has a typographical error relating to the version number in the
header of its odd pages,17 Exhibit 1006 is identical to the version of Part 1 of
FIPA 97 Version 1.0 that was released on October 10, 1007. Id. ¶ 21.
Additionally, as evidence that FIPA 97 Version 1.0 has been continuously
available on fipa.org since the website migration, Dr. Finin cites a set of
Internet Archive records beginning in 2000 showing where each part of that
version of the specification could be accessed. Id. (citing Ex. 1054).
Dr. Finin testifies that the original FIPA website,
http://drogo.cselt.stet.it/fipa, remained live for several years after the website
transitioned to fipa.org. Ex. 1049 ¶ 23. The Internet Archive Wayback
Machine contains an archived screen capture of the website
http://drogo.cselt.stet.it/fipa from December 1, 1998. See Ex. 1051;
Ex. 1049 ¶¶ 23, 54. According to Dr. Finin, the archived page shows that
the FIPA 97 specification was live and freely available on the FIPA website
home page at that time. Ex. 1049 ¶ 23; Ex. 1051, 1 (“FIPA has already
developed a seven-part specification called FIPA 97. Implementations of
FIPA 97 are undergoing field trials that will last until October 1998 when
version 2 of FIPA 97 will be produced.”); Ex. 1049 ¶ 23. The home page
17 The header on odd pages reads “FIPA 1997 Part 1: Version 2.4.” E.g.,
Ex. 1006, 3.
IPR2019-00811
Patent 6,851,115 B1
42
provided links (i.e., “FIPA 97”) to actual specification documents, but those
links have not been maintained in the archive. Ex. 1049 ¶ 23. Dr. Finin
testifies that the screen capture from the Internet Archive is consistent with
the mirror of the former FIPA site hosted by Leonardo Chiariglione. Id. ¶ 51
(citing Ex. 1103 (http://leonardo.chiariglione.or/standards/fipa/)).
Dr. Chiariglione’s mirror is still available, and documents hosted on the
mirror are “accurate and complete copies of FIPA documents that were
available to FIPA members.” Id.
6. Analysis
Petitioner argues that FIPA97 was a publicly accessible printed
publication as of October 10, 1997, when it was posted on the FIPA website.
Pet. 23; Pet. Reply 15–21. More specifically, Petitioner argues that
(1) FIPA 97 Version 1.0 was publicly accessible on October 10, 1997, and
(2) the version of the FIPA specification found in Exhibits 1006–1012,
referred to herein as FIPA97, is the same specification that was posted on
the FIPA website on October 10, 1997, and named FIPA 97 Version 1.0.
See, e.g., Pet. 24 (arguing FIPA 97 Version 1.0 was publicly accessible in
late 1997); Pet. Reply 17–20 (arguing that Exhibits 1006–1012 are the same
documents that were made public as FIPA 97 Version 1.0). Patent Owner
addresses the first contention when it argues the evidence does not show any
version of FIPA 97 Version 1.0 was available on the FIPA website. See,
e.g., PO Resp. 47–48. It also challenges the second contention when it
argues Petitioner has failed to show that Exhibits 1006–1012 are identical to
the version of the FIPA specification that allegedly was publicly accessible
before the critical date. See, e.g., id. at 48. In our analysis below, we
address the two questions in turn.
IPR2019-00811
Patent 6,851,115 B1
43
Considering all the facts and circumstances surrounding the
development of FIPA 97 Version 1.0 and its publication on the FIPA
website, we find that FIPA 97 Version 1.0 was disseminated or otherwise
made available to the extent that persons interested and ordinarily skilled in
the art, exercising reasonable diligence, could locate it. See Jazz Pharm.,
895 F.3d at 1355–56. First, the uncontested evidence shows that members
of the relevant public (i.e., persons of ordinary skill in the field of software
agent-based technology) knew of the release of FIPA 97 Version 1.0 in late
1997, prior to the critical date. As detailed above, FIPA was open to
corporations and educational and governmental organizations without
restriction. Ex. 1006, iv. FIPA’s members included many industry leaders
whose employees attended FIPA meetings, including the October 1997
meeting in Munich where FIPA 97 Version 1.0 was approved for
publication. See id.; Ex. 1095, 22. FIPA meetings and activities were
publicized among the larger community of software agent researchers via
the AgentWeb website and its associated Software Agents mailing list and
AgentWeb newsletter, which had more than 1,300 subscribers. Ex. 1090, 5;
Ex. 1088, 1; Ex. 1087, 1; Ex. 1091, 1–2; Ex. 1092, 3; Ex. 1093, 2; Ex. 1094,
1–2; Ex. 1086, 1–2; see Ex. 1049 ¶¶ 8, 10, 12, 14, 17, 19. The newsletter in
particular informed subscribers of the development of the FIPA specification
during 1996 and 1997 and provided detailed information regarding the
release of FIPA 97 Version 1.0 in a November issue. Ex. 1086, 1–2. News
articles in periodicals published in late 1997 also announced that FIPA had
released and published FIPA 97 Version 1.0. Ex. 1095, 22; Ex. 1096, 93;
see Ex. 1049 ¶ 22. In addition, the record contains several papers published
IPR2019-00811
Patent 6,851,115 B1
44
in late 1997 and 1998, prior to the critical date, describing FIPA 97
Version 1.0 and its release in October 1997. See Ex. 1097, 1–7; Ex. 1099,
12–13; Ex. 1100, 51–52; Ex. 1049 ¶¶ 26–30; see also Ex. 1049 ¶¶ 31–32
(citing Exs. 1027–33 (other publications and patents referring to FIPA
specification)).
The evidence of record also demonstrates that members of the
relevant public, exercising reasonable diligence, could have located FIPA 97
Version 1.0 in late 1997 on the FIPA website, housed at the time at
http://drogo.cselt.stet.it/fipa. FIPA published FIPA 97 Version 1.0 on its
website at the conclusion of the Munich meeting on October 10, 1997, or
shortly thereafter. See Ex. 1049 ¶ 20. In the several months leading up to
the Munich meeting, persons of ordinary skill in the art would have known
about the FIPA website through the widely distributed AgentWeb
newsletters and the Software Agents mailing list, which provided website
information or direct links to drafts of the specification and meeting
announcements and reports. See Ex. 1091, 1–2; Ex. 1092, 3; Ex. 1093, 2;
Ex. 1094; Ex. 1082, 1; Ex. 1083, 1. Therefore, people working in the field
would have been sufficiently familiar with FIPA’s activities and its website
to look for FIPA 97 Version 1.0 on the FIPA website when it was posted
after the Munich meeting. Indeed, both Dr. Finin, who was affiliated with
FIPA, and Dr. Lieberman, who worked in the software agent field, testify
that they accessed FIPA 97 Version 1.0 on the FIPA website in late 1997.
Ex. 1049 ¶ 21; Ex. 1003 ¶ 125.
Moreover, the AgentWeb newsletter dated November 18, 1997,
provided a link to the FIPA website and individual links to the seven parts of
IPR2019-00811
Patent 6,851,115 B1
45
FIPA 97 Version 1.0 published on the website. Ex. 1086, 1–2; see Ex. 1049
¶ 21. Thus, that issue of the newsletter provided more than 1,300 interested
members of the public with direct access to FIPA 97 Version 1.0 at least as
of November 18, 1997.
News articles and papers published in late 1997 also provided readers
with the website address where FIPA 97 Version 1.0 could be found. See
Ex. 1095, 22; Ex. 1096, 93; Ex. 1097, 7 n.9. The news articles in particular
likely reached a very wide audience because they appeared in publications
directed to subject matter broader than agent-based technology. See
Ex. 1095 (EETimes); Ex. 1096 (IEEE Internet Computing). The EETimes
article was dated October 20, 1997, suggesting that FIPA 97 Version 1.0 was
available on the FIPA website at least as of that date.
The undisputed evidence also demonstrates that FIPA 97 Version 1.0
was freely available on the FIPA website without a password or any other
restrictions and access was not subject to any requirement or expectation of
secrecy or confidentiality. See Ex. 1049 ¶ 21; Ex. 1095, 22; see also
Ex. 1006, iv (stating FIPA intended to make its results available to all
interested parties). This evidence further supports our finding that FIPA 97
Version 1.0 was publicly accessible in late 1997. See Jazz Pharm., 895 F.3d
at 1358–59; Voter Verified, Inc. v. Premier Election Solutions, Inc., 698 F.3d
1374, 1380–81 (Fed. Cir. 2012); MIT v. AB Fortia, 774 F.2d 1104, 1109
(Fed. Cir. 1985).
Patent Owner’s arguments, many of which attack pieces of evidence
individually, do not persuade us that Petitioner has failed to show that
FIPA 97 Version 1.0 was publicly accessible. First, Patent Owner argues
IPR2019-00811
Patent 6,851,115 B1
46
that no evidence shows the original FIPA website
(http://drogo.cselt.stet.it/fipa) existed prior to a December 1, 1998, Internet
Archive capture. PO Resp. 55 (citing Ex. 1051). To the contrary, much of
the documentary evidence from late 1997 and testimonial evidence identifies
that website or its mirror as the FIPA website during the relevant time
period. See Ex. 1049 ¶ ¶ 20–23; Ex. 1092, 3; Ex. 1093, 2; Ex. 1095, 22;
Ex. 1096, 93; Ex. 1097, 7 n.9. Notwithstanding the absence of an Internet
Archive capture for the website from a date prior to December 1998, we find
that, based on the totality of evidence in the record, the original FIPA
website existed in late 1997.
Patent Owner also argues that none of the documents cited by
Petitioner provide active links to FIPA 97 Version 1.0 and therefore are
insufficient proof of public accessibility. PO Resp. 54; PO Sur-reply 21–22.
Again, based on the totality of evidence in the record, including but not
limited to the testimony of Dr. Finin that he accessed FIPA 97 Version 1.0
on the FIPA website in late 1997, we find that FIPA 97 Version 1.0 was
available on the FIPA website at that time, notwithstanding the lack of active
links today, more than twenty years later. Patent Owner finds fault with
Dr. Finin’s testimony on this point because he does not claim to have used
any of the links in the AgentWeb newsletters to access FIPA 97 Version 1.0.
PO Sur-reply 35 (citing Ex. 1049 ¶ 21). Patent Owner, however, overlooks
Dr. Finin’s testimony that he was the founder and author of the AgentWeb
newsletter, implying that he created the links in the newsletter. See Ex. 1049
¶ 8. Therefore, based on Dr. Finin’s personal knowledge regarding the links
in the AgentWeb newsletter dated November 18, 1997, and the other
IPR2019-00811
Patent 6,851,115 B1
47
evidence corroborating his testimony, we find that FIPA 97 Version 1.0 was
available on the FIPA website in late 1997.
Patent Owner further asserts that, to the extent the Internet Archive
capture from December 1998 provides an operable download link, it links to
Version 2.0 of the FIPA 97 specification. PO Resp. 55–56 (citing Ex. 1051).
It is not surprising, however, that FIPA’s website in December 1998
provided links to FIPA 97 Version 2.0, which FIPA released in
October 1998 to provide minor updates to two parts of the specification. See
Ex. 1049 ¶ 35; Ex. 1053, vi. In any event, the December 1998 Internet
Archive capture at least shows that the FIPA website was live at that time,
which was prior to the critical date. See Ex. 1049 ¶ 23 (citing Ex. 1051, 1).
In its Sur-reply, Patent Owner argues for the first time that Petitioner
has not demonstrated that a person of ordinary skill in the art in 1997 could
have found FIPA 97 Version 1.0 using search tools or that the FIPA website
was indexed so the specification could be located. PO Sur-reply 31.
Further, Patent Owner argues that the existence of different FIPA websites at
various times would have made it even more difficult to find the correct
version of the FIPA specification without sufficient indexing or search tools.
Id. at 33.
Even if we consider this late argument, we find it unpersuasive. As
detailed above, the uncontested evidence shows that the address of the FIPA
website itself was widely known among those skilled in the art and that
FIPA routinely provided links to the FIPA specification on the website’s
home page. See Ex. 1095, 22; Ex. 1096, 93; Ex. 1097, 7 n.9; Ex. 1051, 1;
Ex. 1021, 1; Ex. 1025, 1. The November 1997 AgentWeb newsletter also
IPR2019-00811
Patent 6,851,115 B1
48
provided direct links to FIPA 97 Version 1.0. Ex. 1086, 1–2. Under these
circumstances, we find that a member of the interested public could have
found FIPA 97 Version 1.0 without search tools or indexing by using the
links in the AgentWeb newsletter or navigating to the well-known FIPA
website home page. We also find that the record, including Dr. Finin’s
testimony, satisfactorily explains the various FIPA websites and website
addresses over time and clearly identifies the FIPA website that was being
used in late 1997. See Ex. 1049 ¶¶ 22–23, 48–49, 51.
Having determined that FIPA 97 Version 1.0 was publicly accessible
on the FIPA website in late 1997, as early as October 10 and no later than
November 18, we turn to whether Exhibits 1006–1012 (collectively referred
to as FIPA97) contain the same version of the FIPA specification that was
available on the FIPA website during that timeframe. Dr. Finin testifies that
Exhibits 1006–1012, documents available on the current FIPA website
(fipa.org) and identified as FIPA 97 Version 1.0, are the same version with
the same content as Parts 1–7 of FIPA 97 Version 1.0 that were released at
the FIPA Munich meeting on October 10, 1997, and were publicly available
on the FIPA website (http://drogo.cselt.stet.it/fipa) in late 1997. Ex. 1049
¶¶ 21, 34.
As an initial matter, the October 10, 1997, date on the cover pages of
several parts of FIPA97 supports Dr. Finin’s testimony that FIPA97 is the
same as FIPA 97 Version 1.0 as published on the FIPA website in late 1997.
See Ex. 1006, Cover; Ex. 1007, Cover; Ex. 1008, Cover; Ex. 1010, Cover;
Ex. 1012, Cover; see also Ex. 1009, Cover (Part 4 identifying a 1997
copyright date); Ex. 1011, Cover (Part 6 identifying a 1997 copyright date).
IPR2019-00811
Patent 6,851,115 B1
49
Patent Owner contends that Petitioner cannot rely on the date stamps or
copyright notices on Exhibits 1006–1012 to establish public accessibility
because they are hearsay. PO Resp. 47; PO Sur-reply 23–25. We agree with
Petitioner that Patent Owner has waived this argument because it did not
raise an objection on evidentiary grounds pursuant to 37 C.F.R.
§ 42.64(b)(1). See Pet. Reply 18. Furthermore, we agree with Petitioner
that the dates on the cover pages are not hearsay because they are evidence
tending to show that FIPA97 is identical to FIPA 97 Version 1.0, not
statements offered to show Exhibits 1006–1012 were published on a certain
date. See id. (citing Apple Inc. v. VirnetX Inc., IPR2016-01585, Paper 32 at
58 (PTAB Feb. 20, 2018)).
Patent Owner also argues that Petitioner cannot rely on the cover page
dates alone as proof of the documents’ public accessibility. PO Sur-
reply 23. The date stamps, however, are only part of the totality of evidence
offered by Petitioner to establish public accessibility. As such, they are
relevant evidence supporting a finding that FIPA97 was publicly accessible.
See Hulu, Paper 29 at 17–18 (citing Nobel Biocare, 903 F.3d at 1377).
Patent Owner argues that the existence of multiple versions of
FIPA 97 Version 1.0 undercuts Petitioner’s position and Dr. Finin’s
testimony that Exhibits 1006–1012 are the version of FIPA 97 Version 1.0
that was publicly accessible in late 1997. See PO Resp. 48–53; PO Sur-
reply 27. For example, Patent Owner points to Dr. Finin’s description of the
version released on October 10, 1997, as an “initial” version, after which
final minor edits were to be made. PO Resp. 50 (citing Ex. 2068, 86:11–15,
88:9–89:25); see Ex. 1049 ¶ 21. Patent Owner cites an Internet Archive
IPR2019-00811
Patent 6,851,115 B1
50
capture in the record referring to “FIPA 97 specification ver. 1.0 (Reston
meeting).” PO Resp. 50–51 (citing Ex. 1105, 27). Patent Owner also cites
evidence, including Dr. Finin’s testimony, that a different version of
FIPA 97 Version 1.0 was published in hardcopy books. Id. at 51 (citing
Exs. 1055, 1056; Ex. 1049 ¶ 25 (stating that “FIPA printed the
November 28, 1997, edited version of the specification” in the books)).
Patent Owner argues that because these various versions of FIPA 97
Version 1.0 have not been substantively introduced into evidence so that the
differences can be ascertained, Petitioner has not shown that Exhibits 1006–
1012 are the version that was publicly accessible on the FIPA website in
late 1997. See PO Resp. 52–53. We disagree. Dr. Finin testifies that
Exhibits 1006–1012 are the same documents that were released at the
conclusion of the FIPA meeting in Munich on October 10, 1997, and were
made publicly available at the time. Ex. 1049 ¶¶ 21, 34. Other evidence of
record corroborates Dr. Finin’s testimony and adequately explains the
multiple versions of the specification. See Pet. Reply 17. As set forth in
detail above, FIPA released a version of FIPA 97 Version 1.0 on October 10,
1997, and made it available on the FIPA website soon thereafter. See
Ex. 1049 ¶ 21; Ex. 1086, 1–2; Ex. 1095, 22; Ex. 1096, 93. It is clear from
the record that any earlier versions of the specification, such as the version
made available after the Reston meeting six months before FIPA 97
Version 1.0, were drafts made available for public comment. See Pet.
Reply 20–21; Ex. 1049 ¶¶ 16–18; Ex. 1024; Ex. 1061.
The record also shows that the version published in book form is a
later version, dated November 28, 1997, containing minor formatting and
IPR2019-00811
Patent 6,851,115 B1
51
clerical edits. See Ex. 1049 ¶ 25; Ex. 1055, 5. Petitioner does not rely on
the book version as a prior art reference or assert that it was publicly
accessible. See Pet. 23–24; Pet. Reply 21. The hardcopy book excerpts in
the record merely corroborate Dr. Finin’s testimony that FIPA’s activities
and release of the FIPA specification were known in the art. See Ex. 1049
¶ 24.
Patent Owner also points to the erroneous header on odd pages in
Exhibit 1006 as calling into question which version of the specification is in
Exhibits 1006–1012. PO Resp. 52. We credit Dr. Finin’s unrebutted
testimony that the header on some pages in Exhibit 1006 is a typographical
error. See Ex. 1049 ¶ 21. Patent Owner has not cited, nor do we see,
evidence in the record suggesting a FIPA 97 Version 2.4 was made publicly
available. See PO Resp. 52.
Patent Owner finds fault with Exhibits 1006–1012 because they were
obtained after the critical date and could not have been obtained from a
website prior to the critical date. See PO Resp. 61–62; PO Sur-reply 21. In
his declaration, Dr. Finin testifies that Exhibits 1006–1012 are available on
the FIPA website today (i.e., fipa.org), and the earliest Internet Archive
record of the relevant website pages shows the documents were posted in
2000. Ex. 1049 ¶ 21 (citing Ex. 1054). Patent Owner contends that in his
deposition Dr. Finin acknowledged that Exhibits 1006–1012 could have
been obtained from files saved in the Internet Archive between 2003 and
2004. PO Resp. 47–48 (citing Ex. 2013, 93:12–94:2, 96:7–20). In either
case, Patent Owner argues, the dates are after the critical date of the
’115 patent. Id. at 48; PO Sur-reply 22. We agree with Petitioner, however,
IPR2019-00811
Patent 6,851,115 B1
52
that the date Exhibits 1006–1012 were obtained is immaterial if they are
digital copies of documents that are shown to have been publicly accessible
before the critical date. See Pet. Reply 19 (citing In re Wyer, 655 F.2d at
226–27).
Finally, Patent Owner argues that the testimony of Dr. Finin and
Dr. Lieberman is conclusory, unreliable, and uncorroborated and therefore
insufficient to establish public accessibility. PO Resp. 57–60; PO Sur-
reply 20–23. “[C]orroboration is required of any witness whose testimony
alone is asserted to invalidate a patent, regardless of his or her level of
interest.” Nobel Biocare, 903 F.3d at 1377–78 (quoting Finnigan Corp. v.
ITC, 180 F.3d 1354, 1369 (Fed. Cir. 1999)). This corroboration requirement
applies when relying on witness testimony to establish public accessibility of
a prior art reference. See id. at 1377–81; Typeright Keyboard Corp. v.
Microsoft Corp., 374 F.3d 1151, 1158–60 (Fed. Cir. 2004). Corroborating
evidence may include documentary or testimonial evidence, and
circumstantial evidence may provide sufficient corroboration. Nobel
Biocare, 903 F.3d at 1378. Determining whether testimony has been
corroborated “involves an assessment of the totality of the circumstances
including an evaluation of all pertinent evidence.” Id.
We are not persuaded by Patent Owner’s argument regarding the
testimony of Dr. Finin and Dr. Lieberman. Dr. Finin’s testimony is based on
his personal knowledge as someone who was directly involved in FIPA’s
activities by submitting comments on draft specifications and attending
meetings, for example, and providing independent advice as a FIPA Fellow.
His testimony that FIPA approved FIPA 97 Version 1.0 at the Munich
IPR2019-00811
Patent 6,851,115 B1
53
meeting on October 10, 1997, and posted it to the FIPA website so that it
was publicly accessible is corroborated by the evidence in the record of the
facts and circumstances surrounding the development and release of the
specification as analyzed above. Dr. Lieberman’s testimony that as a
member of the interested public he accessed FIPA 97 Version 1.0 in late
1997 is corroborated by the same evidence.
As for Dr. Finin’s testimony, based on his personal knowledge, that
Exhibits 1006–1012 are the same as the version of FIPA 97 Version 1.0 that
was publicly accessible on the FIPA website in late 1997, we find it to be
credible and corroborated by other evidence of record. We are not
persuaded that Dr. Finin’s inability to recall many of the details of a
300-page specification more than twenty years after it was created
undermines his testimony. See PO Resp. 58–60. Furthermore, we find that
the evidence discussed above, including the identification of Exhibits 1006–
1012 on their cover pages or elsewhere as FIPA 97 Version 1.0 with a date
of October 10, 1997, corroborates Dr. Finin’s testimony that the version of
the specification in Exhibits 1006–1012 is the version that was publicly
accessible on the FIPA website in late 1997, and not some other version of
the FIPA specification.
For these reasons, we find that a preponderance of the evidence shows
that Exhibits 1006–1012 are the version of FIPA 97 Version 1.0 released on
October 10, 1997, and posted on the FIPA website in late 1997. As
discussed above, a preponderance of the evidence shows that the October 10,
1997 version of FIPA 97 Version 1.0 was publicly accessible in late 1997.
Accordingly, we find that Petitioner has met is burden to show that FIPA97,
IPR2019-00811
Patent 6,851,115 B1
54
submitted as Exhibits 1006–1012 and relied on by Petitioner in its
unpatentability challenges, was publicly accessible prior to the critical date
of January 5, 1999, and therefore is available as prior art to the ’115 patent
under 35 U.S.C. § 102(a).
E. Prior Art Status of Kiss
We next address the threshold issue of whether Kiss (Ex. 1005)
qualifies as prior art in this proceeding. As indicated above, Petitioner relies
on Kiss as a prior art reference in all of the asserted grounds of
unpatentability presented in the Petition. See Pet. 3–4.
Petitioner has the burden of persuasion to prove unpatentability by a
preponderance of the evidence. Dynamic Drinkware, LLC v. Nat’l
Graphics, Inc., 800 F.3d 1375, 1379 (Fed. Cir. 2015). Petitioner also has the
initial burden of production to show that a reference is prior art to the
challenged claims under a relevant section of 35 U.S.C. § 102. See id. Once
Petitioner has met that initial burden, the burden of production shifts to
Patent Owner to argue or produce evidence that the asserted reference is not
prior art to the claims. Id. at 1380. Once Patent Owner has met that burden
of production, the burden of production returns to Petitioner. Id.
The filing date of the ’115 patent is January 5, 1999. Ex. 1001,
code (22). As noted above, Kiss is a United States patent that arose from an
application filed July 21, 1999, claiming the benefit of priority to a
provisional application filed July 21, 1998. Ex. 1005, codes (22), (60).
Petitioner asserts that Kiss is prior art to the ’115 patent under
35 U.S.C. § 102(e) because Kiss is entitled to the benefit of priority to the
filing date of its provisional application (Ex. 1036, “Kiss Provisional
IPR2019-00811
Patent 6,851,115 B1
55
Application”). Pet. 16–20. “A reference patent is only entitled to claim the
benefit of the filing date of its provisional application if the disclosure of the
provisional application provides support for the claims in the reference
patent in compliance with § 112, ¶ 1.” Dynamic Drinkware, 800 F.3d at
1381.
In Ex parte Mann, the Board held that “under Dynamic Drinkware, a
non-provisional child can be entitled to the benefit of a provisional
application’s filing date if the provisional application provides sufficient
support for at least one claim in the child.” 2016 WL 7487271, at *6 (PTAB
Dec. 21, 2016) (emphasis added) (discussing whether Dynamic Drinkware
requires “support in the provisional . . . for all claims, any claim, or
something in between”). In addition, the Board held that a “subject matter
test” is also required—that is, “the [party claiming priority] also must show
that the subject matter relied upon in the non-provisional is sufficiently
supported in the provisional application [and that t]his subject matter test is
in addition to the comparison of claims required by Dynamic Drinkware.”
Id. at *5.
Recognizing these requirements, Petitioner asserts that “each element
of [claim 1 of Kiss] has written description support in the Kiss Provisional,”
providing detailed citations to the supporting disclosures from the Kiss
Provisional Application for each limitation of claim 1 of Kiss. Pet. 16–18.
Addressing the “subject matter test,” Petitioner asserts that “the teachings
that Petitioner relies upon were carried forward from the Kiss Provisional to
Kiss” and provides a detailed mapping of the relied-upon portions of Kiss to
IPR2019-00811
Patent 6,851,115 B1
56
the corresponding portions of the Kiss Provisional Application. Id. at 18–
20.
In the Institution Decision, we determined, based on the preliminary
record, that Petitioner has presented sufficient argument and evidence to
meet its initial burden of production on the issue of whether Kiss is entitled
to the benefit of the filing date of the Kiss Provisional Application so as to
shift the burden of production to Patent Owner to argue or produce evidence
that Kiss is not prior art to the challenged claims. Inst. Dec. 25.
In its Preliminary Response, Patent Owner asserted that Kiss is not
entitled to the priority date of its provisional application. Prelim. Resp. 67–
68. In the Institution Decision, we determined, based on the preliminary
record, the argument and evidence presented by Patent Owner was
insufficient to shift the burden of production back to Petitioner. Inst. Dec.
25–27. We noted, however, a final determination on this particular issue
will be made based on a full record developed during the course of trial. Id.
at 27.
During the trial, Patent Owner did not present any argument or
evidence on this issue, nor did it dispute Kiss qualifies as prior art in this
proceeding. See generally PO Resp.
Based on the complete record, we determine that Petitioner has
established by a preponderance of the evidence that Kiss is prior art to the
’115 patent under 35 U.S.C. § 102(e) because Kiss is entitled to the benefit
of priority to the filing date of its provisional application.
IPR2019-00811
Patent 6,851,115 B1
57
F. Relevant Principles of Law
To prevail in challenging Patent Owner’s claims, Petitioner must
demonstrate by a preponderance of the evidence that the claims are
unpatentable. 35 U.S.C. § 316(e); 37 C.F.R. § 42.1(d). “In an [inter partes
review], the petitioner has the burden from the onset to show with
particularity why the patent it challenges is unpatentable.” Harmonic Inc. v.
Avid Tech., Inc., 815 F.3d 1356, 1363 (Fed. Cir. 2016) (citing 35 U.S.C.
§ 312(a)(3) (requiring inter partes review petitions to identify “with
particularity . . . the evidence that supports the grounds for the challenge to
each claim”)). This burden never shifts to Patent Owner. See Dynamic
Drinkware, LLC v. Nat’l Graphics, Inc., 800 F.3d 1375, 1378 (Fed. Cir.
2015) (citing Tech. Licensing Corp. v. Videotek, Inc., 545 F.3d 1316, 1326–
27 (Fed. Cir. 2008)) (discussing the burden of proof in inter partes review).
A claim is unpatentable under 35 U.S.C. § 103(a) if the differences
between the claimed subject matter 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 the 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) where in evidence, so-called secondary
considerations. Graham v. John Deere Co., 383 U.S. 1, 17–18 (1966).
Petitioner cannot satisfy its burden of proving obviousness by employing
“mere conclusory statements.” In re Magnum Oil Tools Int’l, Ltd., 829 F.3d
IPR2019-00811
Patent 6,851,115 B1
58
1364, 1380 (Fed. Cir. 2016). We analyze Petitioner’s asserted grounds
based on obviousness with the principles identified above in mind.
G. Objective Indicia of Nonobviousness
Before determining whether a claim is obvious in light of the prior art,
we consider any relevant evidence of secondary considerations of non-
obviousness. See Graham, 383 U.S. at 17. Notwithstanding what the
teachings of the prior art would have suggested to one of ordinary skill in the
art at the time of the invention, the totality of the evidence submitted,
including objective evidence of non-obviousness, may lead to a conclusion
that the challenged claims would not have been obvious to one of ordinary
skill. In re Piasecki, 745 F.2d 1468, 1471–72 (Fed. Cir. 1984). Patent
Owner presents evidence of two such considerations: (1) industry praise (PO
Resp. 120–141) and (2) long-felt but unresolved need (id. at 141–145).
“In order to accord substantial weight to secondary considerations in
an obviousness analysis, the evidence of secondary considerations must have
a nexus to the claims, i.e., there must be a legally and factually sufficient
connection between the evidence and the patented invention.” Fox Factory,
Inc. v. SRAM, LLC, 944 F.3d 1366, 1373 (Fed. Cir. 2019) (internal
quotations omitted). For instance, commercial success is relevant if it flows
from the merits of the claimed invention. Sjolund v. Musland, 847 F.2d
1573, 1582 (Fed. Cir. 1988).
A nexus is presumed when “the patentee shows that the asserted
objective evidence is tied to a specific product and that product ‘embodies
the claimed features, and is coextensive with them.’” Id. (quoting Polaris
Indus., Inc. v. Arctic Cat, Inc., 882 F.3d 1056, 1072 (Fed. Cir. 2018)). If the
IPR2019-00811
Patent 6,851,115 B1
59
product is not coextensive with the claims at issue—for example, if the
patented invention is only a component of the product—the patentee is not
entitled to a presumption of nexus. See id. (citing Demaco Corp. v. F. Von
Langsdorff Licensing Ltd., 851 F.2d 1387, 1392 (Fed. Cir. 1988)). The
burden of showing nexus is on the patent owner. In re Huang, 100 F.3d
135, 139–40 (Fed. Cir. 1996).
Patent Owner contends that objective indicia of nonobviousness
confirm the nonobviousness of the ’115 patent. PO Resp. 119–145. Patent
Owner argues that industry praise for a software application known as
“Siri,” “which was first made available for download to iPhone users on
Apple, Inc.’s App Store on February 4, 2010” (id. at 121), “bears a nexus to
the claims of the ’156 Patent” (id. at 141).18 Patent Owner argues that
industry praise for the “personal context awareness and service delegation
‘technical components’ of the Siri application are made possible by the []
claim limitations of independent claim 1 of the ’115 Patent.” Id. at 136.
Patent Owner asserts that “[t]he ’115 Patent was based on SRI
International’s development of version 2 of the Open Agent Architecture
(‘OAA’).” Id. at 135 (citing Ex. 2032; Ex. 1001). For support, Patent
Owner relies on Exhibit 2054, an SRI web page, which states, “SRI
developed OAA® software, which is designed with intelligent ‘agents’ that
track human interactions and work processes to streamline electronic and
computer interchange.” Id. (quoting Ex. 2054). Patent Owner further
18 Although Patent Owner refers to the ’156 Patent here, we assume this is a
mistake and that Patent Owner actually meant to refer to the patent at issue
in this proceeding, the ’115 patent.
IPR2019-00811
Patent 6,851,115 B1
60
asserts that “[a]fter version 2 of OAA was developed, SRI International spun
out Siri, Inc. as an SRI venture, where the Siri application was created based
on an OAA groundwork.” Id. (quoting Ex. 2054 (“OAA laid the
groundwork for the DARPA-funded CALO project, from which Siri, the
first virtual personal assistant, was born. Siri was spun out into an SRI
venture that Apple acquired from SRI.”). Apple Inc. apparently acquired
Siri, Inc. in April 2010, approximately two months after the Siri
application’s initial release.19 Id. at 10, 121, 135.
Patent Owner asserts that “Petitioner proposed a combination of Kiss
with FIPA97, which relies on FIPA97 as teaching a common
communications and protocol language that is combined with Kiss’s
teachings of electronic agents and their general functions and operations.”
PO Resp. 134 (citing Paper 12, 45). Patent Owner argues a “nexus between
the evidence of industry praise of the Siri application [] and the Challenged
Claims can be shown if the industry praise has a nexus to the combination of
these (allegedly) prior art features, i.e., the teachings of electronic agents and
their general functions and operations in Kiss and the common
communications protocol and language of FIPA97.” Id. Patent Owner
argues that its “evidence of industry praise [] satisfies the nexus requirement,
as it shows praise for the Siri application’s integration of personal context
awareness into a service-delegating virtual personal assistant.” Id. at 136;
see also id. at 122–133. Patent Owner also argues that “industry journalists
19 We note for the record that the filing date for the ’115 patent is Jan. 5,
1999, approximately ten years prior to the Siri application’s initial release.
IPR2019-00811
Patent 6,851,115 B1
61
recognized that the Siri application filled an unmet need for a true virtual
personal assistant.” Id. at 141; see also id. at 141–145.
To establish a nexus between claim 1 of the ’115 patent and the Siri
application, Patent Owner provides the following chart. Id. at 136–138.
Patent Owner relies, in significant part, on the opinion of its declarant,
Dr. Medvidovic, to establish a nexus between the Siri application and the
claims of the ’115 patent. See Ex. 2032 ¶¶ 205–229, App. A, 1–32.
IPR2019-00811
Patent 6,851,115 B1
62
IPR2019-00811
Patent 6,851,115 B1
63
Patent Owner argues that “the Siri application’s integration of
personal context awareness is made possible because of the claimed inter-
agent language limitations.” PO Resp. 138 (citing Ex. 2032 ¶ 222). Patent
Owner explains that “[b]y having a layer of conversational protocol that has
parameter lists, the ’115 Patent describes how agents can then use the
parameters in the parameter lists to provide feedback or advice to agents
who directly, or indirectly through the facilitating agent, receive the service
request.” Resp. 138. For example, Patent Owner explains that a “requesting
agent can utilize such a parameter to provide context on a user’s preference
to the facilitator agent and/or the tasking agents. Id. at 139 (citing Ex. 2032
¶ 224). Patent Owner argues that “[t]his same functionality of being able to
include individual preferences, for example, was described by Siri, Inc. as
representing personal context awareness.” Id. at 140 (citing Ex. 2040, 2).
IPR2019-00811
Patent 6,851,115 B1
64
Patent Owner also argues that:
the Siri application’s service delegation corresponds to the
claimed dynamic interpretation limitations. Siri, Inc. described
the service delegation feature of the Siri application as including
“an assistant [that] can reason about what specific set of
resources or services would best be combined to help you
accomplish a particular task.” This is precisely what is claimed
in the dynamic interpretation limitations, e.g., “constructing a
goal satisfaction plan . . . [including] a suitable delegation of sub-
goal requests to best complete the requested service request-by
using reasoning.”
Id. (citing Ex. 2032 ¶ 221).
Petitioner argues that Patent Owner’s “secondary considerations
position rests primarily on an unsubstantiated, conclusory opinion that
certain aspects of the ’115 Patent are embodied in the Apple product ‘Siri.’”
Pet. Reply 82. Petitioner argues that as “Dr. Medvidovic conceded, the
evidence he relied on is ‘circumstantial,’ or a ‘suggestion.’” Id. (citing
Ex. 1129, 147:12–15). Petitioner argues that Dr. Medvidovic “never
reviewed the Siri source code, (Ex. 1129, 146:20–147:6), and indeed never
even asked to review that source code (Ex. 1129, 147:22–148:9).”
Pet. Reply 82–83. Petitioner argues,
[t]he failure to review the code is fatal, because [Dr. Medvidovic]
further testified that an “agent”—an element in every claim—is
a “piece of functionality that is capable of performing certain
tasks that is embodied in at least for a software engineer a clearly
identifiable body of code,” and that to distinguish an “agent”
from any other software he “would have to look inside the
code.”
Id. at 83 (quoting Ex. 1129, 9:14–11:8).
Petitioner faults Patent Owner for not “not seek[ing] that code through
a motion for additional discovery or other means, nor did [Petitioner] seek
IPR2019-00811
Patent 6,851,115 B1
65
any other objective evidence (such as specifications) to support its
assertions.” Pet. Reply 83. Petitioner also faults Patent Owner for not
seeking the testimony of Mr. Adam Cheyer, one of the named co-inventor’s
on the ’115 patent and a co-founder of Siri, Inc., on this topic, even though
Mr. Cheyer’s testimony was apparently obtained in a related proceeding. Id.
at 83–84 (citing Ex. 2034, 58:15–63:16; Ex. 1127); see also id. at 121.
During the Oral Hearing in this proceeding, counsel for Patent Owner was
asked about this potential source of evidence:
12 JUDGE TROCK: Counsel, this is Judge Trock
13 again.
14 Do you have any evidence from Mr. Cheyer
15 that this Siri version that you're discussing right
16 now incorporated the claim limitations of the '115
17 patent?
18 MS. ABDULLAH: Your Honor, we do not have
19 any direct evidence from Mr. Cheyer. And the reason
20 is because he was never deposed in this
21 proceeding and he is also not a consultant or
22 anybody that's working with IPA at the moment. I
23 know --
24 JUDGE TROCK: Wasn't he deposed -- sorry.
25 Wasn't he deposed in the Google proceeding?
26 MS. ABDULLAH: He was, your Honor. And that
1 was pursuant to a subpoena that we had to get
2 permission to serve from the Board because he
IPR2019-00811
Patent 6,851,115 B1
66
3 refused to simply cooperate with us voluntarily.
4 JUDGE TROCK: Was there a reason why you
5 didn't approach the Board in this case for a
6 subpoena?
7 MS. ABDULLAH: Your Honor, because we don't
8 believe that his testimony is as probative as the
9 expert testimony that we presented drawing the
10 direct connection as well as the actual industry
11 praise and long-felt need evidence that we --
12 JUDGE TROCK: But wouldn't he be a witness
13 with personal knowledge of whether or not that
14 original Siri version incorporated the claim
15 limitations of the '115 patent?
16 MS. ABDULLAH: Your Honor, respectfully, I
17 believe that that was not -- that would not be a
18 question of personal knowledge but rather of expert
19 testimony. And also, your Honor, I would note that
20 I don't believe that we would have the requisite
21 showing under the Garmin factors that he would be
22 able to testify to those aspects.
23 JUDGE TROCK: But he is the inventor, right?
24 You've told us that.
25 MS. ABDULLAH: Yes, your Honor, he is the
26 inventor.
1 TROCK: All right. So he might be a
IPR2019-00811
Patent 6,851,115 B1
67
2 person -- a witness with personal knowledge; isn't
3 that right?
4 MS. ABDULLAH: Absolutely. And were this a
5 district court case, you know, I'm sure we would
6 have pursued his deposition. I'm sure Microsoft
7 would have as well. But given the high burden for
8 us to obtain additional discovery, your Honor, we
9 determined that the probative value of his testimony
10 was not such that it would meet the Garmin factors
11 here.
12 JUDGE TROCK: But you did pursue that in the
13 Google case though?
14 MS. ABDULLAH: Your Honor, the Google case
15 was a different issue. The question there had to do
16 with authorship of a prior art reference that
17 Mr. Cheyer was an author of. And it had to do with
18 whether that could be considered the work of another
19 versus the inventor. So inventorship was front and
20 center in the Google proceedings where it is not
21 here. Here this is a 103 analysis for which, you
22 know, obviously I can't pretend to say what the
23 Board would have done. But I think the evidence
24 that we have, that we have been able to present
25 demonstrates that link better than Mr. Cheyer's
26 testimony would be able to.
IPR2019-00811
Patent 6,851,115 B1
68
Tr. 114:12–116:26.
Petitioner also points to evidence that potentially undermines a nexus
between the challenged claims of the ’115 patent and the Siri application.
For example, Petitioner argues that “Adam Cheyer testified that ‘Siri’ did
not implement the agent registry functionality described in the ’115 Patent.”
Pet. Reply 84 (citing Ex. 2034, 46:12-47:5). Likewise, Petitioner argues,
“an email authored by Adam Cheyer indicates that as of August 2005
OAAv2 had only been adapted to run on ‘Windows CE’ mobile devices—
there is no mention of iOS, Apple, or any other mobile platforms.” Id.
(citing Ex. 1137).
Petitioner also argues that Patent Owner’s “evidence credits the
CALO project with being the progenitor of the ideas that led to Siri—not
any version of OAA.” Pet. Reply 84 (citing Ex. 2042, 1 (“Siri traces its
origins to a military-funded artificial-intelligence project called CALO”);
Ex. 2044, 1; Ex. 2047, 2; Ex. 2054, 1; Ex. 2001, 2). “Nor is there any
evidence in the record,” Petitioner argues, “that OAA was ultimately utilized
in SRI’s CALO project other than that the original OAA (i.e., the version
first created in 1994 prior to the ’115 Patent work) laid ‘groundwork’ for
what would ultimately become CALO.” Id. (citing Ex. 2054). Petitioner
argues that “substantial evidence shows that a later agent system named
“SPARK” (SRI Procedural Agent Realization Kit) was used as the basis of
CALO,” but that Patent Owner “ignores this intervening SPARK system and
makes no attempt to show it practices the claims, nor does it meaningfully
address the CALO project.” Id. at 85 (citing Ex. 1136 §7).
IPR2019-00811
Patent 6,851,115 B1
69
Petitioner argues that Patent Owner “failed to provide evidence ‘that
the industry praised a claimed invention or a product that embodies the
patent claims.’” Pet. Reply 85 (quoting Apple Inc. v. Samsung Elecs. Co.,
839 F.3d 1034, 1053 (Fed. Cir. 2016) (emphasis omitted). In particular,
Petitioner argues, Patent Owner “failed to ‘establish that the evidence relied
upon traces its basis to a novel element in the claim and not to something in
the prior art.’” Id. (quoting BioMarin Pharm. Inc. v. Genzyme Therapeutic
Prods. LP, IPR2013-00537, Paper 79, 22 (PTAB Feb. 23, 2015) (emphasis
omitted).
“Furthermore,” Petitioner argues, Patent Owner “has provided no
showing that the Siri App ‘embodies the claimed features and is co-
extensive with them,’ and accordingly can gain no presumption from any
Siri success.” Pet. Reply 85–86 (quoting Polaris Indus., Inc. v. Arctic Cat,
Inc., 882 F.3d 1056, 1072 (Fed. Cir. 2018) (emphasis omitted). “Even
assuming portions of Siri embodied the claims,” Petitioner argues, “the
record is undisputed that Siri provides significant functionality well beyond
anything covered by the claims—voice recognition, for example.” Id. at 86.
Petitioner further argues “the table provided by [Patent Owner]
merely lists unrelated claim language and ‘examples’ of ‘personal context
awareness’ with no further analysis.” Pet. Reply 87. “For example,”
Petitioner argues, Patent Owner “provides no explanation about how the
alleged ‘Personal Context Awareness’ feature in Siri practices ‘a layer of
conversational protocol.’” Id. Moreover, Petitioner argues, “despite
claiming that ‘the Siri application’s integration of personal context
awareness is made possible because of the claimed inter-agent language
IPR2019-00811
Patent 6,851,115 B1
70
limitations,’ [Patent Owner] provides no evidence. [Patent Owner] cites
only to a single paragraph of Dr. Medvidovic’s declaration, which is itself
bare ipse dixit and further undermined by his failure to examine the source
code or any other descriptive technical materials for Siri.” Id. at 88 (citing
Ex. 2032 ¶ 222).
Petitioner argues that Patent Owner’s evidence that Siri embodies the
ICL limitations of claim 1 “is no better.” Pet. Reply 88 (citing PO Resp.
140). Petitioner argues Patent Owner:
does not discuss the actual limitations of claim 1 nor explain how
giving different answers depending on individual preferences
embodies the claims. For example, [Patent Owner] does not even
attempt to articulate how Siri has a conversational layer, or event
types, or how those event types are refined by parameters. Nor
does [Patent Owner] address how Siri embodies the claimed
agent registry limitation, particularly when the testimony of the
inventor Mr. Cheyer indicates that it does not.
Id. (citing Ex. 2034, 46:12–47:5).
Petitioner argues that Dr. Medvidovic’s testimony should be given no
weight. Pet. Reply 89. Petitioner argues that:
Dr. Medvidovic admitted that he would be unable to determine
whether a piece of software was an “agent” without examining
the source code for that software, and further admitted that he
had not reviewed the Siri source code nor ever asked to review
such source code. Ex. 1129, 11:9–16. Yet, Dr. Medvidovic
opines at length that Siri must embody the claims of the ’115
Patent because, for example, “personal context awareness and
service delegations are made possible by the functionality
claimed in the independent claims of the ’115 Patent”—an
opinion without any evidentiary support. Ex. 2032 ¶ 220. Indeed,
by Dr. Medvidovic’s own admission, he lacks the necessary
personal knowledge to opine on the architecture of Siri.
Id.
IPR2019-00811
Patent 6,851,115 B1
71
In its Sur-reply, Patent Owner argues that there is “no prohibition
against the use of circumstantial evidence in IPR proceedings.”
PO Sur-reply 66. Patent Owner argues, “Dr. Medvidovic provided
unrebutted testimony, including a lengthy appendix, supported by this
evidence, which must be credited. Id. at 68 (citing Ex. 2032, §§ 205-229,
App. A). Patent Owner also points out that “Dr. Medvidovic testified that
looking at the code is not the exclusive means of identifying an agent – other
available information can be utilized.” Id. at 69; see Ex. 1129, 10:14–11:16.
Moreover, Patent Owner argues, “under the Garmin factors, Patent Owner
was unable to seek Apple’s highly confidential source code in connection
with these proceedings.” Id. at 70 (citing Garmin Int’l, Inc. v. Cuozzo
Speed Techs., LLC, IPR2012-00001, Paper 26 at 6–7 (PTAB Mar. 5, 2013).
With respect to Mr. Cheyer’s testimony, Patent Owner explains that “Patent
Owner was only able to depose Mr. Cheyer after the Board granted Patent
Owner’s motion under 37 C.F.R. § 42.52(a) for leave to serve a subpoena to
compel Mr. Cheyer to testify” in a related proceeding. Id. at 71.
We agree with Petitioner that Patent Owner does not to provide
persuasive evidence that the Siri application is reasonably commensurate
with the scope of the challenged claims. Patent Owner’s cited evidence
attempting to tie the Siri application to the limitations of claim 1 relies
substantially on the testimony of Dr. Medvidovic (Ex. 2032) and a
web.archive.org Internet page (Ex. 2040) Patent Owner identifies as “Siri,
Inc., Technology – “About Siri – Your Virtual Personal Assistant.” See
PO Resp. 136–140 (citing Ex. 2032 ¶¶ 216–222, 224; Ex. 2040).
IPR2019-00811
Patent 6,851,115 B1
72
We note at the outset that Dr. Medvidovic’s opinion does not rely on
an examination of the Siri application’s source code. During his deposition,
Dr. Medvidovic testified as follows:
[Q] Let me ask you, have you ever reviewed the Siri source code?
[A] I have not.
[Q] So you don't know the details of how Siri is implemented?
[A] Not beyond what I'm stating in the declaration as my sources of
information that I'm drawing on.
Ex. 1129, 146:10–17. This is significant because Dr. Medvidovic testified at
his deposition that an “agent”—an element in every claim—is a “piece of
functionality that is capable of performing certain tasks that is embodied in
at least for a software engineer a clearly identifiable body of code,” and that
“if there is no accompanying specification language, then we would have to
look inside the code [to determine] what kinds of facilities the agent
provides.” See Ex. 1129, 9:13–11:8). As far as we can determine, neither
party has provided “accompanying specification language” for the Siri
application source code. Instead, Dr. Medvidovic testified, the evidence he
relied upon to form his opinion is “circumstantial,” or a “suggestion.” See
Ex. 1129, 147:7–148:9.
In Section XI.B of his declaration, “Nexus of Industry Praise to the
Independent Claims of the ’115 Patent,” Dr. Medvidovic cites to
Exhibit 2040 as support for his understanding of how the Siri application
works. See e.g., Ex. 2032 ¶¶ 215–216, 218, 226, 229.
Exhibit 2040 describes “Siri” as:
the first mainstream consumer application of a Virtual Personal
Assistant. Siri is an intelligent software agent designed to have
IPR2019-00811
Patent 6,851,115 B1
73
a back-and-forth conversational interaction with you as it helps
you get tasks done. The three main technical components behind
Siri’s differentiation correspond to the essential qualities of an
assistant: a conversational interface, personal context awareness
and service delegation.
Ex. 2040, 1.
In his declaration, Dr. Medvidovic discusses the “personal context
awareness” and “service delegation” aspects of Siri in some detail (see, e.g.,
Ex. 2032 ¶¶ 216–220, 222–229). Dr. Medvidovic, however, only mentions
Siri’s “conversational interface” in passing, even though Exhibit 2040
describes Siri as an intelligent software agent “designed to have a back-and-
forth conversational interaction with you as it helps you get tasks done,” and
lists Siri’s “conversational interface” as the first of Siri’s “three main
technical components.” See Ex. 2032 ¶ 215; Ex. 2040, 1.
In his declaration, Dr. Medvidovic attempts to link the Siri application
and the limitations of the independent claims of the ’115 patent. See
Ex. 2032 ¶¶ 205–229, App. A, 1–32. Dr. Medvidovic does this by first
separating the limitations of independent claims 1 and 29 into two
categories, an “Inter-agent Communication Layer” (ICL) and “Service
Request Planning and Execution.” See Ex. 2032, 70–71. Dr. Medvidovic
then states that “[i]t is my opinion that the Siri application’s integration of
personal context awareness is made possible because of the claimed ICL.”
Id. ¶ 222. Dr. Medvidovic explains that “within the conversational protocol
layer of the ICL, there are further claimed “parameter lists” that may
“further refine the one or more events.” Id. ¶ 223. Dr. Medvidovic
concludes that “[t]he claimed parameter lists are what allows the Siri
application to provide ‘different answers’ depending on the context, as they
IPR2019-00811
Patent 6,851,115 B1
74
are able to refine the meaning of a service request.” Id. ¶ 226.
Dr. Medvidovic, however, provides no citation or further support for this
conclusion nor does he point to any evidence to show that the Siri
application utilizes the claimed parameter lists. See id.
In his declaration, Dr. Medvidovic also states that “[i]t is further my
opinion that the Siri application’s integration of service delegation is made
possible because of the claimed service request planning and execution
claim limitations.” Id. ¶ 227. Dr. Medvidovic attempts to support this
opinion by comparing the language of claim 1 relating to a goal satisfaction
plan that includes “a suitable delegation of sub-goal requests to best
complete the requested service request by using reasoning” to language used
in Exhibit 2040 to describe the service delegation feature of the Siri
application, “Service Delegation. An assistant can reason about what
specific sets of resources of services would best be combined to help you
accomplish a particular task.” See id. ¶ 229; Ex. 2040, 2.
In our view, such a comparison is speculative at best to demonstrate
that the Siri application utilizes claim 1’s “service request planning and
execution limitations,” as Dr. Medvidovic opines. The limitations of claim 1
that Dr. Medvidovic puts in this category include:
A computer-implemented method for communication and
cooperative task completion among a plurality of distributed
electronic agents, comprising the acts of:
registering a description of each active client agent’s functional
capabilities as corresponding registered functional capabilities,
receiving a request for service as a base goal in the inter-agent
language, in the form of an arbitrary complex goal expression;
IPR2019-00811
Patent 6,851,115 B1
75
dynamically interpreting the arbitrarily complex goal expression,
said act of interpreting further comprising:
generating one or more sub-goals expressed in the inter-agent
language;
constructing a goal satisfaction plan wherein the goal satisfaction
plan includes:
a suitable delegation of sub-goal requests to best complete the
requested service request-by using reasoning that includes one or
more of domain-independent coordination strategies, domain-
specific reasoning, and application-specific reasoning
comprising rules and learning algorithms; and
dispatching each of the sub-goals to a selected client agent for
performance, based on a match between the sub-goal being
dispatched and the registered functional capabilities of the
selected client agent.
Ex. 2032, 70. Yet, Dr. Medvidovic does not persuasively show that the Siri
application “register[s] a description of each active client agent’s functional
capabilities,” “receiv[es] a request for service . . . in the form of an arbitrary
complex goal expression,” “generat[es] one or more sub-goals expressed in
the inter-agent language,” “construct[s] a goal satisfaction plan,” or
“dispatch[es] each of the sub-goals to a selected client agent . . . based on a
match between the sub-goal . . . and the registered functional capabilities of
the selected client agent.” See id. ¶¶ 227–229.
In fact, as Petitioner points out, there is evidence of record that the
Siri application did not utilize the claimed agent registry functionality as
Dr. Medvidovic opines. According to Petitioner, Mr. Cheyer, the named co-
inventor on the ’115 patent and the co-founder of Siri, Inc., “testified that
‘Siri’ did not implement the agent registry functionality described in the
IPR2019-00811
Patent 6,851,115 B1
76
’115 Patent.” Pet. Reply 84 (citing Ex. 2034, 46:12-47:5). At a deposition
in a related proceeding, Mr. Cheyer testified as follows:
[Q] So you say "it's likely." Are there features in OAA that you think
have yet to be implemented in systems today?
[A] Yes.
[Q] Could you describe those?
[A] Every system is different, but one feature of the original OAA 1
demo was that independent agents -- today we would call them
more like web services -- could connect dynamically to the -- to
the platform, register new capabilities with that platform, and
without knowing about other services on the network, explicitly,
the user could make a single request that would be -- that would
involve a collaboration, and sometimes competition, among those
network ecosystem of services. And so one user request would be
delegated to many services which were independently developed,
and none of those services knew precisely about each other. And
that when I look at Siri and Google Assistant and Alexa and
Bixby, that is not a capability that I have seen in any of the
services today.
Ex. 2034, 45:25–46:2, 46:12–47:5.
Petitioner also points out, and we agree, that there is evidence of
record that credits the CALO project with being the progenitor of the ideas
that led to Siri. See, e.g., Ex. 2042, 1 (“Siri traces its origins to a military-
funded artificial-intelligence project called CALO”); see also Ex. 2044, 1;
Ex. 2047, 2; Ex. 2054, 1; Ex. 2001, 2). Opus Research explains that “Siri is
IPR2019-00811
Patent 6,851,115 B1
77
set apart because it applies the depth of knowledge its founders and software
specialists have built at SRI and elsewhere in creating a “cognitive assistant
that learns and organizes” (CALO). Ex. 2044, 1. Petitioner points out that
there is also evidence of record that indicates a later agent system named
“SPARK” (SRI Procedural Agent Realization Kit) was used as the basis of
CALO. See Ex. 1136 § 7. The SPARK-based Personal Assistant is
described as follows:
SPARK provides general-purpose agent technology for a range
of domains that require reactive task execution. To date, the
driving application for SPARK has been the development of an
intelligent personal assistant for a high-level knowledge worker.
This assistant, called CALO, will be able to perform routine tasks
on behalf of its user (e.g., arrange meetings, complete online
forms, file email), as well as undertake open-ended processes
(e.g., purchasing a computer online), and anticipate future needs
of its user.
Ex. 1136 § 7. This evidence, however, does not appear to have been
considered or addressed by Dr. Medvidovic. See Ex. 2032 ¶¶ 205–229.
We are mindful that objective evidence of nonobviousness need only
be “reasonably commensurate with the scope of the claims.” Rambus Inc. v.
Rea, 731 F.3d 1248, 1257 (Fed. Cir. 2013). We also bear in mind that
“[w]hen the thing that is commercially successful is not coextensive with the
patented invention—for example, if the patented invention is only a
component of a commercially successful machine or process—the patentee
must show prima facie a legally sufficient relationship between that which is
patented and that which is sold.” Demaco Corp., 851 F.2d at 1392.
Here, Patent Owner relies on industry praise for the Siri application
and evidence of an unmet need to establish a nexus between the Siri
IPR2019-00811
Patent 6,851,115 B1
78
application and the claimed invention. However, we are concerned that
Patent Owner and its declarant, Dr. Medvidovic, pay little attention to one of
the “three main technical components behind Siri’s differentiation,” the
“conversational interface,” when evaluating the basis for the proffered
“industry praise.” See Ex. 2040, 1.
Petitioner argues, and we agree, that “[e]ven assuming portions of Siri
embodied the claims, the record is undisputed that Siri provides significant
functionality well beyond anything covered by the claims—voice
recognition, for example.” Pet. Reply 86. The impact of Siri’s
“conversational interface” on Siri’s reviewers is notable. For example, MIT
Technology Review explains that with Siri, “[u]sers can type or speak
commands in casual sentences, and the software deciphers their intent from
the context.” Ex. 2042, 1. Opus Research reports that “I’ve had [Siri] for a
couple of days and here are my initial reactions. My overall experience has
been quite positive. The quality of voice recognition (powered by the same
‘engine’ that supports Dragon Dictation and Dragon Search on the iPhone) is
quite good.” Ex. 2044, 1. Fast Company.Com reports, “Siri, which
launches today as a free iPhone app, is a virtual personal assistant that
amazingly resembles . . . an actual personal assistant. It understands plain
English commands, which can be spoken or typed (e.g. ‘Get me tickets to
that Matt Damon movie’).” Ex. 2045, 2.
Further, The Chronicle of Higher Education explains that Siri “takes a
multi-step search process and turns it into one, simple, voice-driven
request.” Ex. 2046, 2. USA Today reported that “[t]here were occasional
mistakes in speech recognition, but the overall accuracy [of Siri] was
IPR2019-00811
Patent 6,851,115 B1
79
impressive, at least when I spoke slowly and clearly in a quiet environment.”
Ex. 2047, 2. Spatial Computing Strategy & Research reports that, with Siri,
“You ask it to do stuff like ‘find me a pizza place near me’ or ‘tell me the
weather in Chicago this weekend,’ [w]ith your voice or by typing
commands.” Ex. 2048, 2. Search Engine Land explains that Siri is
“intended to enable you to do more with your voice and your phone in fewer
clicks or moves.” “[O]ne can speak to Siri in a more natural way. Rather
than speaking like a robot and saying ‘Open Table’ to minimize error, you
can say something more elaborate and ‘conversational’ such as, ‘I’d like a
reservation tonight around 7:30 at Le Cheval.’” Ex. 2050.
Each of these reported examples indicate that the reviews Patent
Owner relies upon to demonstrate “industry praise” for the Siri application
considered Siri’s conversational interface to be a significant part of its
functionality. Patent Owner and Dr. Medvidovic, however, give Siri’s
“conversational interface” little, if any, consideration when evaluating the
reasons for the “industry praise” used to try and establish a nexus between
the Siri application and the claimed invention. See, e.g., Ex. 2032 ¶ 215.
Even if the Siri application incorporated some of the functionality of the
claimed invention, the evidence of record relied on by Patent Owner shows
that the Siri application, and its attendant “industry praise,” is not reasonably
commensurate with the scope of the claimed invention. Patent Owner does
not argue that the Siri application’s “conversational interface” is claimed by
the ’115 patent.
For these reasons and based on the complete record, we determine that
Patent Owner’s proffered evidence of industry praise for the Siri application
IPR2019-00811
Patent 6,851,115 B1
80
and evidence of a long felt unresolved need does not establish a sufficient
nexus between the Siri application and the challenged claims of the ’115
patent. In the absence of an established nexus with the claimed invention,
secondary consideration factors are not entitled to much, if any, weight and
generally have no bearing on the legal issue of obviousness. See In re
Vamco Mach. & Tool, Inc., 752 F.2d 1564, 1577 (Fed. Cir. 1985).
H. Obviousness over Kiss and FIPA97
In this asserted ground of obviousness, Petitioner contends that claims
1, 26–29, 33, 45–47, and 86–89 of the ’115 patent are unpatentable under
35 U.S.C. § 103(a) as obvious over the combination of Kiss and FIPA97.
Pet. 32–68.
1. Proposed Combination of Kiss and FIPA97; Reason to Combine
In its proposed combination of Kiss with FIPA97, Petitioner relies on
Kiss as teaching electronic agents and their general functions and operations,
including agent collaboration, agent registry, and inter-agent messaging.
Pet. 28. Petitioner acknowledges, however, Kiss does not disclose an inter-
agent communication language and relies on FIPA97 to teach a common
communications protocol and language (FIPA ACL), which is combined
with the teachings of Kiss to provide a communication language for inter-
agent communication between the agents of Kiss. Id. Petitioner explains
that “[FIPA97’s] techniques are used to implement the functionality
described in Kiss, including facilitating agent collaboration, agent registry,
and inter-agent messaging.” Id. (citing Ex. 1005, 3:32–36, 3:67–4:1, 6:67–
7:1). Petitioner asserts that “[w]here the two systems disclose analogous
IPR2019-00811
Patent 6,851,115 B1
81
functionality, such as facilitating cooperation and agent registry, their
techniques are combined, as a Skilled Artisan would understand that to be an
efficient and common sense way to implement the combined system in order
to obtain the benefits of both, and therefore be motivated to do so.” Id.
(citing Ex. 1003 ¶¶ 217–219).
Petitioner argues the proposed combination of Kiss and FIPA97
would have been obvious for several reasons. Petitioner argues Kiss,
FIPA97, and the ’115 patent are analogous art, because “each is directed to
the same field of endeavor, i.e., distributed computing environments
generally and architectures for communication and cooperation among
distributed electronic agents specifically.” Pet. 29 (citing, e.g., Ex. 1001,
1:25–29; Ex. 1005, Abstract; Ex. 1007, Foreword, 1; Ex. 1003 ¶ 220).
Petitioner also argues, “FIPA97 and Kiss are reasonably pertinent to the
problem addressed by the ’115 patent—e.g., the need for ‘flexible, fault-
tolerant, distributed problem solving’ using ‘agent-based technologies.’” Id.
(citing Ex. 1001, 4:34–40; Ex. 1005, 2:33–40; Ex. 1007, 1; Ex. 1003 ¶ 221).
Petitioner argues,
[i]t would have been obvious to make this combination because
to do so would have been the arrangement of old elements (i.e.,
the functionality of FIPA97, and the system of Kiss) with each
performing the same function it had been known to perform
(communication between distributed agents (FIPA97v1));
cooperative task competition and problem solving (Kiss)) and
yielding no more than what one would expect from such an
arrangement (a system of distributed agents, able to
communicate to conduct cooperative task completion and
problem solving).
Id. (citing Ex. 1020, 2; Ex. 1046, 50; Ex. 1003 ¶ 222).
IPR2019-00811
Patent 6,851,115 B1
82
Petitioner also argues the combination
would have been obvious because a Skilled Artisan would have
known that agents, such as in Kiss, must use a common
communications protocol, i.e., an inter-agent communication
language, and related management functionality, and therefore
would have been motivated to consider and employ one of the
available inter-agent communications languages and its related
functionality, and because doing so would have been a
convenient and efficient way to facilitate effective
communication between agents.
Id. at 29–30 (citing Ex. 1003 ¶ 223; Ex. 1020, 2; Ex. 1046, 50).
Petitioner further argues,
[a] Skilled Artisan would also have been motivated based on
earlier descriptions of KQML’s similar use of an inter-agent
communication language, their disclosure that ‘[t]here are
several levels at which agent-based systems must agree, at least
in their interfaces, in order to successfully interoperate:
Transport … Language … Policy … Architecture,’ and their
disclosure of the necessity of a robust inter-agent communication
language in a system like Kiss.
Id. at 30 (citing Ex. 1016, Abstract, 4, 29, 30; Ex. 1006, 7; Ex. 1007, 41, 40–
42; Ex. 1005, Abstract; Ex. 1003 ¶¶ 224–226).
Petitioner argues it “would also have been obvious to try this
combination because the necessity of a common inter-agent communication
language was a known design need in a system such as Kiss, and there were
only a small number of identified, predictable solutions to that need, such as
FIPA ACL.” Id. (citing Ex. 1020, 2; Ex. 1046, 50; Ex. 1035, 1; Ex. 1003
¶ 227).
Petitioner argues, “[a] Skilled Artisan would have been further
motivated to make the combination because of the known advantages of
FIPA97, including that it specifies, ‘key agents necessary for the
IPR2019-00811
Patent 6,851,115 B1
83
management of an agent system, the ontology necessary for the interaction
between systems, and it defines also the transport level of the protocols.’”
Id. at 30–31 (citing Ex. 1035, 1; Ex. 1003 ¶ 229).
Further, Petitioner argues,
a Skilled Artisan would have been motivated to make the
combination because FIPA97 represented a substantial attempt
to formulate an industry standard designed to encourage
interoperability and uniformity amongst agent systems. A
Skilled Artisan would have understood that agent systems that
complied with such a standard would be more likely to be
interoperable with other systems and would likely have more
resources available.
Id. at 31 (citing Ex. 1007, vii; Ex. 1047; Ex. 1003 ¶ 230).
Petitioner argues, “[a] Skilled Artisan would have been further
motivated to include the administrative functionality and exemplary
practices described in FIPA97 with Kiss because the description of such
functionality and practices in FIPA97 suggests that their use with the FIPA
ACL would be advantageous and workable.” Id. Petitioner also argues, “[a]
Skilled Artisan would also have been motivated to employ such additional
functionality and practices in the combined system in order to obtain the full
benefit of using FIPA ACL.” Id. (citing Ex. 1003 ¶ 231).
Finally, Petitioner argues, “a Skilled Artisan could have made the
combination cited above without undue experimentation including because
by January 1999 the components parts of this combination were well-known,
conventional technology and because others had combined FIPA97v1
technology with agent systems.” Id. at 31–32 (citing Ex. 1007; Ex. 1016;
Ex. 1030; Ex. 1003 ¶ 228).
IPR2019-00811
Patent 6,851,115 B1
84
Patent Owner asserts that “Petitioner utterly defaults in offering any
reason to combine Kiss and FIPA97.” PO Resp. 116. Patent Owner argues
that Petitioner’s motivation to combine Kiss and FIPA97 suffers from
“hindsight bias” and that Petitioner “offers no reason or explanation that a
person of skill in the art looking at Kiss would be motivated to find a
particular language with the characteristics of FIPA97’s ACL.” Id. Patent
Owner also argues that Petitioner does not “explain what ‘efficiencies’ a
person of ordinary skill in the art would be looking for” and “fails to identify
how the combination could be achieved.” Id. at 116–117.20
In the Reply, Petitioner argues that Patent Owner “ignores the several
pages of analysis supported by expert testimony (including citations to
evidence in the prior art) articulating the rational underpinning of the
combination” of Kiss and FIPA97. Pet. Reply 30 (citing Pet. 28–32).
Petitioner argues that “while there was no requirement to show that a Skilled
Artisan would search for ‘a particular language with the characteristics of
FIPA97’s ACL,’” (PO Resp. 116), the Petition does explain that “agents,
such as in Kiss, must use a common communications protocol,” so a Skilled
Artisan “would have been motivated to ‘employ one of the available inter-
agent communications languages . . . because doing so would have been a
convenient and efficient way to facilitate effective communication between
agents.’” Pet. Reply 30–31 (citing Pet. 29–30; Ex. 1003 ¶ 223).
20 We need not address the parties’ dispute over the motivation to combine
Kiss’s meta-agent and FIPA97 directory facilitator (see, e.g., PO Resp. 79–
85; Pet. Reply 40–46) because whether the combination of Kiss and FIPA97
teaches the claimed “facilitator agent” is not at issue for this Final Written
Decision.
IPR2019-00811
Patent 6,851,115 B1
85
Petitioner also argues that Patent Owner’s criticism that “Petitioner
fails to identify how the combination could be achieved,” (PO Resp. 117), is
“legally irrelevant” (citing In re Keller, 642 F.2d 413, 425 (CCPA 1981),
and “also ignores the detailed description in the Petition (supported by
Dr. Lieberman) that demonstrates how the combination would work.” Pet.
Reply 31 (citing Pet. 28–32; Ex. 1003 ¶¶ 217–219; Ex. 1005, Figs. 4–6, 21
(annotated)). Petitioner points out that:
[t]he Petition explains, for example, that FIPA97 provides a
common communications protocol and language (FIPA ACL)
between the agents of Kiss, and also adds its administrative
functionality and exemplary practices to the Kiss system, as
described with specificity below. These techniques are used to
implement the functionality described in Kiss, including
facilitating agent collaboration, agent registry, and inter-agent
messaging, (Ex. 1005, 3:32–36, 3:67-4:1, 6:67–7:1), as well
adding functionality that is disclosed in FIPA97.
Pet. Reply 34.
In its Sur-reply, Patent Owner argues that Petitioner’s evidence of a
motivation to combine are simply “[p]ages of boilerplate language” or
exhibit “ex post reasoning colored by hindsight.” PO Sur-reply 58, 62.
Patent Owner argues that the FIPA97 and Kiss protocols are “different” and
“mutually exclusive” and that “Petitioner offers no argument or evidence
relating to how these completely different components would be
reconciled.” Id. at 59–61. Patent Owner argues that FIPA97 “discouraged
interoperability between different and incompatible systems.” Id. at 61.
Patent Owner also argues that “[t]he Board must disregard” portions
of Petitioner’s Reply, which discuss “a series of four figures from Dr.
Lieberman’s declaration,” “because it constitutes new argument.” Id. at 63.
IPR2019-00811
Patent 6,851,115 B1
86
Patent Owner argues that “the Petition must contain a ‘full statement of the
reasons for the relief requested’” and “cannot simply incorporate by
reference its expert’s declaration.” Id. (citing 37 C.F.R § 42.6(a)(3)).
In identifying a reason that would have prompted a person of ordinary
skill in the relevant field to combine the prior art teachings, the Petitioner
“must show some articulated reasoning with some rational underpinning to
support the legal conclusion of obviousness.” KSR, 550 U.S. at 418. A
reason to combine teachings from the prior art “may be found in explicit or
implicit teachings within the references themselves, from the ordinary
knowledge of those skilled in the art, or from the nature of the problem to be
solved.” WMS Gaming Inc. v. Int’l Game Tech., 184 F.3d 1339, 1355 (Fed.
Cir. 1999) (citing In re Rouffet, 149 F.3d 1350, 1357 (Fed. Cir. 1998)).
Here, Petitioner explains that:
FIPA97 provides a common communications protocol and
language (FIPA ACL) between the agents of Kiss, and also adds
its administrative functionality and exemplary practices to the
Kiss system, as described with specificity below. These
techniques are used to implement the functionality described in
Kiss, including facilitating agent collaboration, agent registry,
and inter-agent messaging, Ex. 1005, 3:32–36, 3:67–4:1, 6:67–
7:1, as well adding functionality that is disclosed in FIPA97.
Where the two systems disclose analogous functionality, such as
facilitating cooperation and agent registry, their techniques are
combined, as a Skilled Artisan would understand that to be an
efficient and common sense way to implement the combined
system in order to obtain the benefits of both, and therefore be
motivated to do so.
Pet. 28 (citing Ex. 1003 ¶¶ 217–219).
Petitioner provides persuasive evidence that Kiss and FIPA97 are
analogous art because “each is directed to the same field of endeavor, i.e.,
IPR2019-00811
Patent 6,851,115 B1
87
distributed computing environments generally and architectures for
communication and cooperation among distributed electronic agents
specifically.” Pet. 29 (citing Ex. 1001, 1:25–29; Ex. 1005, Abstract;
Ex. 1007, Foreword, 1; Ex. 1003 ¶ 220). Petitioner also provides persuasive
evidence that FIPA97 and Kiss “are reasonably pertinent to the problem
addressed by the 115 Patent—e.g., the need for ‘flexible, fault-tolerant,
distributed problem solving’ using ‘agent-based technologies.’” Pet. 29
(citing Ex. 1001, 4:34–40; Ex. 1005, 2:33–40; Ex. 1007, 1; Ex. 1003 ¶ 221).
Petitioner argues persuasively that
[i]t would have been obvious to make this combination because
to do so would have been the arrangement of old elements (i.e.,
the functionality of FIPA97, and the system of Kiss) with each
performing the same function it had been known to perform
(communication between distributed agents (FIPA97v1));
cooperative task competition and problem solving (Kiss)) and
yielding no more than what one would expect from such an
arrangement (a system of distributed agents, able to
communicate to conduct cooperative task completion and
problem solving).
Pet. 29 (citing Ex. 1020, 2; Ex. 1046, 50; Ex. 1003 ¶ 222).
Petitioner also argues, and we agree, that it
would have been obvious because a Skilled Artisan would have
known that agents, such as in Kiss, must use a common
communications protocol, i.e., an inter-agent communication
language, and related management functionality, Ex. 1020, 2;
Ex. 1046, 50, and therefore would have been motivated to
consider and employ one of the available inter-agent
communications languages and its related functionality, and
because doing so would have been a convenient and efficient
way to facilitate effective communication between agents.
Pet. 29–30 (citing Ex. 1003 ¶ 223).
IPR2019-00811
Patent 6,851,115 B1
88
Petitioner goes on to argue that “a Skilled Artisan would also have
been motivated based on earlier descriptions of KQML’s similar use of an
inter-agent communication language,” (citing Ex. 1016, Abstract) and their
disclosure that “[t]here are several levels at which agent-based systems must
agree, at least in their interfaces, in order to successfully interoperate:
Transport … Language … Policy … Architecture,” (citing Ex. 1016, 4, 29,
30) as well as their disclosure of “the necessity of a robust inter-agent
communication language in a system like Kiss.” Pet. 30 (citing Ex. 1006, 7;
Ex. 1007, 41, 4–-42; Ex. 1005, Abstract; Ex. 1003 ¶¶ 224–226).
Petitioner further argues that “[s]uch a combination would also have
been obvious to try this combination because the necessity of a common
inter-agent communication language was a known design need in a system
such as Kiss,” (citing Ex. 1020, 2; Ex. 1046, 50) and there were only a small
number of identified, predictable solutions to that need, such as FIPA ACL.”
Pet. 30 (citing Ex. 1035, 1; Ex. 1003 ¶ 227).
Petitioner goes on to argue that a “Skilled Artisan” would also have
been motivated to combine FIPA97 and Kiss in the manner described in the
Petition “because of the known advantages of FIPA97, including that it
specifies, ‘key agents necessary for the management of an agent system, the
ontology necessary for the interaction between systems, and it defines also
the transport level of the protocols’” (citing Ex. 1035, 1; Ex. 1003 ¶ 229),
and “because FIPA97 represented a substantial attempt to formulate an
industry standard designed to encourage interoperability and uniformity
amongst agent systems.” Pet. 30–31 (citing Ex. 1007, vii; Ex. 1047).
IPR2019-00811
Patent 6,851,115 B1
89
Petitioner also explains that:
[a] Skilled Artisan would have been further motivated to include
the administrative functionality and exemplary practices
described in FIPA97 with Kiss because the description of such
functionality and practices in FIPA97 suggests that their use with
the FIPA ACL would be advantageous and workable. A Skilled
Artisan would also have been motivated to employ such
additional functionality and practices in the combined system in
order to obtain the full benefit of using FIPA ACL.
Pet. 31 (citing Ex. 1003 ¶ 231).
In light of the reasons and the explanations Petitioner provides that
would have motivated a person of ordinary skill in the art to combine the
teachings of FIPA97 and Kiss in the manner Petitioner describes, Patent
Owner’s argument that “Petitioner utterly defaults in offering any reason to
combine Kiss and FIPA97” (see PO Resp. 116) (emphasis added), is directly
contrary to the evidence of record and is unpersuasive.
Patent Owner also argues that Petitioner “offers no reason or
explanation that a person of skill in the art looking at Kiss would be
motivated to find a particular language with the characteristics of FIPA97’s
ACL.” Id. at 116. Patent Owner’s argument, however, is misplaced. The
test for obviousness is not whether a person of ordinary skill in the art
“looking at Kiss would be motivated to find a particular language with the
characteristics of FIPA97’s ACL.” Rather, “[u]nder the correct
[obviousness] analysis, any need or problem known in the field of endeavor
at the time of invention and addressed by the patent can provide a reason for
combining the elements in the manner claimed.” KSR, 550 U.S. at 420.
Here, Petitioner provides persuasive evidence and argument that
FIPA97 and Kiss “are reasonably pertinent to the problem addressed by the
IPR2019-00811
Patent 6,851,115 B1
90
115 Patent—e.g., the need for ‘flexible, fault-tolerant, distributed problem
solving’ using ‘agent-based technologies.’” Pet. 29 (citing Ex. 1001, 4:34–
40; Ex. 1005, 2:33–40; Ex. 1007, 1; Ex. 1003 ¶ 221). This evidence
includes the ’115 patent’s observation that:
[a]gent-based systems have shown much promise for flexible,
fault-tolerant, distributed problem solving. Several agent-based
projects have helped to evolve the notion of facilitation.
However, existing agent-based technologies and architectures
are typically very limited in the extent to which agents can
specify complex goals or influence the strategies used by the
facilitator.
Ex. 1001, 4:34–40, as well as Kiss’ observation that “a need exists for a
knowledge management system for dynamic, distributed problem-solving
systems. Ex. 1005, 2:39–40. Indeed, Dr. Lieberman notes that the FIPA97
specification “defines a language and supporting tools, such as protocols, to
be used by intelligent software agents to communicate with each other,” and
that “the terms used and the mechanisms used [by FIPA97] support such a
higher-level, often task based, view of interaction and communication.”
Ex. 1003 ¶ 221. This evidence support’s Petitioner’s position that a person
of ordinary skill in the art at the time of the claimed invention would have
combined FIPA97 and Kiss in the manner described in the Petition to
provide a flexible, fault-tolerant, distributed problem solving system using
agent-based technologies.
Finally, addressing Patent Owner argument that Petitioner’s
motivation to combine Kiss and FIPA97 suffers from “hindsight bias” (PO
Resp. 116), we note that “[a]ny judgment on obviousness is in a sense
necessarily a reconstruction based upon hindsight reasoning, but so long as it
takes into account only knowledge which was within the level of ordinary
IPR2019-00811
Patent 6,851,115 B1
91
skill at the time the claimed invention was made and does not include
knowledge gleaned only from applicant's disclosure, such a reconstruction is
proper.” In re McLaughlin, 443 F.2d 1392, 1395 (CCPA 1971).
Here, Petitioner provides persuasive evidence that a person of
ordinary skill in the art at the time of the invention would have been
motivated to combine FIPA97’s communications protocol and inter-agent
communication language (FIPA ACL) between the agents of Kiss, adding its
administrative functionality and practices to the Kiss system. See, e.g., Pet.
28–32; Ex. 1003 ¶¶ 217–232. This is not evidence of hindsight bias. Patent
Owner does not identify, and we do discern, any particular knowledge used
by Petitioner in its combination of FIPA97 and Kiss that was gleaned from
the ’115 patent’s disclosure and was not within the level of ordinary skill in
the art at the time of the invention.
Based on the complete record, we find that Petitioner has shown
persuasively by a preponderance of the evidence that one of ordinary skill in
the art at the time of the invention would have been motivated to combine
the teachings of FIPA97 and Kiss in the manner proffered by Petitioner.
2. Discussion of Independent Claim 1 —
Differences Between the Claimed Subject Matter and the Prior Art
Claim 1 is directed to a method for communication and cooperative
task completion among distributed electronic agents comprising the steps of
registering a description of each active client agent’s functional capabilities
using an inter-agent language, receiving a request for service as a base goal
in the form of an arbitrarily complex goal expression in the inter-agent
language, and dynamically interpreting the arbitrarily complex goal
IPR2019-00811
Patent 6,851,115 B1
92
expression. Ex. 1001, 29:10–43. Claim 1 also recites that the “inter-agent
language” includes a “layer of conversational protocol” and a “content
layer.” Id. at 29:16–24. In addition, claim 1 recites that the step of
“dynamically interpreting the arbitrarily complex goal expression”
comprises “generating sub-goals” and “constructing a goal satisfaction
plan.” Id. at 29:28–32. The recited “goal satisfaction plan” in turn includes
“a suitable delegation of sub-goal requests” and “dispatching each of the
sub-goals to a selected client agent for performance.” Id. at 29:32–41.
In its proposed combination of Kiss with FIPA97, Petitioner relies on
Kiss to teach electronic agents and their general communication and
collaboration functions, including the steps recited in claim 1, whereas
FIPA97 is relied upon to teach the recited “inter-agent language” and the
limitations relating to the inter-agent language, such as “layer of
conversational protocol,” a “content layer,” an “arbitrarily complex goal
expression,” and “sub-goals expressed in the inter-agent language.” Pet. 28,
32–54. Patent Owner argues that the proposed combination of Kiss and
FIPA97 does not teach the “goal satisfaction plan” recited in the claim. PO
Resp. 85–92. Patent Owner also asserts that FIPA97 does not teach the
“inter-agent language” (or Interagent Communication Language (ICL)) and
related limitations. Id. at 95–108. In what follows, we analyze the
recitations of claim 1 in more detail, addressing Petitioner’s contentions and,
where applicable, Patent Owner’s responsive arguments.
a. Preamble
Claim 1 recites a preamble as follows: “A computer-implemented
method for communication and cooperative task completion among a
IPR2019-00811
Patent 6,851,115 B1
93
plurality of distributed electronic agents.”21 Petitioner asserts that Kiss
teaches the recited “distributed electronic agents” because Kiss describes
that “[i]nteraction between a user and the knowledge resources is mediated
by a collection of cooperative intelligent agents” (Pet. 32 (citing Ex. 1005,
2:43–49, Fig. 1)), where the agents are “distributed” in a “hierarchical
architecture” (id. at 33 (citing Ex. 1005, Abstract, 2:50–55, 4:57–59)).
Petitioner contends that Kiss also teaches the recited “communication
and cooperative task completion” among the electronic agents because Kiss
describes that “cooperative intelligent agents [that] incorporate generalized
automated negotiation and distributed inference (i.e. problem-solving)
processes,” analyze “problem statements,” and reorganize them as “sets of
tasks.” Pet. 33 (citing Ex. 1005, 2:50–55).
Patent Owner does not specifically dispute that Kiss teaches the
preamble of claim 1. See generally PO Resp. Based on the complete record
and for the reasons explained by Petitioner, we are persuaded that Petitioner
has demonstrated sufficiently that Kiss teaches the preamble of claim 1.22
b. Registering a Description
The first step recited in claim 1 is “registering a description of each
active client agent’s functional capabilities as corresponding registered
21 Because Petitioner has shown that the recitations in the preamble is
satisfied by Kiss, we need not determine whether the preamble is limiting.
See Vivid Techs., 200 F.3d at 803.
22 We also find that Patent Owner has waived any argument directed to the
preamble of claim 1. See Paper 16 (Scheduling Order), 7 (“Patent Owner is
cautioned that any arguments for patentability not raised in the response may
be deemed waived.”).
IPR2019-00811
Patent 6,851,115 B1
94
functional capabilities.” Ex. 1001, 29:13–15. Petitioner asserts that Kiss
teaches this limitation because Kiss describes a “registry of agents,” which is
constructed through “registering a description of each active client agent’s
functional capabilities as corresponding registered functional capabilities.”
Pet. 34 (citing Ex. 1005, 3:37–45, 6:66–7:19, 8:41–48, 10:32–35, 12:18–20,
Fig. 1).
Patent Owner does not specifically dispute that Kiss teaches this
“registering a description” limitation. See generally PO Resp. Based on the
complete record and for the reasons explained by Petitioner, we are
persuaded that Petitioner has demonstrated sufficiently that Kiss teaches
“registering a description of each active client agent’s functional capabilities
as corresponding registered functional capabilities,” as recited in claim 1.23
c. An Inter-agent Language Including a Layer of Conversational
Protocol and a Content Layer
Claim 1 recites that “registering a description of each active client
agent’s functional capabilities” is performed using an “expandable,
platform-independent, inter-agent language” (Ex. 1001, 29:13–16) and that
the “inter-agent language” includes:
a layer of conversational protocol defined by event types and
parameter lists associated with one or more of the events,
wherein the parameter lists further refine the one or more
events;
23 We also find that Patent Owner has waived any argument directed to this
claim limitation. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
95
a content layer comprising one or more of goals, triggers and data
elements associated with the events.
Id. at 29:16–24. Petitioner asserts that FIPA97 teaches each of these
limitations relating to an “inter-agent language.” Pet. 34–43.
Patent Owner asserts that FIPA97 does not teach “refin[ing] . . .
events” and “triggers” recited in the claim. PO Resp. 97–108. In addition,
Patent Owner argues that FIPA ACL is not “expandable” because, among
other reasons, FIPA ACL does not have the capability of refining events. Id.
at 105–106. Patent Owner does not dispute FIPA97 teaches the rest of the
limitations relating to an “inter-agent language” recited in clam 1, as set
forth above. See id. at 95–108.
(i) Inter-agent Language
Petitioner relies on Agent Communication Language of FIPA97
(FIPA ACL) as teaching the “inter-agent language” recited in claim 1.
Pet. 28 (citing Exs. 1006, 1007, 1008), 35. In particular, Petitioner contends
that FIPA97 discloses Agent Communication Language (ACL) “with
precisely defined syntax, semantics and pragmatics” for “communication
between independently designed and developed software agents.” Id. at 35
(citing Ex. 1007, 3).
(ii) Platform-independent, Expandable Inter-agent Language
Petitioner asserts that FIPA ACL is “platform-independent,” as recited
in claim 1, because it is designed to “maximise interoperability across agent-
based applications, services and equipment,” and to facilitate the
“construction and management of an agent system composed of different
agents,” even when those systems are “built by different developers.”
IPR2019-00811
Patent 6,851,115 B1
96
Pet. 35 (citing Ex. 1007, vii, ix). Patent Owner does not dispute that FIPA
ACL is “platform-independent,” as recited in the claim. See PO Resp. 95–
108.
Petitioner also contends that FIPA ACL is “expandable,” as recited in
claim 1, because in FIPA97 “[a]gents may use communicative acts with
other names, not defined” in the specification (Pet. 35 (citing Ex. 1007, 12))
and agents “may choose to implement any subset (including all, though this
is unlikely) of the pre-defined message types and protocols” (id. at 35–36
(citing Ex. 1007, 12)). In addition, Petitioner asserts that in FIPA97 the
“formal basis” of the language is “supplemented with pragmatic extensions”
that help “ease the practical implementation of effective inter-agent
communications.” Id. at 36 (citing Ex. 1007, 11). Petitioner argues that
FIPA97 also discloses possible extensions to the inter-agent communication
language, such as MIME header support. Id. (citing Ex. 1007, 82).
Patent Owner asserts that FIPA ACL is not “expandable” basically for
two reasons. PO Resp. 105–106. First, citing the prosecution history, Patent
Owner argues that “the concept of expandability” relates to the feature of
“refin[ing] . . . events” recited in the claim and that FIPA ACL is not
“expandable” because FIPA ACL does not have the capability of refining
events. Id. Second, Patent Owner contends that FIPA ACL is not
“expandable” in a general sense, citing various criticism of FIPA ACL from
“[r]esearchers in the field” regarding FIPA ACL’s alleged rigidity and
inflexibility in certain aspects. Id. at 106. We address each of these
arguments in turn.
IPR2019-00811
Patent 6,851,115 B1
97
Addressing first Patent Owner’s argument about “the concept of
expandability” as it relates to the feature of “refin[ing] . . . events” recited in
the claim, Patent Owner cites the following statement by the Examiner in the
Notice of Allowability:
The disclosed agent communication language [KQML] does not
read upon the cited agent language because the layer does not
define an event type as well as the parameter lists that further
refines the event. Nwana’s language at best has separate layers
for the event and the parameters associated with the event. By
Applicant providing these parameters in the same layer as the
event such that they further refine the event, a standard set of
events are dynamically extensible based upon the parameter list
which is not possible with the teachings of Nwana. Therefore,
the claims are allowable over the prior art of record.
Ex. 1002, 14 (emphases added); PO Resp. 105 (citing Ex. 1002, 14). As
indicated in the Amendment submitted before the allowance of the claims,
claim 1 was amended to recite “wherein the parameter lists further refine the
one or more events” to distinguish the claim from the combination of
Nwana/KQML and Kiss cited by the Examiner. See Ex. 1002, 13–14, 18,
28, 51. Patent Owner argues, therefore, that the limitation “expandable”
recited in claim 1 relates to the feature of the parameter lists that “refine . . .
events,” as recited in the claim. PO Resp. 105. Patent Owner asserts that
FIPA ACL is not “expandable,” as recited, because FIPA ACL does not
teach parameters that further refine events. Id.
In response, Petitioner asserts that the claim language “using an
expandable, platform-independent, inter-agent language” was part of the
originally filed claims, which were rejected by the Examiner over the
combination of Nwana/KQML and Kiss. Pet. Reply 59–60 (citing Ex. 1002,
IPR2019-00811
Patent 6,851,115 B1
98
763). Petitioner argues that, therefore, the Examiner’s discussion of
“dynamically extensible” events resulting from the “refin[ing] . . . events”
feature added in the Amendment is unrelated to the “expandable” language
that was present in the claim as originally filed. Id. at 60. Petitioner also
asserts that, even if the “expandable” language relates to the “refin[ing] . . .
events” feature, FIPA ACL is “expandable” because FIPA ACL teaches
parameter lists that refine events. Id.
We need not determine whether the “expandable” language relates to
the “refin[ing] . . . events” feature added in the Amendment because, as
discussed below, we agree with Petitioner that FIPA ACL teaches
“parameter lists” that “further refine the one or more events,” as recited in
claim 1. Thus, to the extent Patent Owner asserts that the limitation
“expandable” recited in claim 1 relates to the feature of the parameter lists
“refin[ing] . . . events,” Petitioner demonstrates sufficiently that FIPA ACL
teaches an “expandable . . . inter-agent language.”
Turning to the Specification, the ’115 patent describes that “[a]s new
agents connect to the facilitator, registering capability specifications and
natural language vocabulary, what the user can say and do dynamically
changes; in other words, the ICL is dynamically expandable.” Ex. 1001,
8:41–44 (emphases added). That is, the ’115 patent suggests that the
expandability of the ICL (or “inter-agent language,” as recited in claim 1)
results from or flows from the ICL’s feature of “registering capability
IPR2019-00811
Patent 6,851,115 B1
99
specifications and natural language vocabulary.”24 Id. As noted in the
Institution Decision, this portion of the Specification appears to describe an
exemplary embodiment. Inst. Dec. 54 (citing Ex. 1001, 8:40–49).
As discussed above, Petitioner demonstrates sufficiently Kiss teaches
“registering a description of each active client agent’s functional capabilities
as corresponding registered functional capabilities,” as recited in claim 1.
Thus, to the extent the portion of the Specification discussed above describes
an embodiment of an “expandable . . . inter-agent language,” we are
persuaded that the combination of Kiss and FIPA97 teaches “registering a
description of each active client agent’s functional capabilities as
corresponding registered functional capabilities, using an expandable,
platform-independent, inter-agent language,” as recited in claim 1 (see Pet.
34–36).
Next, Patent Owner asserts that FIPA ACL is not “expandable” in a
more general sense, citing various “criticisms” from “[r]esearchers in the
field” about FIPA ACL’s alleged “rigidity and lack of expandability.”
PO Resp. 106 (citing Ex. 2031, 2 (“[i]t is unlikely that a single set of axioms
will cover all eventualities because communication is inherently context
dependent”); Ex. 1081, 2 (“[T]he current FIPA-ACL spec tends to
unnecessarily over-constrain the feasibility of communication acts . . . .”).
Patent Owner also contends that Petitioner’s declarants, Dr. Finin and
Dr. Lieberman, made allegedly similar statements. Id. (citing Ex. 1082, 2
24 As described in the Specification, agents connecting to the facilitator is
part of the operation of the system of the ’115 patent, not a feature of the
ICL. See Ex. 1001, 6:24–44; 7:13–32.
IPR2019-00811
Patent 6,851,115 B1
100
(“The FIPA proposal only covers the most rudimentary message types.”);
Ex. 2014, 207:5–20 (“FIPA is a standard, like a miles-per-gallon standard
. . . So that doesn’t say how you’re going to build a car; it just says a car has
to get . . . 50 miles a gallon.”)).
We are not persuaded by Patent Owner’s argument and evidence
because Patent Owner does not explain adequately why the cited extrinsic
evidence regarding FIPA ACL’s alleged “rigidity” or inflexibility in certain
aspects shows that FIPA ACL is not “expandable” as recited in the claim. In
the Institution Decision, based on the preliminary record, we did not find
anything in the intrinsic record that justifies deviating from the plain
meaning of the claim term “expandable.” Inst. Dec. 54–55. Upon
considering the complete record, we discern no reason to depart from our
preliminary assessment.25
We may “at any time in order to better understand the underlying
technology and may also rely on dictionary definitions when construing
claim terms, so long as the dictionary definition does not contradict any
definition found in or ascertained by a reading of the patent documents.”
Phillips, 415 F.3d at 1322–23 (citation omitted). Accordingly, in the
Institution Decision, finding no readily discernable meaning of “expandable”
in the intrinsic record, we consulted a technical dictionary, the 7th edition of
25 In the Patent Owner Response, Patent Owner does not discuss the
Specification or argue that the meaning of “expandable” recited in the claim
can be discerned from the Specification, although we invited the parties in
the Institution Decision to address the term “expandable” further in their
papers during trial. See PO Resp. 105–107 (citing Inst. Dec. 55–56). Thus,
Patent Owner has waived its argument regarding the meaning of
“expandable” based on any disclosures in the Specification. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
101
the IEEE Dictionary (Ex. 3001).26 The IEEE Dictionary defines
“expandability” as follows: “expandability (1) (supervisory control, data
acquisition, and automatic control) The capability of a system to be
increased in capacity or provided with additional functions. (2) See also:
extendability.” Ex. 3001, 3.27 The same dictionary defines “extensible
language” as follows.
A computer language that can be altered or can alter itself to
provide a programmer with additional user-specified functions
or capabilities. Examples include Ada, ALGOL, FORTH, and
LOGO, because each can be used in a building block fashion to
construct increasingly complex functions.
Id. at 4. Upon considering the complete record, we determine that these
dictionary definitions are consistent with the intrinsic record, including the
description of an embodiment in the Specification regarding “the ICL
[being] dynamically expandable.” See Ex. 1001, 8:41–44.
In the Institution Decision, based on the preliminary record, we found
that the characteristics of FIPA ACL argued by Petitioner—such as the
ability to use communicative acts with other names not defined in the
language specification, the pragmatic extensions, and the MIME
extension—appear to fall within the plain meaning of the term
“expandable,” as defined in the IEEE Dictionary. Inst. Dec. 55; see Pet. 35–
36 (citing Ex. 1007, 11, 12, 82). Patent Owner does not argue otherwise.
26 IEEE 100 THE AUTHORITATIVE DICTIONARY OF IEEE STANDARDS TERMS
(7th ed. 2000) (Ex. 3001).
27 The page numbers for Exhibit 3001 refer to the page numbers inserted in
the bottom, right-hand corner of each page.
IPR2019-00811
Patent 6,851,115 B1
102
See PO Resp. 105–107. Based on the complete record and for the reasons
discussed above, we determine that Petitioner has shown sufficiently that the
combination of Kiss and FIPA97 teaches “registering a description of each
active client agent’s functional capabilities as corresponding registered
functional capabilities, using an expandable, platform-independent, inter-
agent language,” as recited in claim 1.
(iii) A Layer of Conversational Protocol Defined by Event Types and
Parameter Lists Associated With the Events
Claim 1 recites that the “inter-agent language” includes “a layer of
conversational protocol defined by event types and parameter lists associated
with one or more of the events, wherein the parameter lists further refine the
one or more events.” Ex. 1001, 29:16–21. Petitioner contends that FIPA97
teaches each of these limitations. Pet. 36–42.
(1) A Layer of Conversational Protocol
Petitioner asserts that FIPA97 teaches “a layer of conversational
protocol” recited in claim 1 because FIPA ACL “defines the semantics of
messages between agents as including a communicative act, followed by
various parameters and parameter expressions,” as shown in Figure 1 of Part
2 of FIPA97 reproduced above in the Overview of FIPA97 section
(§ III.E.3.). Pet. 36–37 (emphasis added) (citing Ex. 1007, 2; Ex. 1003
¶ 235). Petitioner also argues that FIPA97 “discloses a number of different
performatives and rules governing FIPA ACL messages and the meaning of
the different parts” and, therefore, discloses “a layer of conversational
protocol.” Id. at 37 (citing Ex. 1007, 11–22; Ex. 1003 ¶¶ 236–239).
IPR2019-00811
Patent 6,851,115 B1
103
(2) Events and Event Types
Petitioner maps the “event” of claim 1 to FIPA ACL’s “message” and
asserts that FIPA ACL’s “message type” teaches an “event type” recited in
claim 1. Id. at 37–38 (citing Ex. 1007, 3, 4, 11–13, 18). Petitioner argues
that FIPA ACL’s “message” is an “event” recited in claim 1 because the
FIPA ACL communicative acts “are performed by an agent sending a
message to another agent.” Id. at 37 (emphasis added) (citing Ex. 1007, 3, 4,
11).
Referencing Figure 1 of Part 2 of FIPA97 reproduced above,
Petitioner asserts that in FIPA ACL, “[t]he first element of the message is a
word which identifies the communicative act being communicated, which
defines the principal meaning of the message,” and “the message’s
communicative act type corresponds to that which in KQML is called the
performative.” Pet. 37 (emphasis added) (citing Ex. 1007, 12, 13).
Petitioner argues that FIPA97 further defines “message type” as
corresponding to the “communicative acts/performatives of the FIPA ACL.”
Id. at 38 (citing Ex. 1007, 18).
As discussed above, Petitioner asserts that FIPA97 teaches “a layer of
conversational protocol” because FIPA ACL “defines the semantics of
messages between agents as including a communicative act.” Pet. 36
(emphasis added) (citing Ex. 1007, 2). Petitioner contends that FIPA97
teaches the layer of conversational protocol in FIPA ACL is “defined by
event types,” as recited in claim 1, because in FIPA97, “[t]he message types
are a reference to the semantic acts defined in this specification.” Id. at
37–38 (underlined emphases added) (citing Ex. 1007, 11).
IPR2019-00811
Patent 6,851,115 B1
104
(3) Parameter Lists
Claim 1 recites that “a layer of conversational protocol” is also
“defined by . . . parameter lists associated with one or more of the events.”
Petitioner asserts that in FIPA97, FIPA ACL messages may include multiple
different parameters (“parameter lists associated with one or more of the
events”) that can “help the message transport service to deliver the message
correctly” or “the receiver to interpret the meaning of the message,” or “to
respond co-operatively,” as shown in Figure 1 of Part 2 of FIPA97
(reproduced above). Pet. 38–39 (citing Ex. 1007, 12, 14, Fig. 1). Petitioner
contends that FIPA97 discloses additional parameters in Table 1. Id. at 39
(citing Ex. 1007, 13–14).
(4) Refining Events
Claim 1 recites that “the parameter lists further refine the one or more
events.” Petitioner asserts that this phrase should be construed to mean that
“a list of parameters associated with an event can refine the event by
affecting the meaning of the event.” Pet. 13. Petitioner discusses the
disclosures in the Specification and the prosecution history of the ’115
patent in support of its proposed construction. Id. at 13–16 (citing Ex. 1001,
11:2–37; Ex. 1002, 13–14, 51, 55; Ex. 1020, 5, 6; Ex. 1003 ¶¶ 90, 91, 94–
98).
Although Patent Owner discusses constructions for six terms in the
Patent Owner Response, Patent Owner does not discuss Petitioner’s
proposed construction for the claim term “refine . . . events” or propose its
own construction. See PO Resp. 24–38. Instead, Patent Owner argues that
FIPA97 does not teach parameters that “refine . . . events” under Petitioner’s
IPR2019-00811
Patent 6,851,115 B1
105
proposed construction. See PO Resp. 97–98 (“the cited parameters in
FIPA97 do not, contrary to Petitioner’s assertion, affect the meaning of the
message”), 100 (“this parameter does not affect the meaning of the
message”), 102 (“In contrast, the FIPA97 “receiver” parameter does not
affect the meaning of the “inform” communicative act.”). Based on the
arguments and evidence presented by Patent Owner, we understand Patent
Owner not to dispute Petitioner’s proposed construction for the term “the
parameter lists further refine the one or more events.”28
Upon considering the complete record, we agree with Petitioner that
the claim term “refine . . . events” encompasses “affecting the meaning of
the events.” As discussed above, during prosecution, the limitation “wherein
the parameter lists further refine the one or more events” was added to
claim 1 in an amendment to distinguish the claim from the combination of
Nwana/KQML and Kiss cited by the Examiner. See Ex. 1002, 13–14, 18,
28, 51; Pet. 14–16 (citing Ex. 1002, 13–14, 51); PO Resp. 105 (citing
Ex. 1002, 14). Citing the disclosures of Nwana/KQML discussed by the
Examiner, Petitioner persuasively argues that the parameters of Nwana’s
KQML message (i.e., the claimed “event”) “tell” do not change the meaning
of the message, whereas the parameters of an event of the ’115 patent’s
change the meaning of the event by “refin[ing] the event.” Pet. 14–16
(citing Ex. 1002, 13–14, 51, 55; Ex. 1020, 5, 6). In support of its argument,
Petitioner cites the following passage from the Specification that
28 We also find that Patent Owner has waived any argument regarding claim
construction of this term. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
106
distinguishes the ’115 patent’s ICL (i.e., the recited “inter-agent language”)
from KQML.
For example, in KQML, a request to satisfy a query can employ
either of the performatives ask_all or ask_one. In ICL, on the
other hand, this type of request preferably is expressed by the
event type evost solve, together with the solution_limit(N)
parameter--where N can be any positive integer. (A request
for all solutions is indicated by the omission of the solution limit
parameter.) [The request can also be accompanied by other
parameters, which combine to further refine its semantics.] In
KQML, then, this example forces one to choose between two
possible conversational options, neither of which may be
precisely what is desired. In either case, the performative chosen
is a single value that must capture the entire conversational
characterization of the communication.
Pet. 13 (alteration and underlined emphasis added) (quoting Ex. 1001,
11:21–37). Petitioner argues that a message having the KQML performative
“ask_all” always has the same meaning, requesting all solutions to the
request, just as the KQML performative “ask_one” always requests one
solution. Id. at 14. According to Petitioner, in contrast, the meaning of the
’115 patent’s event “evost_solve” depends on the value of its parameter
“solution_limit(N)”—e.g., an agent may request the identity of three agents
capable of translating a document by including “solution_limit(N)” in the
message and setting the value of N to 3. Id. Thus, Petitioner argues that the
parameters of an event of the ’115 patent affect the meaning of the event.
Id. at 13–16.
We are persuaded by Petitioner’s argument and evidence that the
“parameterized approach” of the ’115 patent (Ex. 1001, 11:16–21) can refine
an event by changing the meaning of the event. As discussed above in
IPR2019-00811
Patent 6,851,115 B1
107
Section III.B., we construe “event” to mean “a message or goal
communicated between agents” and “goal” to mean “a request for service.”
The ’115 patent describes that “[i]n one embodiment, a request for one of an
agent’s services normally arrives in the form of an event from the agent’s
facilitator.” Id. at 12:44–46 (emphasis added). In a section titled “Refining
Service Requests,” the ’115 patent describes as follows:
In a preferred embodiment of the present invention, parameters
associated with a goal (or sub-goal) can draw on useful features
to refine the request’s meaning. For example, it is frequently
preferred to be able to specify whether or not solutions are to be
returned synchronously; this is done using the reply parameter,
which can take any of the values synchronous, asynchronous, or
none. As another example, when the goal is a non-compound
query of a data solvable, the cache parameter may preferably be
used to request local caching of the facts associated with that
solvable.
Id. at 15:49–58 (emphases added). Thus, the ’115 patent describes that in an
embodiment, parameters of a request for service, i.e., an event, can refine the
“meaning” of the request or event.
Nonetheless, we note that this passage and all of the disclosures of the
’115 patent cited by Petitioner describe exemplary embodiments. See id. at
11:25–28 (“In ICL . . . this type of request preferably is expressed by the
event type evost solve, together with the solution_limit(N) parameter--where
N can be any positive integer.” (emphasis added)), 15:49–51 (“In a preferred
embodiment of the present invention, parameters associated with a goal (or
sub-goal) can draw on useful features to refine the request’s meaning.”
(emphases added)). Thus, the term “refine . . . events” recited in claim 1
may encompass but is not necessarily limited to “affecting the meaning of
IPR2019-00811
Patent 6,851,115 B1
108
events.” See SuperGuide, 358 F.3d at 875 (“a particular embodiment
appearing in the written description may not be read into a claim when the
claim language is broader than the embodiment”); WesternGeco, 889 F.3d at
1323–24 (“It is well established that claims are not limited to preferred
embodiments, unless the specification clearly indicates otherwise.”). Indeed,
Petitioner’s proposed construction of the term “the parameter lists further
refine the one or more events” expresses the meaning of the term in a
permissive fashion—“an event can refine the event by affecting the meaning
of the event.” Pet. 13 (emphasis added).
Based on the complete record, we determine that the term “the
parameter lists further refine the one or more events” encompasses
parameters “affecting the meaning of the events.”
Petitioner asserts that FIPA97 teaches this limitation “the parameter
lists further refine the one or more events” because several of the
conversational layer parameters disclosed in FIPA97 affect the meaning of
the performative included within the message, and therefore affect the
meaning of the message, or event. Pet. 40. According to Petitioner, FIPA97
discloses a parameter “:receiver,” which “can alter the meaning [of] an
event, such as ‘inform,’ to be ‘inform one’” or “inform a number’ of agents,
depending on the value of the parameter.” Id. (citing Ex. 1007, 13). Citing
the testimony of Dr. Lieberman, Petitioner asserts that naming a tuple
“corresponds to the action of multicasting the message” such that “semantics
of this multicast” is refined so that “the message is sent to each agent named
in the tuple.” Id. (citing Ex. 1007, 13; Ex. 1003 ¶ 258). Petitioner also
argues, “[t]his ‘tuple naming’ functionality is analogous to the example
IPR2019-00811
Patent 6,851,115 B1
109
provided in the ’115 Patent of ‘solution_limit(N),’ in which ‘N’ modifies
how many solutions are requested, and is not found in the KQML Nwana
reference that was before the examiner.” Id. (citing Ex. 1001, 11:16–42;
Ex. 1020; Ex. 1003 ¶ 258). Petitioner also asserts that FIPA97 discloses two
other parameters—“:protocol” when used with the “call for proposals”
performative and “:conversation-id”—that affect the meaning of a message.
Id. at 41–42 (citing Ex. 1007, 13–14, 46–49; Ex. 1003 ¶¶ 261–265).
Patent Owner asserts that FIPA97 does not teach parameters
“refin[ing] . . . events,” as recited in the claim, because the parameters of
FIPA97 cited by Petitioner does not affect the meaning of the message.
PO Resp. 97–98. In particular, Patent Owner contends that FIPA97’s
“:receiver” parameter of the FIPA97 “inform” message relied upon by
Petitioner is no different from KQML’s “:receiver” parameter for the KQML
“tell” message in that both modify the message but do not affect the meaning
of the message. Id. at 98–99 (citing Ex. 1007, 12; Ex. 1020, 5). Patent
Owner asserts that FIPA97’s “:receiver” parameter does not affect the
meaning of the FIPA97 “inform” message because “the result of this
parameter is only that the message is multicast; i.e., it is ‘sent to each agent
named in the tuple.’” Id. at 100. According to Patent Owner, “the content
of the message is unchanged; the only thing that changes is who (as in,
recipient) receives the message.” Id. at 101.
IPR2019-00811
Patent 6,851,115 B1
110
The portion of FIPA97 cited by Petitioner is reproduced below.
Ex. 1007, 13. Pointing to this disclosure in FIPA97, Petitioner asserts that
the “receiver” parameter affects the meaning of the event by affecting
“whether the performative communicates the message to a single agent, a
selection of agents, or if [the message is] broadcast to every agent.”
Pet. Reply 56 (citing Ex. 1007, 13). According to Petitioner, “in FIPA97, an
event using ‘inform’ can mean three or more different things (i.e., inform
(1) one person, (2) a select group of people, (3) everybody) depending on the
value of just that parameter.” Id.
We agree with Petitioner’s argument and disagree with Paten Owner’s
argument. As shown above in the portion of FIPA97 relied on by Petitioner,
FIPA97 describes the “meaning” of the message affected by the “:receiver”
parameter (under the column heading “Meaning:”) as follows:
Note that the recipient may be a single agent name, or a tuple of
agent names. This corresponds to the action of multicasting the
IPR2019-00811
Patent 6,851,115 B1
111
message. Pragmatically, the semantics of this multicast is that
the message is sent to each agent named in the tuple, and that the
sender intends each of them to be recipient of the CA encoded in
the message. For example, if an agent performs an inform act
with a tuple of three agents as receiver, it denotes that the sender
intends each of these agent to come to believe the content of the
message.
Ex. 1007, 13 (emphases added). We agree with Petitioner this passage
describes, under the column heading “Meaning,” the meaning of the
message modified or affected by the “:receiver” parameter.
Patent Owner argues that the FIPA97 “:receiver” parameter does not
change “the content of the message” and that “the only thing that changes is
who (as in, recipient) receives the message.” PO Resp. 101. Patent Owner
contends that, in contrast, the solution_limit(N) parameter of the ’115 patent
“modifies the actual request/communicative act/performative itself—i.e.,
evost solve, because it imposes a modification on the limits of solutions
presented in response to a request.” Id. at 102 (citing Ex. 2032 ¶161).
We are not persuaded by Patent Owner’s argument. First, we disagree
with Patent Owner that the FIPA97 “:receiver” parameter changes only
“who (as in, recipient) receives the message.” As described in FIPA97, the
“:receiver” parameter does not merely specify “who receives the message,”
e.g., agent A as opposed agent B. Rather, as persuasively explained by
Petitioner, the “:receiver” parameter affects the qualitative aspect of the
message—i.e., whether to inform (1) one agent, (2) a select group of agents,
or (3) every known agent. See Pet. Reply 56.
In addition, Patent Owner does not explain adequately why the
solution_limit(N) parameter of the ’115 patent specifying the number of
IPR2019-00811
Patent 6,851,115 B1
112
agents to be queried “modifies the actual request/communicative
act/performative itself” but the “:receiver” parameter of FIPA97 specifying
the number of agents to which to send a message does not. See PO Resp.
100–101. We agree with Petitioner that Patent Owner does not identify “a
practical distinction between limiting the number of agents to be queried
(solution_limit(N)), and querying a selection of agents (:receiver).”
Pet. Reply 57.
Next, Patent Owner contends that FIPA97 does not teach parameters
“refin[ing] . . . events,” as recited in the claim, because FIPA97’s “:receiver”
parameter is the same as or similar to the KQML “:receiver” parameter,
which does not modify the meaning of a message. PO Resp. 98–99 (citing
Ex. 1007, 12; Ex. 1020, 5). In support of its argument, Patent Owner cites a
page from Dr. Finin’s presentation slides at an unidentified meeting or class,
which compares “KQML tell and FIPA ACL inform.” Id. 99–100 (citing
Ex. 2019, 15; Ex. 2013, 132:6–22.). Patent Owner also cites the testimony
from Dr. Finin at his deposition that “the general idea of agent
communication language that was embodied in KQML is similar to the one
that is embodied in FIPA.” Id. at 100 (citing Ex. 2012, 32:10–17).
Patent Owner, however, does not discuss the disclosure of
KQML/Nwana (Ex. 1020) sufficiently (other than pointing to the similarity
of the format between the “:receiver” parameter of KQML and FIPA97) or
explain adequately why KQML’s “:receiver” parameter is similar to the
FIPA97 “:receiver” parameter. See PO Resp. 98–100. For example, Patent
Owner does not explain whether KQML’s “:receiver” parameter specifies,
similar to FIPA97, that “the recipient may be a single agent name, or a tuple
IPR2019-00811
Patent 6,851,115 B1
113
of agent names,” that “the message is sent to each agent named in the tuple,”
and that “the sender intends each of them to be recipient of the message.”
See id.; Ex. 1007, 13 (emphasis added). The general statements from
Dr. Finin (or a slide purportedly from his unidentified presentation) cited by
Patent Owner (PO Resp. 99–100 (citing Ex. 2019, 15; Ex. 2013, 132:6–22,
32:10–17) regarding the alleged similarity between KQML and FIPA97 at a
general level are simply insufficient to override the specific disclosures in
FIPA97 cited by Petitioner regarding the meaning of the message modified
or affected by the FIPA97 “:receiver” parameter.
Patent Owner also cites testimony from Petitioner’s declarant,
Dr. Lieberman, as “confirm[ing]” that the FIPA97 “:receiver” parameter
does not change “the content of the message” and that “the only thing that
changes is who (as in, recipient) receives the message.” PO Resp. 101
(citing Ex. 2014, 110:14–111:16, 168:5–12). The cited testimony of
Dr. Lieberman, however, discusses FIPA97’s messages and multicasting in
general (responding to general questions from Patent Owner’s counsel) and
does not discuss the specific disclosure in FIPA97 cited by Petitioner
regarding the meaning of the message modified or affected by the FIPA97
“:receiver” parameter. See Ex. 2104, 108:4–111:16, 168:5–12. Thus, we
find Patent Owner’s citation to the testimony of Dr. Lieberman to be
unpersuasive (if not misleading) to “confirm” that the FIPA97 “:receiver”
parameter does not change “the content of the message” and that “the only
thing that changes is who (as in, recipient) receives the message.” See
PO Resp. 101.
IPR2019-00811
Patent 6,851,115 B1
114
Next, Patent Owner cites 9 paragraphs from the Declaration of
Dr. Cohen (Ex. 2033) to argue that “FIPA97 does not contain any
conversational layer parameters that affect the semantics of FIPA97
messages.” PO Resp. 103 (citing Ex. 2033 ¶¶ 80–89).29 Patent Owner,
however, does not discuss Dr. Cohen’s testimony in the Patent Owner
Response or explain how Dr. Cohen’s testimony supports its contention. See
id. Thus, to the extent Petitioner purports to rely on Dr. Cohen’s testimony,
this amounts to improper incorporation by reference in violation of
37 C.F.R. § 42.6(a)(3). Accordingly, we decline to consider the cited
paragraphs from the Cohen Declaration. We note, nonetheless, that
Dr. Cohen’s testimony appears to dwell in generalities, e.g., citing the works
of logicians/philosophers Tarski and Frege (see Ex. 2033 ¶¶ 80–81), and
discusses tuples and multicasting in general terms (see id. ¶¶ 85–88).
Lastly, Patent Owner contends that the limitation “the parameter lists
further refine the one or more events” requires “the parameters of an event
(or message) can refine the message or goal itself” (PO Resp. 97) or “change
the nature of the communicative act itself” (id. at 102). Although it is not
entirely clear what Patent Owner argues,30 to the extent Patent Owner argues
“refin[ing]” a messages or goal requires changing the message or goal itself,
we disagree with Patent Owner’s argument. As discussed above, in a
section titled “Refining Service Requests,” the ’115 patent describes
29 Patent Owner’s citation to the Cohen Declaration appears to be mistaken
since the last paragraph of Cohen Declaration is paragraph 88.
30 As discussed above, Patent Owner does not dispute Petitioner’s proposed
construction of the term “the parameter lists further refine the one or more
events.” See PO Resp. 24–38.
IPR2019-00811
Patent 6,851,115 B1
115
examples of “parameters” associated with a request or goal that “refine the
request’s meaning,” including the “reply parameter” that specifies whether
solutions are to be returned synchronously or asynchronously, and the
“cache parameter” that specifies local caching of facts related to the request.
See Ex. 1001, 15:49–58. These parameters do not appear to change the
request itself but, rather, specify how the request should be handled—e.g.,
synchronous response, local caching, etc. Thus, we are not persuaded by
Patent Owner’s argument that the claim requires the recited “parameters”
“refine the message or goal itself” (PO Resp. 97) or “change the nature of
the communicative act itself” (id. at 102). Such an interpretation, which
“excludes a [disclosed] embodiment from the scope of the claim is rarely, if
ever, correct.” Broadcom Corp. v. Emulex Corp., 732 F.3d 1325, 1333 (Fed.
Cir. 2013) (quoting Accent Pkg., Inc. v. Leggett & Platt, Inc., 707 F.3d 1318,
1326 (Fed. Cir. 2013)) (citing Phillips, 415 F.3d at 1312–13).
Based on the foregoing and upon considering the complete record, we
determine that Petitioner has shown sufficiently that FIPA97 teaches
“wherein the parameter lists further refine the one or more events,” as
recited in claim 1. Based on the complete record and for the reasons
explained by Petitioner, we determine that Petitioner has shown sufficiently
that FIPA97 teaches “a layer of conversational protocol defined by event
types and parameter lists associated with one or more of the events, wherein
the parameter lists further refine the one or more events,” as recited in
claim 1.
IPR2019-00811
Patent 6,851,115 B1
116
(vi) Content Layer Comprising Goals and Triggers
Claim 1 recites “a content layer comprising one or more of goals,
triggers and data elements associated with the events.” Petitioner asserts
that FIPA97 teaches the recited “content layer” because FIPA97 describes
various rules and standards governing the content that may be embedded
within messages (“a content layer”). Pet. 42 (citing Ex. 1007, 3, 9, 14–16).
Petitioner argues that the ’115 patent consistently explains that the “content
layer” of the disclosed ICL “consists of the specific goals, triggers, and data
elements that may be embedded within various events.” Id. at 9 (Ex. 1001,
11:13–15).
Petitioner also contends that FIPA97 teaches “a content layer
comprising one or more of goals . . . associated with the events” because in
FIPA97 “goals can be communicated among agents through the use of an
achieve domain-language primitive.” Pet. 42 (emphasis added) (citing
Ex. 1007, 81).
Petitioner asserts that FIPA97 also teaches “data elements associated
with the events” because FIPA ACL supports content expression through
objects and other data (“data elements”) associated with various
performatives. Pet. 43 (citing Ex. 1007, 14).
(1) Triggers
Petitioner asserts that a “trigger” recited in claim 1 is “a general
mechanism for requesting some action be taken when one or more
conditions is met.” Pet. 11–12 (citing Ex. 1001, 21:21–22:31; Ex. 1003
¶¶ 80–81). Patent Owner does not dispute Petitioner’s proposed meaning of
“trigger” (see PO Resp. 24–38) and argues that FIPA97 does not teach
IPR2019-00811
Patent 6,851,115 B1
117
triggers under Petitioner’s interpretation of the term (see id. at 107–108).
We adopt this undisputed interpretation of “trigger” for this Final Written
Decision because the construction is consistent with the disclosure in the
Specification. See Ex. 1001, 21:22–24 (“triggers [provide] a general
mechanism for requesting some action be taken when a set of conditions is
met”).
Petitioner asserts that FIPA97 discloses “triggers” for various
performative actions within the content layer, including for use with the
performative “accept-proposal” which “informs the receiver that it intends
that (at some point in the future) the receiving agent will perform the action,
once the given precondition is, or becomes, true.” Pet. 43 (emphasis added)
(citing Ex. 1007, 23). Petitioner argues that FIPA97 also discloses that the
performative “propose” can be used “to make a proposal or respond to an
existing proposal during a negotiation process by proposing to perform a
given action subject to certain conditions being true.” Id. (emphasis added)
(citing Ex. 1007, 34).
Petitioner argues that the ’115 patent describes “triggers” as
“providing a general mechanism for requesting some action be taken when a
set of conditions is met,” and that such triggers “preferably specif[y] at least
a condition and an action . . . [t]he condition indicates under what
circumstances the trigger should fire, and the action indicates what should
happen when it fires.” Pet. 11–12 (emphases added) (citing Ex. 1001,
21:21–34).
Patent Owner contends that the portions of FIPA97 cited by Petitioner
do not describe “a settable trigger that kicks in when a specified event takes
IPR2019-00811
Patent 6,851,115 B1
118
place.” PO Resp. 108. Patent Owner asserts that the communicative acts (or
performative actions) cited by Petitioner instead are “performatives that
contain parameters dictating what the associated action is.” Id. at 107 (citing
Ex. 2032 ¶¶ 168–169).
Patent Owner also cites 35 paragraphs from the Cohen Declaration in
support of its argument. Id. at 108 (citing Ex. 2033 ¶¶ 30–64). Patent
Owner, however, does not discuss Dr. Cohen’s testimony in the Patent
Owner Response or explain how Dr. Cohen’s testimony supports its
contention. See id. at 107–108. Thus, to the extent Petitioner purports to
rely on Dr. Cohen’s testimony, this amounts to improper incorporation by
reference in violation of 37 C.F.R. § 42.6(a)(3). Accordingly, we decline to
consider the cited paragraphs from the Cohen Declaration.
Petitioner argues that the performative “accept-proposal” “informs the
receiver that it intends that (at some point in the future) the receiving agent
will perform the action, once the given precondition is, or becomes, true.”
Pet. Reply 63 (citing Ex. 1007, 23). Petitioner cites the following testimony
from Dr. Lieberman:
[T]he “precondition” is sent to an agent through the “content
layer,” then at some future point when that “precondition” is met
the receiving agent will “trigger” and take an action based upon
the information encoded in the “content layer” of the message.
EX1007, 23.
Id. (citing Ex. 1003 ¶ 276). Pointing to the description in the cited portion of
FIPA97, Petitioner argues
[A]s explained by Dr. Lieberman, and shown in FIPA97, the
performative “accept-proposal” (“[a] general mechanism for
requesting”) will take an action—in this example, streaming a
multimedia channel (“some action be taken”)—when “the
IPR2019-00811
Patent 6,851,115 B1
119
customer is ready” (“when one or more conditions is met”).
Further, the example above shows how this trigger information
is stored within the “content layer.”
Id. at 64 (reproducing the description of the performative “accept-proposal”
in Ex. 1007, 23).
Petitioner contends that Dr. Lieberman and the Petition describe many
other possible triggers, including the “propose” performative and the
“request-when” performative. Id. at 65. Petitioner asserts that the “propose”
performative “informs the receiver that the proposer will adopt the intention
to perform the action once the given precondition is met” (id. (citing
Ex. 1007, 34; Ex. 1003 ¶¶ 277–279; Pet. 43)) and that the “request-when”
performative is used when “[t]he sender wants the receiver to perform some
action when some given proposition becomes true” (id. (citing Ex. 1007,
40; Ex. 1003 ¶ 414; Pet. 63–64)). Petitioner points to the description of the
“request-when” performative in FIPA97 and argues that the performative
“request-when” (“[a] general mechanism for requesting”) will take an
action—in this example, sending a message (“some action be taken”)—
when “an alarm occurs” (“when one or more conditions is met”). Id. at 66.
Petitioner concludes that FIPA97, therefore, discloses multiple “triggers”
recited in claim 1. Id.
Patent Owner does not dispute Petitioner’s responsive arguments in
the Reply, but instead contends that Petitioner’s arguments are presented
“too late” and that “Petitioner cannot fix the deficiencies of its Petition on
Reply.” PO Sur-reply 54. To the extent Patent Owner argues that
Petitioner’s arguments in the Reply is improper new arguments outside the
scope of a proper reply, we disagree. As discussed above, Petitioner’s
IPR2019-00811
Patent 6,851,115 B1
120
arguments in the Petitioner Reply on whether FIPA97 teaches the “triggers”
recited in the claim are directly responsive to Patent Owner’s arguments
raised in the Patent Owner Response. See Belden Inc. v. Berk-Tek LLC, 804
F.3d 1064, 1078–80 (Fed. Cir. 2015) (explaining that the Board may rely on
new evidence submitted with a reply because that evidence was responsive
to the arguments in patent owner’s response).
Based on the complete record, we determine that Petitioner has shown
sufficiently that FIPA97 teaches “triggers” recited in claim 1.
In sum, based on the complete record, we determine that Petitioner
has demonstrated sufficiently that FIPA97 teaches the claimed “expandable,
platform-independent, inter-agent language” including “a layer of
conversational protocol” and “a content layer,” notwithstanding these
arguments from Patent Owner.
d. Receiving a Request for Service in an Arbitrarily Complex Goal
Expression
The second step of claim 1 recites “receiving a request for service as a
base goal in the inter-agent language, in the form of an arbitrarily complex
goal expression.” Ex. 1001, 29:25–27. Petitioner relies on Kiss to teach the
“receiving a request for service” step and combines Kiss’s teaching with
FIPA97’s ACL to teach “a request for service . . . in the form of an
arbitrarily complex goal expression.” Pet. 43–47.
(i) Receiving a Request for Service
Petitioner asserts that Kiss teaches “receiving a request for service”
because the reference describes that a “user agent” receives a “query” from a
user, and then “parses the user inquiry and directs the parsed inquiry to a
IPR2019-00811
Patent 6,851,115 B1
121
metaagent 119 for action.” Pet. 44 (citing Ex. 1005, 8:28–34). Petitioner
points to the example process depicted in Figures 8–20 of Kiss as showing
how a “problem presented by a user is solved through distributed
inferencing.” Id. (citing Ex. 1005, 12:21–22). Petitioner asserts that the
initial request illustrated in Figure 8 (not reproduced herein) “when the user
asks of the user agent 705 ‘what is the effect of increasing sales by 20%’?
(FIG.8)” is a “base goal” because it is the basic request initiated by the user
that inspires a number of additional, more particular requests in order to be
satisfied. Id. at 44–45 (citing Ex. 1005, 12:23–26, Fig. 8).
(ii) Arbitrarily Complex Goal Expression
As discussed above in Section III.B. and the Institution Decision,
based on the definitions provided in the ’115 patent, we construe “arbitrarily
complex goal expression” to mean “a goal expressed in a language or syntax
that allows an expression, when appropriate or when desired, that expresses
multiple sub-goals and can potentially include more than one type of logical
connector (e.g., AND, OR, NOT), and/or more than one level of logical
nesting (e.g., use of parentheses), or the substantive equivalent, although not
every goal is itself necessarily complex.” Inst. Dec. 35–37. Petitioner
argues that this definition of “arbitrarily complex goal expression” does not
require multiple sub-goals to be expressed within the request, but instead
requires only that the goal be expressed in a language “that allows the
expression of multiple sub-goals . . . .” Pet. 45.
Petitioner asserts that FIPA ACL allows the expression of multiple
sub-goals, which can potentially include more than one type of logical
connector and/or more than one level of logical nesting (e.g., use of
IPR2019-00811
Patent 6,851,115 B1
122
parentheses), or the substantive equivalent, because FIPA97 “allows the
construction of s-expressions of arbitrary depth and complexity.” Id. (citing
Ex. 1007, 15). Petitioner argues that FIPA97 also discloses complex “goal
expressions” that “express multiple sub-goals” and which may also include
“logical nesting” using parentheses. Id. at 46 (citing Ex. 1007, 40, 41).
Petitioner contends that, in the combined system of Kiss and FIPA97,
the request discussed above— “What is the effect of increasing sales by
20%?”—is an “arbitrary complex goal expression” because it is a goal
expressed in FIPA ACL. Id.
Patent Owner does not specifically dispute that Kiss teaches this
“receiving a request for service” limitation. See generally PO Resp. Based
on the complete record and for the reasons explained by Petitioner, we are
persuaded that Petitioner has demonstrated sufficiently that Kiss teaches
“receiving a request for service as a base goal in the inter-agent language, in
the form of an arbitrarily complex goal expression,” as recited in claim 1.31
e. Dynamically Interpreting the Arbitrarily Complex Goal Expression
Claim 1 recites “dynamically interpreting the arbitrarily complex goal
expression,” the act of interpreting comprising “generating one or more
sub-goals expressed in the inter-agent language,” “constructing a goal
satisfaction plan,” and “dispatching each of the sub-goals to a selected client
agent for performance.” Ex. 1001, 29:28–44.
31 We also find that Patent Owner has waived any argument directed to this
claim limitation. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
123
Petitioner contends that Kiss teaches “dynamically interpreting the
arbitrarily complex goal expression,” as recited in claim 1. Pet. 47–49.
Petitioner argues
Kiss discloses that, in response to a request for service, the “meta
agent 119 formulates a goal statement for the problem-solving
phase of the process, then formulates a solution plan for the
problem” and “dynamically assesses the problem and its solution
states (between users and the knowledge agents 121), divides the
problem, and assigns the appropriate knowledge agents 121 to
work on the solution.”
Id. at 47–48 (emphasis added) (citing Ex. 1005, 8:32–48).
In addition, Petitioner explains how Kiss interprets the exemplary
“arbitrary complex goal expression” discussed above—“What is the effect of
increasing sales by 20%?”—including the meta agent formulating a plan to
respond to the request, identifying other agents and knowledge modules that
could assist in the response, and doing both dynamically as additional
information is obtained. Id. at 48 (citing Ex. 1005, 12:21–14:29, Figs. 8–
20). Petitioner further provides a detailed description of how this example
goal expression is dynamically interpreted in Kiss. Id. at 48–49 (citing
Ex. 1005, 12:29–14:30, Figs. 9–11, 13–20).
Patent Owner does not specifically dispute that Kiss teaches this
“dynamically interpreting the arbitrarily complex goal expression”
limitation. See generally PO Resp. Based on the complete record and for
the reasons explained by Petitioner, we are persuaded that Petitioner has
IPR2019-00811
Patent 6,851,115 B1
124
demonstrated sufficiently that Kiss teaches “dynamically interpreting the
arbitrarily complex goal expression,” as recited in claim 1.32
f. Generating One or More Sub Goals Expressed in the Inter-agent
Language
Petitioner contends that Kiss teaches “generating one or more
sub-goals expressed in the inter-agent language,” as recited in claim 1,
because Kiss discloses, in the example “What is the effect of increasing sales
by 20%?” discussed above, that the process of responding to the user’s
request includes several additional requests for service (“sub-goals”) to
various agents, including two sub-requests each from the sales agent and the
meta agent, and three sub-requests from the production agent, where each of
the sub-requests is made in FIPA ACL. Pet. 49–50 (citing Ex. 1005, 12:54–
56, 13:25–27, 13:37–39, 13:56–57, 14:3–5). According Petitioner, each of
these sub-requests is a “sub-goal” because each is a subset of the process of
responding to the user’s base request—“base goal”—of “what is the effect of
increasing sales by 20%?” Id. at 50. Petitioner also argues that Kiss’s meta-
agent is capable of “formulating a sub-problem query.” Id. (citing Ex. 1005,
11:15–16).
Patent Owner does not specifically dispute that Kiss teaches this
“generating one or more sub-goals expressed in the inter-agent language”
limitation. See generally PO Resp. Based on the complete record and for
the reasons explained by Petitioner, we are persuaded that Petitioner has
32 We also find that Patent Owner has waived any argument directed to this
claim limitation. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
125
demonstrated sufficiently that Kiss teaches “generating one or more
sub-goals expressed in the inter-agent language,” as recited in claim 1.33
g. Constructing a Goal Satisfaction Plan Including Delegation of
Sub-goal Requests
Claim 1 recites “constructing a goal satisfaction plan,” the goal
satisfaction plan including “a suitable delegation of sub-goal requests” and
“dispatching each of the sub-goals to a selected client agent for
performance.” Ex. 1001, 29:32–43. Petitioner contends that Kiss teaches
each of these limitations. Pet. 50–54.
(i) Goal Satisfaction Plan
Petitioner asserts that Kiss teaches “constructing a goal satisfaction
plan” because Kiss describes that the meta-agent “formulates a goal
statement for the problem-solving phase of the process,” then “formulates a
solution plan for the problem.” Pet. 50 (citing Ex. 1005, 8:33–34, 5:30–64,
12:25–40, Figs. 8–20). Petitioner contends that Figures 8–20 of Kiss
illustrate adding tasks to an agenda, i.e., constructing a plan, to satisfy the
user request (i.e., a “goal”) “what is the effect of increasing sales by 20%?”
Id. at 50–51 (citing Ex. 1005, 13:27–29, Figs. 8–20).
Patent Owner contends that Kiss does not teach the claim “goal
satisfaction plan” because Kiss does not disclose “the use of any reasoning,
optimization, or taking into account any advice parameters or constraints.”
PO Resp. 89. Patent Owner also argues that Kiss “does not have the
33 We also find that Patent Owner has waived any argument directed to this
claim limitation. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
126
capability of formulating an ‘optimal or near-optimal’ ‘goal satisfaction
plan’ utilizing reasoning as described in the ‘115 patent.” Id. at 91 (citing
Ex. 2032 ¶ 141). These arguments are predicated on Patent Owner’s
proposed construction of “goal satisfaction plan” to mean “a plan for the
satisfaction of a complex goal expression in an optimal or near-optimal
manner that is consistent with any advice parameters or constraints.” See
PO Resp. 28. As discussed above in Section III.B.1., we disagree with
Patent Owner that the claim term “goal satisfaction plan” is limited as Patent
Owner contends. Thus, for the reasons explained above in Section III.B.1.,
we disagree with Patent Owner’s argument that Kiss does not disclose the
“goal satisfaction plan” recited in claim 1.
Based on the complete record, we determine that Petitioner has shown
sufficiently that Kiss teaches “constructing a goal satisfaction plan,” as
recited in claim 1.
(ii) Delegation of Sub-goal Requests Using Reasoning
Claim 1 recites “a suitable delegation of sub-goal requests to best
complete the requested service request—by using reasoning that includes
one or more of domain-independent coordination strategies, domain-specific
reasoning, and application-specific reasoning comprising rules and learning
algorithms.” Ex. 1001, 29:32–39.
(1) Delegation of Sub-goal Requests
Petitioner asserts that Kiss teaches a meta-agent that makes “a suitable
delegation of sub-goal requests to best complete the requested service
request,” as recited in claim 1, because Kiss identifies and queries various
IPR2019-00811
Patent 6,851,115 B1
127
agents (e.g., sales agent, production agent) who best to respond to new
sub-requests for specific information as they arrive. Pet. 51 (citing
Ex. 1005, 12:21–14:30, Figs. 8–20). Petitioner contends that in Kiss, the
meta-agent “contains knowledge of problem solving methodologies and
distributed inferencing procedures” and is further responsible for allocating
“tasks to the knowledge agent layer.” Id. (citing Ex. 1005, 5:24–27, 3:26–
27). Petitioner asserts that Kiss’s meta-agent uses these methodologies in
conjunction with the agent registry “to dynamically bring available
knowledge resources together” (“suitable delegation”) by identifying “those
other resources capable of furthering the problem-solving process” (“best
complete the requested service request”). Id. (citing Ex. 1005, 5:27–29).
Patent Owner contends that Kiss does not teach “optimization of any
kind.” PO Resp. 89. Patent Owner asserts that Kiss’s meta agent employs a
trial and error method—it jumps into “executing the solution plan even
before the plan is complete” and when errors are encountered, it backtracks
or takes a new action. Id. at 90–91 (citing Ex. 1005 5:39–46). Patent Owner
argues that Kiss is “forced” to go through this trial-and-error process
because it does not have the capability of formulating an “optimal or near-
optimal” “goal satisfaction plan.” Id. at 91.
As discussed above in Section III.B., claim 1 does not recite an
“optimal or near-optimal” “goal satisfaction plan.” Instead, claim 1 recites
“a suitable delegation of sub-goal requests to best complete the requested
service request,” which may relate to the “optimal or near-optimal” limiting
feature argued by Patent Owner. As discussed in the same section, Patent
Owner asserts that the term “optimal” and “near optimal” has a
IPR2019-00811
Patent 6,851,115 B1
128
well-established meaning in the context of computer engineering, citing a
technical dictionary, which defines “optimization” as “[t]he process of
finding the best solution to some problem, where ‘best’ accords to prestated
criteria.” PO Sur-reply 6 (emphasis added) (citing Ex. 2078, 3). Thus,
contrary to Patent Owner’s argument, the trial and error process of Kiss is a
“process of finding the best solution” to a problem under Patent Owner’s
definition of “optimization.” See Ex. 1005, 5:21–29 (“The meta agent 119
contains knowledge of how to dynamically bring available knowledge
resources together as and when needed to provide a response to the
inquiry.”), 5:32–45 (“The meta agent 119 is responsible for formulating a
dynamic “solution plan” for the distributed inferencing to be performed by
the system 100, and allocates tasks to the knowledge agent layer 109 in
furtherance of the solution plan. The planning capability of the meta agent
119 includes the ability to formulate parallel sub-plans and perform iterative
and recursive procedures. . . . The meta agent 119 is capable of
backtracking or replanning to permit escape during plan execution from
dead-end or otherwise unproductive search paths.” (emphases added)); Pet.
51 (citing Ex. 1005, 5:24–29).
We note that claim 1 does not recite “to best complete the requested
service request at the very first try.” Thus, Kiss’s formulation of a dynamic
“solution plan” (and allocation of tasks) that includes “sub-plans” to perform
the trial and error process for finding the best solution to a problem satisfies
“delegation of sub-goal requests to best complete the requested service
request,” as recited in claim 1.
IPR2019-00811
Patent 6,851,115 B1
129
Patent Owner also argues that it explained during prosecution how the
facilitating engine is able to use reasoning to delegate the sub-goals to
service providing agents in such a way as “to best complete the requested
service request,” as follows:
For example, assume that several agents are able to roast coffee.
The facilitating engine is able to use reasoning to delegate the
sub-goal task of roasting coffee to the service-providing agent
that can roast beans in the least amount of time because the
facilitating engine has reasoned that the least amount of time
taken to make coffee is the best way to accomplish the base goal
of making coffee.
PO Resp. 90 (citing Ex. 1002, 173).
In response, Petitioner argues that Kiss similarly optimizes its solution
plan based on time. Pet. Reply 49–50. Petitioner asserts that Kiss explains
that as part of its matchmaking functionality (which is utilized in the
solution plan), the agent service layer will create “mappings for each
knowledge module 123 relating the capabilities to interests” and that these
mappings are used “to optimize problem solutions.” Id. at 49 (citing
Ex. 1005, 7:8–11). According to Petitioner, Kiss describes that when
utilizing “real time” inferencing, the meta-agent “may deactivate the
knowledge agent whose responses do not meet time delay” or prioritize
“knowledge agents that have a history of rapid response or that can tailor
their problem solving to real time.” Id. at 49–50 (citing Ex. 1005, 7:65–84).
We agree with Petitioner that Kiss’s delegation and prioritization of agents
based on their speed of response satisfies “delegation of sub-goal requests to
best complete the requested service request,” as recited in claim 1.
IPR2019-00811
Patent 6,851,115 B1
130
(2) Using Reasoning
Claim 1 recites that “a suitable delegation of sub-goal requests” is
accomplished “by using reasoning that includes one or more of
domain-independent coordination strategies, domain-specific reasoning, and
application-specific reasoning comprising rules and learning algorithms.”
Petitioner asserts that the ’115 patent describes speech recognition, natural
language processing, and email as “domain-independent technologies,” and
travel planning and reservations agents as “domain specific.” Pet. 52 (citing
Ex. 1001, 6:51–55).
Citing the testimony of Dr. Lieberman, Petitioner argues that the
operations disclosed in Kiss with respect to Figures 8–20 represent
“reasoning that includes one or more of domain-independent coordination
strategies” because the meta agent, by identifying and querying several
different specialized (i.e., domain-specific) agents, is carrying out a strategy
that is domain independent (i.e., is not limited to a particular area of
knowledge but instead seeks information regarding sales, production,
marketing, facilities, and materials) and coordinates the action of those
different domain-specific agents across several domains. Pet. 52 (citing
Ex. 1003 ¶¶ 331–332). Petitioner argues that Kiss’s meta-agent can employ
a distributed inferencing scheme that “assembles a problem-specific rule
network as a distributed object under control by a meta agent.” Id. at 52–53
(citing Ex. 1005, 7:21–26). Petitioner contends that application of such a
problem-specific set of rules to a request constitutes both “domain-specific
reasoning” (because the problem is the domain) and “application-specific
IPR2019-00811
Patent 6,851,115 B1
131
reasoning” (because the problem is the application). Id. at 53 (citing
Ex. 1003 ¶ 333).
We additionally note that the portion of Kiss cited by Petitioner
describes that distributed inferencing schemes are based on the use of
“first-order logic.” Ex. 1005, 7:21–23. The same paragraph also describes
using a “theorem-proving engine.” Id. at 7:26–28. We note that both
“first-order logic” and “theorem-proving engine” may be considered
“domain-independent technologies,” similar to speech recognition and
natural language processing technologies described in the ’115 patent.
Patent Owner contends that the “solution plan” of Kiss is not
constructed by “using reasoning” because Kiss’s inferencing schemes are
used only “[a]fter the solution plan is formulated . . . to perform the search
and execution phases.” PO Resp. 89 (citing Ex. 1005, 8:58–60). In
response, Petitioner asserts that in the Petition it argued that it would have
been obvious to utilize inferencing strategies when constructing the goal
satisfaction plan. Pet. Reply 48 (citing Pet. 53). Thus, Patent Owner’s
argument is inapposite because it does not address the combination proposed
by Petitioner. See 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.”).
As Petitioner also persuasively argues, Petitioner in the Petition
additionally relies on several other ways in which Kiss utilizes reasoning to
construct its solution plan, such as “problem solving methodologies and
distributed inferencing procedures,” how the meta-agent “assigns the
appropriate knowledge agents 121 to work on the solution,” and how the
IPR2019-00811
Patent 6,851,115 B1
132
meta-agent “executes the solution plan by maintaining an agenda,
commitment table, task queue, knowledge manager or equivalent dynamic
control service.” Pet. Reply 48 (citing Pet. 50–51; Ex. 1003 ¶¶ 322–325,
331–335). Petitioner further argues that Kiss’s “iterative and recursive”
planning is another example of how reasoning is used to construct the
solution plan of Kiss. Id. (citing Pet. 50; Ex. 1003 ¶ 322).
Based on the complete record, we determine that Petitioner has shown
sufficiently that Kiss teaches “a suitable delegation of sub-goal requests to
best complete the requested service request—by using reasoning that
includes one or more of domain-independent coordination strategies,
domain-specific reasoning, and application-specific reasoning comprising
rules and learning algorithms,” as recited in claim 1, notwithstanding the
arguments by Patent Owner.
h. Dispatching Each of the Sub-goals
Claim 1 recites “dispatching each of the sub-goals to a selected client
agent for performance, based on a match between the sub-goal being
dispatched and the registered functional capabilities of the selected client
agent.” Ex. 1001, 29:32–43.
Petitioner asserts that in Kiss
the meta agent dispatches queries (“sub-goals”) to various
specialized agents based on a determination that the targeted
agent possesses a knowledge module appropriate for the query
(“based on a match between the sub-goal being dispatched and
the registered functional capabilities of the selected client
agent.”)
IPR2019-00811
Patent 6,851,115 B1
133
Pet. 54 (citing Ex. 1005, 12:21–14:29, Figs. 8–20). Petitioner argues that
Kiss’s specialized agents are not facilitators (and, therefore, are “client
agents”) because they are not responsible for matching requests, from users
and agents, with descriptions of the capabilities of other agents. Id. (citing
Ex. 1001, 4:64–66, 6:41–43).
Patent Owner does not specifically dispute that Kiss teaches this
“dispatching each of the sub-goals” limitation. See generally PO Resp.
Based on the complete record and for the reasons explained by Petitioner,
we are persuaded that Petitioner has demonstrated sufficiently that Kiss
teaches “dispatching each of the sub-goals to a selected client agent for
performance, based on a match between the sub-goal being dispatched and
the registered functional capabilities of the selected client agent,” as recited
in claim 1.34
i. Conclusion
In consideration of the foregoing and based on the complete record,
we are persuaded by Petitioner’s arguments and evidence, notwithstanding
Patent Owner’s arguments, addressed above. Having weighed each of the
Graham factors, including the scope and content of the prior art, the
differences between the prior art and the challenged claim, and the objective
evidence of nonobviousness, we determine that Petitioner has demonstrated
by a preponderance of the evidence that claim 1 is unpatentable under
35 U.S.C. § 103(a) as obvious over the combination of Kiss and FIPA97.
34 We also find that Patent Owner has waived any argument directed to this
claim limitation. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
134
3. Dependent Clams 26–28
Claims 26–28 depends directly or indirectly from claim 1. Petitioner
asserts that FIPA97 teaches all of the additionally recited limitations of these
dependent claims. Pet. 54–57. Petitioner provides detailed explanations and
specific citations to FIPA97 indicating where in the reference each of the
additionally recited limitations of claims 26–28 is disclosed. Id. Petitioner
also relies on the testimony of Dr. Lieberman to support its position. Id.
Addressing claim 26, Petitioner asserts that FIPA97v1 discloses base
goals expressed as “a compound goal having sub-goals separated by
operators” and that it would have been obvious to include a base goal that
“is a compound goal having sub-goals separated by operators” in the system
of Kiss and FIPA97 based on the teachings of FIPA97 and, alternatively, on
the teachings of Kiss. Pet. 54–55 (citing Ex. 1003 ¶¶ 306–313, 477–479).
With respect to claim 27, Petitioner contends that FIPA97 discloses
operators for composing base goals including a “conjunction operator” and
the operator formula “(and )” which in FIPA97 is
defined as a “Conjunction.” Pet. 55 (citing Ex. 1007, 73; Ex. 1034, 112–
113; Ex. 1003 ¶¶ 481–483). Petitioner further asserts that FIPA97 discloses
a “disjunction operator” and the operator formula “(or
),” which in FIPA97 is defined as “Disjunction.” Id. (citing
Ex. 1007, 73; Ex. 1003 ¶¶ 484–485).
Regarding claim 28, Petitioner asserts as follows:
FIPA97 discloses a “parallel disjunction operator” in the form of
an operator that is able to query multiple disjunct agents in a
single expression. FIPA97’s disclosed content language (SL2)
supports grammars for “quantifying-in inside modal operators”.
EX1007, 78. In the disclosed example, an agent is asking for all
IPR2019-00811
Patent 6,851,115 B1
135
x1’s that are true within agent “i” or agent “j.” EX1007, 78. For
example, if one wanted to know if (x1) was a train or Metro
station in D.C.; and agent “i” had a list of train stations in D.C.,
and agent “j” had a list of Metro stations in D.C., the above query
would cause both agents to query their internal databases. The
result of each query would be true or false depending on whether
x1 appeared each respective list. Therefore, agent “i” and “j”
(“by different agents”) are considering “disjunct goals” (i.e., goal
“(p ?x1)” versus “(q ?x1)”) within the context of a “disjunction
operator” (i.e., “or”). Thus, FIPA97v1 discloses “a parallel
disjunction operator that indicates that disjunct goals are to be
performed by different agents.”
Pet. 56–57.
Patent Owner does not respond specifically to Petitioner’s challenge
to claims 26–28 beyond Patent Owner’s arguments advanced with respect to
claim 1 discussed above.
Based on the complete record and for the reasons explained by
Petitioner, we determine that Petitioner has demonstrated by a
preponderance of the evidence that claims 26–28 are unpatentable under
35 U.S.C. § 103(a) over the combination of Kiss and FIPA97.35
4. Independent Claim 29
Although many of the limitations recited in claims 1 and 29 are
similar, there exist material differences in the claim language of these
claims. For example, claim 1 recites
constructing a goal satisfaction plan wherein the goal satisfaction
plan includes:
35 We also find that Patent Owner has waived any argument directed to
claims 26–28, aside from contesting the unpatentability of claim 1, over the
combination of Kiss and FIPA97. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
136
a suitable delegation of sub-goal requests to best complete the
requested service request—by using reasoning that includes
one or more of domain-independent coordination strategies,
domain-specific reasoning, and application-specific
reasoning comprising rules and learning algorithms; and
dispatching each of the sub-goals to a selected client agent for
performance, based on a match between the sub-goal being
dispatched and the registered functional capabilities of the
selected client agent,
whereas claim 29 recites
constructing a base goal satisfaction plan including the sub-acts
of:
determining whether the request service is available,
determining sub-goals required in completing the base goal by
using reasoning that includes one or more of domain-
independent coordination strategies, domain-specific
reasoning, and application-specific reasoning comprising
rules and learning algorithms,
selecting service-providing electronic agents from the agent
registry suitable for performing the determined sub-goals,
and,
ordering a delegation of sub-goal requests complete the
requested service.
Ex. 1001, 29:32–43 (claim 1), 32:9–22 (claim 29). As apparent in the claim
language reproduced above, there exist material differences between these
limitations on the face of the claims.
When addressing the “constructing a base goal satisfaction plan”
limitation and its elements recited in claim 29, however, Petitioner’s analysis
consists of pointing to its analysis for similar but materially different
limitations from claim 1. See Pet. 61. Petitioner does not address the claim
language of claim 29 or explain why its analysis for claim 1 is sufficient to
IPR2019-00811
Patent 6,851,115 B1
137
demonstrate a reasonable likelihood of prevailing with respect to claim 29.
Accordingly, in the Institution Decision, we determined that the information
presented in the Petition did not demonstrate a reasonable likelihood of
Petitioner prevailing in its challenge to claim 29 under 35 U.S.C. § 103(a) as
obvious over the combination of Kiss and FIPA97. Inst. Dec. 78–79.
In the Petitioner Reply, Petitioner presents comparisons of the claim
language of claim 1 and claim 29 and argues the combination of Kiss and
FIPA97 teaches the limitations of claim 29. Pet. Reply 66–74.
Patent Owner asserts that Petitioner’s arguments relating to claim 29
are new arguments presented for the first time in the Reply, and, therefore,
should not be considered. PO Resp. 54–58.
We agree with Patent Owner. It was Petitioner’s burden to
demonstrate sufficiently in the Petition that the cited prior art renders the
challenged claims unpatentable, including showing that the Petition’s
contentions are supported by evidence. See 35 U.S.C. § 314(a); see also
Harmonic Inc. v. Avid Tech., Inc., 815 F.3d 1356, 1363 (Fed. Cir. 2016) (“In
an IPR, the petitioner has the burden from the onset to show with
particularity why the patent it challenges is unpatentable.” (emphases
added)) (citing 35 U.S.C. § 312(a)(3)); Intelligent Bio-Systems, Inc. v.
Illumina Cambridge Ltd., 821 F.3d 1359, 1369 (Fed. Cir. 2016) (requiring
“the initial petition identify ‘with particularity’ the ‘evidence that supports
the grounds for the challenge to each claim.’” (emphases added)). It is
improper to present new arguments or evidence in a reply that should have
been presented and developed in the first instance in the Petition. Under the
IPR2019-00811
Patent 6,851,115 B1
138
particular facts and circumstances in this case, we decline to consider
Petitioner’s argument in the Reply that was not present in the Petition.
Accordingly, Petitioner does not demonstrate by a preponderance of
evidence that claim 29 is unpatentable under 35 U.S.C. § 103(a) over the
combination of Kiss and FIPA97.
5. Dependent Claims 33 and 45–47
Claims 33 and 45–47 depend directly or indirectly from claim 29.
Petitioner’s arguments and evidence presented with respect to dependent
claims 33 and 45–47 only address the additionally recited limitations of
these claims, and, therefore, do not remedy the deficiencies in Petitioner’s
analysis of independent claim 29 discussed above. See Pet. 54–57, 61–63.
Therefore, for the same reasons discussed above with respect to
claim 29, Petitioner does not demonstrate by a preponderance of evidence
that claims 33 and 45–47 are unpatentable under 35 U.S.C. § 103(a) over the
combination of Kiss and FIPA97.
6. Independent Claim 86
Claim 86 recites a preamble as follows: “A data wave carrier
providing a transport mechanism for information communication in a
distributed computing environment having at least one facilitator agent and
at least one active client agent, and an Interagent Communication Language
(ICL).” Ex. 1001, 37:10–14. In the body of the claim, claim 86 recites
substantially the same limitations as claim 1, such as the limitations that
recite that the ICL includes a “layer of conversational protocol” and a
“content layer” (id. at 37:14–20) and that the facilitator agent constructs “a
goal satisfaction plan by using reasoning” (id. at 37:21–38:5). In addition,
IPR2019-00811
Patent 6,851,115 B1
139
claims 86 recites that the data wave carrier comprises “a signal
representation of an inter-agent language description of an active client
agent’s functional capabilities.” Id. at 38:5–7.
Claim 86 recites “wherein said at least one facilitator agent is
operable to construct a goal satisfaction plan.” Id. at 37:21–22 (emphasis
added). Petitioner’s analysis of this limitation consists of a single
sentence—“This claim element is satisfied for the same reasons described
above in §VI.1.h.,” which refers to Petitioner’s analysis of the limitation of
claim 1 that recites “constructing a goal satisfaction plan.” Claim 1,
however, does not recite “facilitator agent,” nor does Petitioner explain in
the context of its analysis of claim 1 how the combination of Kiss and
FIPA97 teaches “at least one facilitator agent is operable to construct a goal
satisfaction plan.” See Pet. 32–54.
As discussed above with respect to claim 29, it was Petitioner’s
burden to demonstrate sufficiently in the Petition that the cited prior art
renders the challenged claims unpatentable, including showing that the
Petition’s contentions are supported by evidence. See 35 U.S.C. § 314(a);
see also Harmonic, 815 F.3d at 1363 (requiring “the initial petition identify
‘with particularity’ the ‘evidence that supports the grounds for the challenge
to each claim.’” (emphases added); Intelligent Bio-Systems, 821 F.3d at 1369
(requiring “the initial petition identify ‘with particularity’ the ‘evidence that
supports the grounds for the challenge to each claim.’” (emphases added)).
Accordingly, Petitioner does not demonstrate by a preponderance of
evidence that claim 86 is unpatentable under 35 U.S.C. § 103(a) over the
combination of Kiss and FIPA97.
IPR2019-00811
Patent 6,851,115 B1
140
7. Dependent Claims 87–89
Claims 87–89 depend directly or indirectly from claim 86.
Petitioner’s arguments and evidence presented with respect to dependent
claims 87–89 only address the additionally recited limitations of these
claims, and, therefore, do not remedy the deficiencies in Petitioner’s analysis
of independent claim 86 discussed above. See Pet. 66–68.
Therefore, for the same reasons discussed above with respect to
claim 86, Petitioner does not demonstrate by a preponderance of evidence
that claims 87–89 are unpatentable under 35 U.S.C. § 103(a) over the
combination of Kiss and FIPA97.
I. Obviousness over Kiss, FIPA97, and Moran
Claims 12–14 and 35–37 depend from claims 1 or 29. Petitioner adds
the teachings of Moran (Ex. 1013) to the basic combination of Kiss and
FIPA97 in an asserted ground of obviousness as to these dependent claims.
Pet. 68–76.
1. Moran (Ex. 1013)
Moran is titled “Multimodal User Interfaces in the Open Agent
Architecture” and describes “multimodal agent-based user interfaces”
implemented under SRI International’s Open Agent Architecture (OAA),
which is mentioned in the ’115 patent as a starting point for the subject
matter described in the patent. Ex. 1013, 6136 (Abstract); Ex. 1001, 4:44–55.
36 The page numbers for Exhibit 1013 refer to the page numbers inserted by
Petitioner in the bottom, right-hand corner of each page.
IPR2019-00811
Patent 6,851,115 B1
141
In Moran, users can “enter commands with a mix of modalities, for
example, a spoken command in which the object to be acted on is identified
by a pen gesture (or other graphical pointing operation),” because the system
“supports speech, handwriting and pen-based gestures in addition to the
conventional keyboard and mouse inputs.” Ex. 1013, 62. Moran discloses
that existing speech recognition systems can be used with its agent-based
applications, once an agent has been created to interface with the system. Id.
Moran also discloses that “[a] major advantage of using an agent-based
architecture is that it provides simple mix-and-match for the components,”
and describes existing natural language systems that can be used
interchangeably. Id. Moran discloses a User Interface (UI) Agent, which
“manages the various modalities and applies additional interpretation to
those inputs as needed,” along with a speech recognition agent and a natural
language understanding agent. Id.
Petitioner asserts that Moran was published in 1997 in conjunction
with the International Conference on Intelligent User Interfaces that took
place on January 6–9, 1997. Pet. 27 (citing Ex. 1013, Cover-2 (evidencing
receipt by libraries on March 1997)). Petitioner asserts that Moran was
republished later in March 1998 in Volume 10, Issue 5 of a journal titled
“Knowledge-Based Systems.” Id. at 27–28 (citing Ex. 1111, cover
(evidencing receipt by libraries by March 1998)). Petitioner contends that
Moran, therefore, is prior art to the ’115 patent under 35 U.S.C. §§ 102(a)
and (b). Id. at 28.
Patent Owner does not contest the applicability of Moran as prior art
in this proceeding. See generally PO Resp.
IPR2019-00811
Patent 6,851,115 B1
142
We note that Moran was published by the Association for Computing
Machinery (ACM) and bears the Copyright date of 1997 from ACM.
Ex. 1013, Cover-2. Like IEEE publications, ACM publications, such as the
one in which Moran appeared, are distributed widely and intended to be
accessible to the public. The wide distribution of publications from IEEE
and ACM distinguishes them from academic papers such as masters and
doctoral theses, which in some cases may not be well catalogued or indexed.
The Board has accepted the copyright date of an IEEE publication as
evidence of its public availability. Ericsson, Inc. v. Intellectual Ventures I
LLC, Case IPR2014-00527, Paper 41 at 10 (PTAB May 18, 2015). In view
of the reliability of ACM publications and their wide distribution, as well as
the evidence provided by Petitioner that the publication containing the
Moran article was provided to libraries for circulation to the public,
Petitioner has demonstrated sufficiently that Moran qualifies as applicable
prior art.
2. Motivation to Combine
Petitioner contends that a person of ordinary skill in the art would
have been motivated to add the teachings of Moran to the basic combination
of Kiss and FIPA97 because Moran is analogous art to Kiss, FIPA97, and
the ’115 patent and is also reasonably pertinent to the problem addressed by
Kiss, FIPA97 and the 115 Patent—e.g., the need for “flexible, fault-tolerant,
distributed problem solving” using “agent-based technologies.” Pet. 68
(citing Ex. 1003 ¶¶ 221, 391).
IPR2019-00811
Patent 6,851,115 B1
143
Petitioner asserts that
It would have been obvious to combine the language processing
functionality of Moran in the system of Kiss/FIPA97, because to
do so would have been the arrangement of old elements (i.e., the
multimodal agent-based language processing of Moran, the
functionality of FIPA97, and the system of distributed
cooperative agents of Kiss) with each performing the same
function it had been known to perform (multimodal interface
supporting non-ICL requests (Moran); implementing
communication between distributed agents (FIPA97V1);
distributed agents conducting cooperative task competition and
problem solving (Kiss)) and yielding no more than what one
would expect from such an arrangement (a system of distributed
agents, able to communicate to conduct cooperative task
completion and problem solving, including multimodal user
interface support), as Kiss demonstrates.
Pet. 67 (citing Ex. 1020, 2; Ex. 1027, 50; Ex. 1003 ¶ 392).
Patent Owner does not dispute Petitioner’s articulation of the
reasons to combine Kiss, FIPA97, and Moran. See generally PO
Resp.
Based on the complete record, we determine Petitioner has
provided sufficient reasons why a person of ordinary skill in the art
would have combined Kiss, FIPA97, and Moran.
3. Dependent Claims 12–14
Claims 12–14 depend directly or indirectly from claims 1. Petitioner
asserts that Moran teaches all of the additionally recited limitations of
dependent claims 12–14. Pet. 71–75. Petitioner provides detailed
explanations and specific citations to Moran indicating where in the
reference each of the additionally recited limitations of claims 12–14 are
IPR2019-00811
Patent 6,851,115 B1
144
disclosed. Id. Petitioner also relies on the testimony of Dr. Lieberman to
support its position. Id.
a. Claim 12
Claim 12 recites “receiving a request for service in a second language
differing from the inter-agent language.” Petitioner relies on “UI Agent” of
Moran to teach this limitation. Pet. 71–72. Petitioner contends that
The UI Agent receives “speech or pen-based input” which is then
routed to a “recognition” agent that produce raw text (“second
language differing from the inter-agent language”). Once the raw
text is produced it is then sent to a “natural language
understanding agent” that is able to produce a “logical form
representation” of the request. Therefore, an agent receives “a
request for service” (“user's request”) in a “second language
differing from the inter-agent language” (“speech or pen-based
input”; raw “text”).
Id. (citing Ex. 1013, 62; Ex. 1003 ¶ 388).
Addressing “selecting a registered agent capable of converting the
second language into the inter-agent language,” as recited in claim 12,
Petitioner asserts that
Moran, further discloses “delegation of the request to the
appropriate agents” based off that agent’s registered capabilities.
When an agent is added to the system in Moran, that agent
“registers its capabilities with the Facilitator.” As part of that
registration the agent also registers any “natural language
vocabulary that can be used to talk about the tasks that the agent
can perform.” (“registered agent capable of converting the
second language into the inter-agent language”). As with the
cited combination of Kiss and FIPA ACL, when an agent needs
a task completed “it sends a request to the Facilitator, which then
delegates it to an agent, or agents, that have registered that they
can handle the needed tasks.” For example, a user might query,
IPR2019-00811
Patent 6,851,115 B1
145
“Where will I be at 2:00 this afternoon?” EX1013, 66 (emphasis
in original). In order to resolve this request the “UI Agent” will
work “in concert with a Facilitator agent” to invoke a “speech
recognition agent or handwriting recognition agent” and thus
delegate the task to the “appropriate agents” based off their
registered capabilities. Therefore, an agent is “selected” to
convert a “second language into the inter-agent language” based
off that agent’s registered capabilities.
Pet. 72–73 (citing Ex. 1013, 62–66).
Next, addressing “selecting a registered agent capable of converting
the second language into the inter-agent language,” as recited in claim 12,
Petitioner contends as follows:
Moran, further discloses “delegation of the request to the
appropriate agents” based off that agent’s registered capabilities.
EX1013, 66. When an agent is added to the system in Moran,
that agent “registers its capabilities with the Facilitator.”
EX1013, 64. As part of that registration the agent also registers
any “natural language vocabulary that can be used to talk about
the tasks that the agent can perform.” (“registered agent capable
of converting the second language into the inter-agent language”)
EX1013, 64. As with the cited combination of Kiss and FIPA
ACL, when an agent needs a task completed “it sends a request
to the Facilitator, which then delegates it to an agent, or agents,
that have registered that they can handle the needed tasks.”
EX1013, 64. For example, a user might query, “Where will I be
at 2:00 this afternoon?” EX1013, 66 (emphasis in original). In
order to resolve this request the “UI Agent” will work “in concert
with a Facilitator agent” to invoke a “speech recognition agent
or handwriting recognition agent” and thus delegate the task to
the “appropriate agents” based off their registered capabilities.
EX1013, 62-66. Therefore, an agent is “selected” to convert a
“second language into the inter-agent language” based off that
agent’s registered capabilities. EX1003,¶399.
Pet. 72–73.
IPR2019-00811
Patent 6,851,115 B1
146
Turning next to claim 12’s limitation that recites “forwarding the
request for service in a second language to the registered agent capable of
converting the second language into the inter-agent language,” Petitioner
asserts that Moran also discloses that when a “UI Agent” detects “speech
input” it will “sends a command to the Speech Recognition agent to process
the audio input and to return the corresponding text.” Pet. 73. Petitioner
further contends that
As disclosed above the delegation of the task to a specific
“Speech Recognition agent” is based on that agent’s registered
capabilities. After the “UI Agent” determines that translation is
necessary and “selects” an appropriate agent the task is then
“forwarded” to the appropriate agent. The “UI Agent sends a
command” (“forwarding”) to the translation agent (“registered
agent capable of converting the second language into the inter-
agent language”), then the “audio input” (“request for service in
a second language”) is processed and the text is returned to the
“UI Agent.
Pet. 73 (citing Ex. 1013, 62.).
Addressing next claim 12’s limitation that recites “implicitly
requesting that such a conversion be performed and the results returned,”
Petitioner asserts as follows:
Moran discloses a UI Agent “implicitly” requesting a
“conversion be performed and the results returned.” When the
“UI Agent” detects “speech or pen-based input” the UI Agent
“invokes a speech recognition agent or handwriting recognition
agent.” The corresponding text is then sent to a “natural
language understanding agent” this agent “produces a logical
form representation of the user's request” (“conversion be
performed and the results returned.”). (emphasis in original).
Moran, further shows that the user interface agents can work “in
concert with a Facilitator agent” to delegate task to an
IPR2019-00811
Patent 6,851,115 B1
147
“appropriate agent” in a “less direct and more anonymous”
manner (“implicitly”).
Pet. 74 (citing Ex. 1013, 62, 66).
Patent Owner does not respond specifically to Petitioner’s challenge
to claim 12 beyond Patent Owner’s arguments advanced with respect to
claim 1 discussed above. See generally PO Resp.
Based on the complete record and for the reasons explained by
Petitioner, we determine that Petitioner has demonstrated by a
preponderance of the evidence that claim 12 is unpatentable under 35 U.S.C.
§ 103(a) over the combination of Kiss, FIPA97, Moran.37
b. Claims 13 and 14
Petitioner contends that Moran discloses a request that includes “a
natural language query” because
Moran shows a UI Agent which receives a request for service in
the form of a natural language question; “Where will I be at 2:00
this afternoon?” EX1013, 66 (emphasis in original). The UI
Agent is then able to work with the Facilitator Agent to delegate
the request to the “appropriate agents.” EX1013, 66. Therefore,
the UI Agent is capable of receiving and processing “a natural
language query.” EX1003,¶406.
Pet. 74. According to Petitioner, Moran further discloses that when the UI
Agent receives natural language text, such as through “speech or pen-based
input” it may invoke a “natural language understanding agent” (“natural
language agent”) that “produces a logical form representation of the user's
37 We also find that Patent Owner has waived any argument directed to
claim 12, aside from contesting the unpatentability of claim 1, over the
combination of Kiss, FIPA97, and Moran. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
148
request.” (“second registered agent capable of converting the second
language into the inter-agent language”). Pet. 75 (citing Ex. 1013, 62;
Ex. 1003 ¶ 407).
Turning next to claim 14, Petitioner asserts that
Moran discloses a “user interface agent” that generates a “natural
language query” in the form of a UI agent that can accept
“spoken and pen-based inputs.” The “spoken or handwritten
inputs” can be treated as raw text or it may be “interpreted by a
natural language understanding agent.” For example, a user can
ask “Where will I be at 2:00 this afternoon?” (“natural language
query”) this question can then be treated as raw text or processed
by a “natural language understanding agent” to produce “a
logical form representation of the user's request.” Therefore, a
“user interface agent” generates a “natural language query” when
it accepts the user’s question, which is phrased in simple English.
Pet. 75 (citing Ex. 1013, 62).
Patent Owner does not respond specifically to Petitioner’s challenge
to claims 13 and 14 beyond Patent Owner’s arguments advanced with
respect to claim 1 discussed above. See generally PO Resp.
Based on the complete record and for the reasons explained by
Petitioner, we are persuaded that Petitioner has demonstrated by a
preponderance of the evidence that claims 13 and 14 are unpatentable under
35 U.S.C. § 103(a) as obvious over the combination of Kiss, FIPA97, and
Moran.38
38 We also find that Patent Owner has waived any argument directed to
claims 13 and 14, aside from contesting the unpatentability of claim 1, over
the combination of Kiss, FIPA97, and Moran. See Paper 16, 7.
IPR2019-00811
Patent 6,851,115 B1
149
4. Dependent Claims 35–37
Claims 35–37 depend directly or indirectly from claim 29.
Petitioner’s arguments and evidence presented with respect to these
dependent claims only address the additionally recited limitations of the
claims, and, therefore, do not remedy the deficiencies in Petitioner’s analysis
of independent claim 29 discussed above. See Pet. 75–76.
Therefore, for the same reasons discussed above with respect to
claim 29, the information presented in the Petition does not demonstrate by a
preponderance of the evidence that claims 35–37 are unpatentable under
35 U.S.C. § 103(a) as obvious over the combination of Kiss, FIPA97, and
Moran.
J. Patent Owner’s Motion to Exclude
Patent Owner filed a Motion to Exclude (Paper 35) certain “a portion
of Exhibit 1129, as well as related testimony relied on by [Petitioner].”
Paper 35, 1. Exhibit 1129 is the deposition transcript of Patent Owner’s
declarant Dr. Medvidovic, taken on May 19, 2020. In particular, Patent
Owner seeks to exclude lines 53:19–54:21 of Exhibit 1129, which contain
certain deposition testimony of Dr. Medvidovic regarding the “facilitator” of
the ’115 patent. We, however, do not rely on lines 53:19–54:21 of
Exhibit 1129 or related testimony of Dr. Medvidovic for this Final Written
Decision because whether the prior art teaches the claimed “facilitator
agent” is not at issue for this Final Written Decision. Accordingly, we
dismiss Patent Owner’s Motion to Exclude as moot.
IPR2019-00811
Patent 6,851,115 B1
150
K. Patent Owner’s Constitutional Challenges
Patent Owner raises two constitutional challenges. First, Patent
Owner argues that subjecting a pre-AIA patent, such as the ’115 patent,
retroactively to inter partes review violates the Takings and Due Process
Clauses of the Fifth Amendment. PO Resp. 146. Second, Patent Owner
asserts that inter partes reviews violate the Appointments Clause of the U.S.
Constitution when conducted by administrative patent judges not nominated
by the President and confirmed by the Senate. Id.
Addressing first Patent Owner’s Appointments Clause challenge, we
are bound by the Federal Circuit’s decision in Arthrex, Inc. v. Smith &
Nephew, Inc., 941 F.3d 1320 (Fed. Cir. 2019), which resolved this issue.
See 941 F.3d at 1337 (“This as-applied severance . . . cures the constitutional
violation.”); see also Arthrex, Inc. v. Smith & Nephew, Inc., 953 F.3d 760,
764 (Fed. Cir. 2020) (Moore, J., concurring in denial of rehearing)
(“Because the APJs were constitutionally appointed as of the
implementation of the severance, inter partes review decisions going
forward were no longer rendered by unconstitutional panels.”).
Accordingly, we do not consider this issue any further.
With regard to the Takings and Due Process Clause challenge, we
note that challenges to retroactive application of IPRs to pre-AIA patents
have been addressed by the Federal Circuit in Celgene Corp. v. Peter, 931
F.3d 1342, 1357–1363 (Fed. Cir. 2019), cert. denied 2020 WL 3405867
(June 22, 2020) (Takings Clause) and Sound View Innovations, LLC v. Hulu,
LLC, Nos. 2019-1865, 2019-1867, 2020 WL 3583556, *3 (Fed. Cir. July 2,
IPR2019-00811
Patent 6,851,115 B1
151
2020) (non-precedential) (Due Process Clause). Accordingly, we do not
consider this issue any further.
IV. CONCLUSION
For the foregoing reasons, we conclude that Petitioner has met its
burden of proof, by a preponderance of the evidence, in showing that
claims 1, 12–14, and 26–28 of the ’115 patent are unpatentable. For the
reasons discussed above, Petitioner has not demonstrated, by a
preponderance of the evidence, claims 29, 33, 35–37, 45–47, and 86–89 of
the ’115 patent are unpatentable. The chart below summarizes our
conclusions.
Claims
35 U.S.C.
§
References/
Basis
Claim(s)
Shown
Unpatentable
Claim(s)
Not Shown
Unpatentable
1, 26–29,
33, 45–47,
86–89
103(a)
Kiss and
FIPA97
1, 26–28
29, 33, 45–47,
86–89
12–14,
35–37
103(a)
Kiss, FIPA97,
and Moran
12–14 35–37
Overall
Outcome
1, 12–14, 26–
28
29, 33, 35–37,
45–47, 86–89
V. ORDER
In consideration of the foregoing, it is hereby:
ORDERED that claims 1, 12–14, and 26–28 of the ’115 patent are
determined to be unpatentable;
IPR2019-00811
Patent 6,851,115 B1
152
FURTHER ORDERED that claims 29, 33, 35–37, 45–47, and 86–89
of the ’115 patent are not determined to be unpatentable;
FURTHER ORDERED that Patent Owner’s Motion to Exclude is
dismissed as moot; and
FURTHER ORDERED that, because this is a Final Written Decision,
a party to the proceeding seeking judicial review of the Decision must
comply with the notice and service requirements of 37 C.F.R. § 90.2.
IPR2019-00811
Patent 6,851,115 B1
153
PETITIONER:
Joseph Micallef
Scott Border
jmicallef@sidley.com
sborder@sidley.com
PATENT OWNER:
Steven Hartsell
Alexander Gasser
Sarah Spires
shartsell@skiermontderby.com
agasser@skiermontderby.com
sspires@skiermontderby.com