1286 757 federal reporter, 3d series - … › wp-content › uploads › 2016 › 05 › ...1286...

1286 757 FEDERAL REPORTER, 3d SERIES

active); see also Simpson v. United States,721 F.3d 875, 876 (7th Cir.2013) (conclud-ing that Alleyne is not retroactive becauseApprendi is not retroactive). Finally, ev-ery one of our sister circuits to considerthe issue in a published decision has con-cluded that Alleyne is not retroactivelyapplicable on collateral review. See Unit-ed States v. Winkelman, 746 F.3d 134, 136(3d Cir.2014); In re Kemper, 735 F.3d 211,212 (5th Cir.2013); Simpson, 721 F.3d at876. We therefore conclude that Jeantyhas not met Bryant ’s third requirement.

[4] Jeanty also fails to satisfyBryant ’s fourth requirement, which is thata retroactive application of Alleyne wouldmean that his current sentence exceedsthe statutory maximum authorized by Con-gress for his crime. See Bryant, 738 F.3dat 1274. He was sentenced to ten years(120 months) in prison, but 21 U.S.C.§ 841(b)(1)(B)(ii) provides that the maxi-mum sentence for Jeanty, who was con-victed of conspiring to possess with intentto distribute at least 500 grams of cocaine,is forty years. Ten is less than forty.Thus even if Alleyne applied retroactively,Jeanty’s current sentence would not ex-ceed the statutory maximum authorized byCongress in § 841(b)(1)(B).

[5] Finally, even assuming that Jeantycould satisfy all five of Bryant ’s require-ments, his petition is without merit. Heargues that he could not receive the man-datory minimum unless the jury—not thejudge—found that his 1997 conviction qual-ified as ‘‘a prior conviction for a felonydrug offense’’ under 21 U.S.C.§ 841(b)(1)(B). But the Alleyne Courttook pains to point out that its holding didnot upset its previous ruling in Almenda-rez–Torres v. United States, which heldthat the fact of a prior conviction is not an

‘‘element’’ that must be found by a jury.See Alleyne, 133 S.Ct. at 2160 n. 1 (citingAlmendarez–Torres, 523 U.S. 224, 118S.Ct. 1219, 140 L.Ed.2d 350 (1998)). SoAlmendarez–Torres forecloses Jeanty’sclaim. See United States v. Greer, 440F.3d 1267, 1273–76 (11th Cir.2006).

AFFIRMED.

,

APPLE INC. and Next Software, Inc.(formerly known as Next Computer,

Inc.), Plaintiffs–Appellants,

v.

MOTOROLA, INC. (now known as Mo-torola Solutions, Inc.) and MotorolaMobility, Inc., Defendants–Cross Ap-pellants.

Nos. 2012–1548, 2012–1549.

United States Court of Appeals,Federal Circuit.

April 25, 2014.

Background: In patent infringement ac-tion between manufacturers of smart-phones, both parties sought declaratoryjudgments of non-infringement and inval-idity of patents related to smartphonetechnology. Following claim construction,the United States District Court for theNorthern District of Illinois, Richard A.

1287APPLE INC. v. MOTOROLA, INC.Cite as 757 F.3d 1286 (Fed. Cir. 2014)

Posner, Circuit Judge, 2012 WL 1959560,excluded damages experts, and 869F.Supp.2d 901, granted summary judg-ment of non-infringement with respect tocertain claims. Cross-appeals were taken.

Holdings: The Court of Appeals, Reyna,Circuit Judge, held that:

(1) term ‘‘heuristic’’ meant one or morerules to be applied to data to assist indrawing inferences from that data;

(2) term ‘‘analyzer server’’ meant a serverroutine separate from a client that re-ceives data having structures from theclient;

(3) term ‘‘realtime application program in-terface (API)’’ meant an API that al-lows realtime interaction between twoor more subsystems;

(4) accused device did not infringe patentdisclosing system for encrypting datacommunications;

(5) expert’s testimony regarding reason-able royalty rate was admissible;

(6) fact issues precluded summary judg-ment on damages; and

(7) patentee was not entitled to injunctionfor infringement of patent related tosmartphone technology.

Affirmed in part, reversed in part, vacatedin part, and remanded.

Rader, Chief Judge, filed an opinion dis-senting-in-part.

Prost, Circuit Judge, filed an opinion con-curring-in-part and dissenting-in-part.

1. Patents O324.5

Claim construction is a question of lawthat the Court of Appeals reviews de novo.

2. Patents O324.5

Whether patent claim language in-vokes the statute governing means-plus-

function limitations is a question of lawthat the Court of Appeals reviews de novo.35 U.S.C.A. § 112.

3. Patents O101(8)

The overall means-plus-function anal-ysis is a two-step process; in the first step,the court must determine if the claim limi-tation is drafted in means-plus-functionformat, in the second step, if the limitationis in means-plus-function format, the courtmust specifically review the specificationfor corresponding structure. 35 U.S.C.A.§ 112.

4. Patents O101(8)

When determining if the claim limita-tion is drafted in means-plus-function for-mat, the court must construe the claimlimitation to decide if it connotes ‘‘suffi-ciently definite structure’’ to a person ofordinary skill in the art, which requires thecourt to consider the specification. 35U.S.C.A. § 112.

5. Patents O101(8)

When determining whether a patentclaim is a means-plus-function claim, it ispossible to find that a claim limitation doesnot connote sufficiently definite structuredespite the presence of some correspond-ing structure in the specification. 35U.S.C.A. § 112.

6. Patents O101(8)

Not ‘‘every’’ mean-plus-function limi-tation is indefinite; only those that lackthe term means, do not connote sufficientlydefinite structure, and lack correspondingstructure. 35 U.S.C.A. § 112.

7. Patents O101(8)

When a claim limitation lacks theterm ‘‘means,’’ it creates a rebuttable pre-sumption that the statute governing


means-plus-function limitations does notapply. 35 U.S.C.A. § 112.

8. Patents O101(8)

The rebuttable presumption that thestatute governing means-plus-function lim-itations does not apply may be overcome ifthe claim fails to recite ‘‘sufficiently defi-nite structure’’ or merely recites a functionwithout reciting sufficient structure forperforming that function. 35 U.S.C.A.§ 112.

9. Patents O101(8)

Whether to draft a patent claim inbroad structural terms is the claim draft-er’s choice, and any resulting risk thatemanates from that choice is not a basisfor the court to rewrite a claim in means-plus-function format. 35 U.S.C.A. § 112.

10. Patents O101(8)

The correct inquiry when determiningwhether statute governing means-plus-function limitations applies, when ‘‘means’’is absent from a limitation, is whether thelimitation, read in light of the remainingclaim language, specification, prosecutionhistory, and relevant extrinsic evidence,has sufficiently definite structure to a per-son of ordinary skill in the art. 35U.S.C.A. § 112.

11. Patents O101(8)

When determining whether statutegoverning means-plus-function limitationsapplies, ‘‘structure’’ to a person of ordinaryskill in the art of computer-implementedinventions may differ from more tradition-al, mechanical structure. 35 U.S.C.A.§ 112.

12. Patents O101(8)

For purposes of determining whetherstatute governing means-plus-function lim-

itations applies, when ‘‘means’’ is absentfrom a limitation, a limitation has sufficientstructure when it recites a claim term witha structural definition that is either provid-ed in the specification or generally knownin the art. 35 U.S.C.A. § 112.

13. Patents O101(8)

For purposes of determining whetherstatute governing means-plus-function lim-itations applies, when ‘‘means’’ is absentfrom a limitation, structure may be provid-ed by describing the claim limitation’s op-eration, such as its input, output, or con-nections. 35 U.S.C.A. § 112.

14. Patents O101(8)

The means-plus-function limitation’soperation is more than just its function; itis how the function is achieved in thecontext of the invention. 35 U.S.C.A.§ 112.

15. Patents O101(8)

Even if a patentee elects to use a‘‘generic’’ claim term, such as a nonce wordor a verbal construct, properly construingthat term in view of the specification, pros-ecution history, etc., may still provide suf-ficient structure such that the presumptionagainst means-plus-function claiming re-mains intact. 35 U.S.C.A. § 112.

16. Patents O101(8)

When determining whether statutegoverning means-plus-function limitationsapplies, when ‘‘means’’ is absent from alimitation, the limitation need not connotea single, specific structure; rather, it maydescribe a class of structures. 35 U.S.C.A.§ 112.

17. Patents O101(8)

Even if the patent describes all struc-tures that perform the recited function,


this, by itself, does not overcome thestrong presumption that means-plus-func-tion claiming does not apply when theterm ‘‘means’’ is not recited in the claim.35 U.S.C.A. § 112.

18. Patents O101(8)

If the claim merely recites a genericnonce word and the remaining claim lan-guage, specification, prosecution history,and relevant external evidence provide nofurther structural description to a personof ordinary skill in the art, then the pre-sumption against means-plus-functionclaiming is rebutted. 35 U.S.C.A. § 112.

19. Patents O101(8)

If a claim recites a generic term that,properly construed in light of the specifica-tion, lacks sufficiently definite structure toa person of ordinary skill in the art, thepresumption that the statute governingmeans-plus-function limitations does notapply is overcome and the patentee hasinvoked means-plus-function claiming. 35U.S.C.A. § 112.

20. Patents O101(2)

Term ‘‘heuristic,’’ in patent disclosinguse of finger contacts to control a touch-screen computer, meant one or more rulesto be applied to data to assist in drawinginferences from that data.

21. Patents O101(2)

Term ‘‘analyzer server,’’ in patent dis-closing system for recognizing certainstructures, such as a telephone number, ona touchscreen and then linking certain ac-tions, such as calling the telephone num-ber, to the structure, meant a server rou-tine separate from a client that receivesdata having structures from the client.

22. Patents O101(2)

Term ‘‘linking actions to the detectedstructures,’’ in patent disclosing system for

recognizing certain structures, such as atelephone number, on a touchscreen andthen linking certain actions, such as callingthe telephone number, to the structure,meant creating a specified connection be-tween each detected structure and at leastone computer subroutine that causes theCPU to perform a sequence of operationson that detected structure.

23. Patents O101(2)

Term ‘‘realtime application programinterface (API),’’ in patent disclosing sys-tem for processing data in ‘‘realtime,’’meant an API that allows realtime interac-tion between two or more subsystems.

24. Patents O101(11)

In patent disclosing method for gener-ating ‘‘preamble sequences,’’ which wereused in communications between cellphones and base stations, third step of‘‘multiplying the outer code by the innercode’’ could not occur until after the firsttwo steps, ‘‘forming an outer code’’ and‘‘forming an inner code’’ were completed.


Steps in a method claim need notnecessarily be performed in the order theyare written; on the other hand, if gram-mar, logic, the specification, or the prose-cution history require the steps to be per-formed sequentially, then the claims are solimited.

26. Patents O235(2)

In patent disclosing system for en-crypting data communications, claimed‘‘transmit overflow sequence number,’’ orTOSN, was ‘‘never transmitted’’ to the re-ceiver in the claimed system, and thusaccused device did not infringe patent,where counterpart to TOSN in accusedproducts was transmitted to the receiver.


27. Patents O168(2.1)

Statements made in unrelated patentapplications are not relevant to claim con-struction.

28. Patents O168(1)

When construing claim terms in pat-ent disclosing system for encrypting datacommunications, district court could relyon statements before the Japan patent of-fice during prosecution of related patent;statements were made in official proceed-ing in which patentee had every incentiveto exercise care in characterizing scope ofits invention, two patents were related andshared familial relationship, both specifica-tions were the same, and, at time patenteemade its statements to Japanese patentoffice, Japanese application contained iden-tical claim.

29. Evidence O555.2

Under the rules governing admissionof expert testimony and precedent, a dis-trict court judge, acting as a gatekeeper,may exclude evidence if it is based uponunreliable principles or methods, or legallyinsufficient facts and data. Fed.RulesEvid.Rules 702, 703, 28 U.S.C.A.

30. Evidence O555.2, 570, 574

When deciding whether to admit ex-pert testimony, a judge must be cautiousnot to overstep its gatekeeping role andweigh facts, evaluate the correctness ofconclusions, impose its own preferredmethodology, or judge credibility, includ-ing the credibility of one expert over an-other; these tasks are solely reserved forthe fact finder. Fed.Rules Evid.Rules 702,703, 28 U.S.C.A.

31. Evidence O570

When determining whether to admitexpert testimony, the jury must still be

allowed to play its essential role as thearbiter of the weight and credibility ofexpert testimony. Fed.Rules Evid.Rules702, 703, 28 U.S.C.A.

32. Evidence O555.9

That the gatekeeping role of the judgeis limited to excluding testimony based onunreliable principles and methods is partic-ularly essential in the context of patentdamages. Fed.Rules Evid.Rules 702, 703,28 U.S.C.A.

33. Patents O319(1)

That one approach may better accountfor one aspect of a royalty estimation inpatent infringement action does not makeother approaches inadmissible.

34. Federal Courts O3600

Seventh Circuit reviews de novowhether the district court applied theproper legal framework, but reviews deci-sions to admit or exclude expert testimonyunder this framework for an abuse of dis-cretion. Fed.Rules Evid.Rules 702, 703,28 U.S.C.A.

35. Evidence O555.9

Proposed expert on damages relatedto infringement of patent disclosing use offinger contacts to control touchscreencomputer employed reliable principles andmethods, reliably applied them, and reliedupon legally sufficient facts or data, andthus testimony regarding reasonable roy-alty rate was admissible; methodology fo-cused on claimed technology, expert beganwith existing product containing featureshe contended were similar to asserted fea-tures, attempted to isolate value of similarfeatures by valuing other, non-claimedfeatures of patented device and subtract-ing this value, to reach reasonable royaltyexpert compared his resulting royalty to


related licenses and rationalized price dif-ferences, expert addressed possibility ofremoving asserted features from accuseddevices and argued that would frustrateconsumers and devalue products, and ex-pert’s application was straightforward andadequately supported by technical expertsand his own experience and expertise.Fed.Rules Evid.Rule 702, 28 U.S.C.A.

36. Evidence O555.9

Reliance of patentee’s damages experton computer chip identified by patentee’stechnical expert did not render inadmissi-ble damages expert’s testimony regardingreasonable royalty rate for infringement ofpatent disclosing system for processingdata in ‘‘realtime’’; rule governing admissi-bility of experts explicitly allowed expertto rely on information he had been madeaware of if experts in particular fieldwould reasonably rely on those kinds offacts or data in forming opinion on subject,regardless of which party employed ex-pert. Fed.Rules Evid.Rules 702, 703, 28U.S.C.A.

37. Evidence O555.9

Proposed testimony related to dam-ages for infringement of patent related tosmartphone technology was admissible;testimony included analysis of licenseagreements between patentee and ‘‘all ofthe major’’ cellular phone makers in theUnited States, except for competitor, at-tempted to separate out value of patentfrom total value of standard-essential pat-ent (SEP) portfolio, and concluded thatpatent was worth 40%–50% of SEP portfo-lio rate, and constructed cost estimate typ-ically relied upon when calculating patentdamages. Fed.Rules Evid.Rules 702, 703,28 U.S.C.A.

38. Evidence O555.9

Proposed testimony of patentee’s li-censing expert related to damages for in-

fringement of patent related to smart-phone technology was not tied to facts ofpatent infringement case, and thus wasinadmissible; expert admitted that heknew nothing about patentee’s standard-essential patent (SEP) portfolio and didnot even purport to link 40%–50% royaltyrate to claimed invention of patent.

39. Patents O312(2)

That a party may choose to pursueone course of proving damages over anoth-er does not render its expert’s damagestestimony inadmissible; nor is there a re-quirement that a patentee value every po-tential non-infringing alternative in orderfor its damages testimony to be admissible.

40. Patents O319(1)

Using sufficiently comparable licensesis a generally reliable method of estimat-ing the value of a patent.


A finding that a royalty estimate maysuffer from factual flaws does not, by it-self, support the legal conclusion that zerois a reasonable royalty.

42. Patents O319(1)

If a patentee’s evidence fails to sup-port its specific royalty estimate, the factfinder is still required to determine whatroyalty is supported by the record.

43. Patents O318(1)

If a patentee seeks lost profits andfails to meet its burden, then an award ofno lost profits is justified; this is because alost profits award is above and beyond thereasonable royalty floor required by thestatute.

44. Patents O319(1)

A fact finder may award no damagesonly when the record supports a zero roy-alty award.


45. Patents O323.2(3)

At summary judgment, a judge mayonly award a zero royalty for infringementif there is no genuine issue of material factthat zero is the only reasonable royalty;thus, if a patentee raises a factual issueregarding whether it is due any non-zeroroyalty, summary judgment must be de-nied.

46. Patents O319(1)

Simply because a patentee fails toshow that its royalty estimate is correctdoes not, by itself, justify awarding a roy-alty of zero at summary judgment.

47. Patents O323.2(3)

Genuine issue of material fact existedregarding whether patentee was due anynon-zero royalty for infringement of patentdisclosing system for recognizing certainstructures, such as a telephone number, ona touchscreen and then linking certain ac-tions, such as calling the telephone num-ber, to the structure, precluding summaryjudgment on damages for infringement.35 U.S.C.A. § 284.

48. Patents O323.2(3)

Genuine issue of material fact existedregarding whether competitor infringedpatent disclosing use of finger contacts tocontrol a touchscreen computer, preclud-ing summary judgment on patentee’s enti-tlement to injunction for infringement ofpatent.

49. Patents O317

Patentee was not entitled to injunctionfor infringement of patent related tosmartphone technology; patentee’s fair,reasonable, and nondiscriminatory(FRAND) licensing commitments, whichyielded many license agreements encom-

passing patent, strongly suggested thatmoney damages were adequate to fullycompensate patentee for any infringement,and patentee failed to demonstrate thatcompetitor’s infringement caused it irrepa-rable harm.

Patents O328(2)

5,319,712, 6,175,559. Construed andRuled Not Infringed.

Patents O328(2)

5,946,647, 6,343,263, 7,479,949. Con-strued.

Patents O328(2)

6,359,898. Cited.

E. Joshua Rosenkranz, Orrick, Herring-ton & Sutcliffe LLP, of New York, NewYork, argued for plaintiffs-appellants.With him on the brief were Mark S. Da-vies, Rachel M. McKenzie and T. VannPearce, Jr. of Washington, DC; and Mat-thew D. Powers, Tensegrity Law GroupLLP, of Redwood Shores, CA. Of counselwas Katherine M. Kopp, Orrick, Herring-ton & Sutcliffe LLP, of Washington, DC.

David A. Nelson, Quinn Emanuel Urqu-hart & Sullivan, LLP, argued for defen-dants-cross appellants. With him on thebrief were Stephen A. Swedlow; Brian C.Cannon, of Redwood Shores, CA; Kath-leen M. Sullivan and Edward J. Defranco,of New York, New York; and Charles K.Verhoeven, of San Francisco, CA. Of coun-sel were Raymond N. Nimrod and DavidElihu, of New York, NY; and AmandaScott Williamson, of Chicago, IL.

Joel Davidow, Cuneo, Gilbert & LaDuca,LLP, of Washington, DC, for amicus curi-


ae The American Antitrust Institute.With him on the brief was Robert J. Cyn-kar.

Charles W. Shifley, Banner & Witcoff,Ltd., of Chicago, IL, for amicus curiae TheIntellectual Property Law Association ofChicago.

Debra J. McComas, Haynes and Boone,LLP, of Dallas, TX, for amici curiae AlteraCorporation, et al. With her on the briefwas David L. McCombs. Of counsel onthe brief were Marta Beckwith, Cisco Sys-tems, Inc., of San Jose, CA; ElizabethLauner, Logitech Inc., of Newark, CA;and Richard J. Lutton, Jr., Nest Labs,Inc., of Palo Alto, CA.

Richard M. Brunell, Senior Advisor forcompetition matters, United States Feder-al Trade Commission, of Washington, DC,for amicus curiae United States FederalTrade Commission. With him on the briefwere David C. Shonka, Acting GeneralCounsel, William Cohen, Deputy GeneralCounsel, William F. Adkinson, Jr., Attor-ney and Suzanne Munck af Rosenchold,Chief Counsel for Intellectual Property.

Peter M. Lancaster, Dorsey & WhitneyLLP, of Minneapolis, MN, for amicus curi-ae The Institute of Electrical and Elec-tronics Engineers, Incorporated. Withhim on the brief were Michael A. Lindsay,of Minneapolis, MN, and Eileen M. Lach,General Counsel and Chief Compliance Of-ficer IEEE, of New York, NY.

Paul D. Clement, Brancroft PLLC, ofWashington, DC, for amici curiae VerizonCommunications Inc., et al. With him onthe brief was D. Zachary Hudson.

Tina M. Chappell, Director of Intellectu-al Property Policy, Intel Corporation, ofChandler, AZ, for amicus curiae Intel Cor-poration. With her on the brief were

Thomas G. Hunger and Matthew D.McGill, Gibson, Dunn & Crutcher LLP, ofWashington, DC.

Richard S. Taffet, Bingham McCutchenLLP, of New York, NY, for amicus curiaeQualcomm Incorporated. With him on thebrief were Patrick Strawbridge, of Boston,MA and David B. Salmons, of Washington,DC.

Brian R. Matsui, Morrison & FoersterLLP, of Washington, DC, for amicus curi-ae Law Professors Thomas F. Cotter, etal. With him on the brief was Natalie R.Ram.

Roy T. Englert, Jr., Robbins, Russell,Englert, Orseck, Untereiner & SauberLLP, of Washington, DC, for amicus curi-ae, BSA The Software Alliance.

Patrick J. Flinn, Alston and Bird LLP,of Atlanta, GA, for amici curiae NokiaCorporation, et al. With him on the briefwas Keith E. Broyles.

Brian C. Riopelle, McGuire Woods LLP,of Richmond, VA, for amicus curiae Re-search in Motion Limited. With him onthe brief were Robert M. Tyler and Kris-ten M. Calleja.

Constantine L. Trela, Jr., Sidley AustinLLP, of Chicago, IL, for amicus curiaeMicrosoft Corporation. With him on thebrief were Richard A. Cederoth and Na-thaniel C. Love. Of counsel on the briefwere T. Andrew Culbert and David E.Killough, Microsoft Corporation, of Red-mond, WV.

Before RADER, Chief Judge, PROSTand REYNA, Circuit Judges.

Opinion for the court filed by CircuitJudge REYNA.


Opinion dissenting in part filed by ChiefJudge RADER.

Opinion concurring in part anddissenting in part filed by Circuit JudgePROST.

REYNA, Circuit Judge.

Plaintiffs Apple Inc. and Next Software,Inc. (‘‘Apple’’) filed a complaint againstDefendants Motorola, Inc. and MotorolaMobility, Inc. (‘‘Motorola’’) on October 29,2010 in the United States District Courtfor the Western District of Wisconsin, as-serting infringement of three patents.Motorola counterclaimed, asserting six ofits own patents. Apple amended its com-plaint to include an additional twelve pat-ents. Both parties also sought declaratoryjudgments of non-infringement and inval-idity.

After claim construction began in Wis-consin, the case was transferred to theUnited States District Court for theNorthern District of Illinois, Judge Posnersitting by designation. The district courtin Illinois completed claim construction.Based upon its claim construction deci-sions, the court granted summary judg-ment of non-infringement with respect tocertain claims and excluded the vast ma-jority of both parties’ damages expert evi-dence for the remaining claims. With lit-tle expert evidence deemed admissible, thecourt granted summary judgment that nei-ther side was entitled to any damages oran injunction. Despite infringement beingassumed, the district court dismissed allclaims with prejudice before trial.

Only six patents are at issue on appeal:Apple’s U.S. Patent Nos. 7,479,949; 6,343,-263; and 5,946,647; and Motorola’s U.S.Patent Nos. 6,359,898; 6,175,559; and5,319,712. The parties contest the district

court’s claim construction, admissibility,damages, and injunction decisions. As de-tailed below, we affirm the district court’sclaim construction decisions, with the ex-ception of its construction of Apple’s 8949patent. With a minor exception, the dis-trict court’s decision to exclude the dam-ages evidence presented by both Appleand Motorola is reversed. We also re-verse the district court’s grant of summaryjudgment of no damages for infringementof Apple’s patents. Based upon our rever-sal of the district court’s claim constructionof the 8949 patent, we vacate the court’sgrant of summary judgment regarding Ap-ple’s request for an injunction. Thecourt’s decision that Motorola is not enti-tled to an injunction for infringement ofthe FRAND-committed 8898 patent is af-firmed. We address these, and all relatedissues, in turn.

CLAIM CONSTRUCTION

[1] The parties raise claim constructionissues regarding Apple’s 8949, 8263, and8647 patents and Motorola’s 8559 and 8712patents. Claim construction is a questionof law that we review de novo. CyborCorp. v. FAS Techs., Inc., 138 F.3d 1448,1454–55 (Fed.Cir.1998) (en banc).

Apple’s 8949 patent

The district court construed claims 1, 2,9, and 10 of the 8949 patent and, basedupon its construction, granted Motorola’smotion for summary judgment of nonin-fringement for the majority of the accusedproducts. Because the district court mis-takenly construed certain limitations asmeans-plus-function limitations, we re-verse its claim construction and vacate thesubsequent summary judgment decision.

The 8949 patent discloses the use offinger contacts to control a touchscreen


computer. Claims 1, 2, 9, and 10 of the8949 patent are recited below, with thelimitations at issue emphasized. Claim 1recites:

A computing device, comprising: a touchscreen display; one or more processors;memory; and one or more programs,

wherein the one or more programs arestored in the memory and configured tobe executed by the one or more pro-cessors, the one or more programs in-cluding:

instructions for detecting one or morefinger contacts with the touch screendisplay;

instructions for applying one or moreheuristics to the one or more fingercontacts to determine a command forthe device; and

instructions for processing the com-mand;

wherein the one or more heuristics com-prise:

a vertical screen scrolling heuristic fordetermining that the one or more fingercontacts correspond to a one-dimension-al vertical screen scrolling commandrather than a two-dimensional screentranslation command based on an angleof initial movement of a finger contactwith respect to the touch screen display;

a two-dimensional screen translationheuristic for determining that the one ormore finger contacts correspond to thetwo-dimensional screen translation com-mand rather than the one-dimensionalvertical screen scrolling command basedon the angle of initial movement of thefinger contact with respect to the touchscreen display; and

a next item heuristic for determiningthat the one or more finger contacts

correspond to a command to transitionfrom displaying a respective item in aset of items to displaying a next item inthe set of items.

8949 patent at col. 122 l. 37–col. 123 l. 2(emphases added). Claim 2 recites:

The computing device of claim 1, where-in the one or more heuristics include aheuristic for determining that the one ormore finger contacts correspond to acommand to translate content within aframe rather than translating an entirepage that includes the frame.

Id. at col. 123, lns 3–7 (emphasis added).Claim 10 recites:

The computing device of claim 9, where-in the first set of heuristics comprises aheuristic for determining that the one ormore first finger contacts correspond toa one-dimensional horizontal screenscrolling command rather than the two-dimensional screen translation commandbased on the angle of initial movementof the finger contact with respect to thetouch screen display.

Id. at col. 124, ll. 1–7. The district courtfirst found that the claim term ‘‘heuristic’’was not indefinite, instead construing it as‘‘one or more rules to be applied to data toassist in drawing inferences from thatdata.’’ Next, the court found that the‘‘heuristic’’ limitations in claims 1, 2, 9, and10 described functions ‘‘without describingthe structure necessary to perform thefunctions.’’ Accordingly, the court con-cluded that these claim limitations weremeans-plus-function limitations under 35U.S.C. § 112, ¶ 6, despite not reciting theword ‘‘means.’’ The court next found thatthe specification contained sufficient ‘‘cor-responding structure’’ capable of perform-ing the claimed functions. 35 U.S.C.§ 112. In doing so, the court limited the


‘‘next item heuristic’’ in claim 1 to ‘‘a heu-ristic that uses as one input a user’s fingertap on the right side of the device’s touchscreen.’’ Based upon this construction,Motorola moved for summary judgment ofnoninfringement. The court concludedthat the only accused products that use a‘‘finger tap’’ in this manner are those thatcome pre-loaded with a specific program:the Amazon Kindle application. The courtgranted Motorola’s motion for summaryjudgment of non-infringement for the re-maining accused products.

On appeal, Motorola again argues that‘‘heuristic’’ is indefinite. In the alternative,Motorola argues that the district court cor-rectly concluded that the heuristic limita-tions were drafted in means-plus-functionformat and correctly limited the ‘‘next itemheuristic’’ limitation to the finger tap ges-ture. Apple points out that the claims donot use the word ‘‘means’’ and that thiscreates a strong presumption against con-struing the limitations as means-plus-func-tion limitations. Apple argues that theheuristic limitations connote sufficientlydefinite structure such that Motorola hasnot overcome this strong presumption.

[2–4] Whether claim language invokesSection 112, ¶ 6 is a question of law thatwe review de novo. Inventio AG v. Thys-senKrupp Elevator Americas Corp., 649F.3d 1350, 1356 (Fed.Cir.2011); LightingWorld, Inc. v. Birchwood Lighting, Inc.,382 F.3d 1354, 1358 (Fed.Cir.2004). Sec-tion 112, ¶ 6 states:

An element in a claim for a combinationmay be expressed as a means or step forperforming a specified function withoutthe recital of structure, material, or actsin support thereof, and such claim shallbe construed to cover the correspondingstructure, material, or acts described inthe specification and equivalents thereof.

35 U.S.C. § 112, ¶ 6. The overall means-plus-function analysis is a two-step pro-cess. Naturally, there is some analyticaloverlap between these two steps. In thefirst step, we must determine if the claimlimitation is drafted in means-plus-functionformat. As part of this step, we mustconstrue the claim limitation to decide if itconnotes ‘‘sufficiently definite structure’’ toa person of ordinary skill in the art, whichrequires us to consider the specification(among other evidence). In the secondstep, if the limitation is in means-plus-function format, we must specifically re-view the specification for ‘‘correspondingstructure.’’ Thus, while these two ‘‘struc-ture’’ inquiries are inherently related, theyare distinct.

[5, 6] The Dissent is concerned that wehave impermissibly looked for correspond-ing structure in the specification beforedeciding that the claim is in means-plus-function format thereby creating a newrule that renders ‘‘every means-plus-func-tion claim term indefinite.’’ J. Prost Dis-sent at 1335–36 (emphasis in original)(‘‘Dissent’’). This is not our analysis. TheDissent correctly notes that the first stepin the means-plus-function analysis re-quires us to determine whether the entireclaim limitation at issue connotes ‘‘suffi-ciently definite structure’’ to a person ofordinary skill in the art. Dissent at 1334–35. In so doing, we naturally look to thespecification, prosecution history, and rele-vant external evidence to construe the lim-itation. While this inquiry may be similarto looking for corresponding structure inthe specification, our precedent requires itwhen deciding whether a claim limitationlacking means connotes sufficiently defi-nite structure to a person of ordinary skillin the art. See, e.g., Inventio, 649 F.3d at1357 (‘‘It is proper to consult the intrinsicrecord, including the written description,


when determining if a challenger has re-butted the presumption that a claim lack-ing the term ‘means’ recites sufficientlydefinite structure.’’); Lighting World, 382F.3d at 1360–64 (examining the writtendescription and external evidence); FloHealthcare Solutions, LLC v. Kappos, 697F.3d 1367, 1374 (Fed.Cir.2012) (examiningremaining claim language, written descrip-tion, and external evidence); Linear Tech.Corp. v. Impala Linear Corp., 379 F.3d1311, 1320 (Fed.Cir.2004) (examining re-maining claim language and external evi-dence). Because these inquiries are dis-tinct, it is possible to find that a claimlimitation does not connote sufficiently def-inite structure despite the presence ofsome corresponding structure in the speci-fication. See, e.g., Massachusetts Inst. ofTech. v. Abacus Software, 462 F.3d 1344,1355 (Fed.Cir.2006) (‘‘MIT’’); WelkerBearing Co. v. PHD, Inc., 550 F.3d 1090,1096–97 (Fed.Cir.2008). As such, not ‘‘ev-ery’’ mean-plus-function limitation is indef-inite under our precedent; only those thatlack the term means, do not connote suffi-ciently definite structure, and lack corre-sponding structure. We do not state orapply a different rule in this case. In thiscase, as we find that the claims connotesufficiently definite structure to a personof ordinary skill in the art, we do not reachthe second step of the means-plus-functionanalysis.

[7, 8] As the district court recognized,when a claim limitation lacks the term‘‘means,’’ it creates a rebuttable presump-tion that Section 112, ¶ 6 does not apply.See, e.g., Lighting World, 382 F.3d at 1358;CCS Fitness, Inc. v. Brunswick Corp., 288F.3d 1359, 1369 (Fed.Cir.2002). This pre-sumption may be overcome if the claimfails to recite ‘‘sufficiently definite struc-ture’’ or merely recites a ‘‘function without

reciting sufficient structure for performingthat function.’’ Linear, 379 F.3d at 1319(quoting Watts v. XL Sys. Inc., 232 F.3d877, 880 (Fed.Cir.2000)); see also Inventio,649 F.3d at 1356. We have repeatedlycharacterized this presumption as ‘‘strong’’and ‘‘not readily overcome’’ and, as such,have ‘‘seldom’’ held that a limitation with-out recitation of ‘‘means’’ is a means-plus-function limitation. Lighting World, 382F.3d at 1358, 1362; Inventio, 649 F.3d at1356; see also Flo Healthcare, 697 F.3d at1374 (‘‘When the claim drafter has notsignaled his intent to invoke § 112, ¶ 6 byusing the term ‘means,’ we are unwilling toapply that provision without a showingthat the limitation essentially is devoid ofanything that can be construed as struc-ture.’’).

[9] The Dissent suggests that choosingto include ‘‘means’’ in a claim limitation isa ‘‘minor drafting decision’’ that corre-spondingly merits little weight in a Section112, ¶ 6 analysis. Dissent at 1337. Wedisagree. The strong presumption createdby not including means in a claim limita-tion provides clarity and predictability forthe public and the patentee alike. It helpsthe public determine when claim elementsare expressly limited to structures dis-closed in the specification (or their equiva-lents) and provides the patentee with thetools for reliably invoking or avoidingmeans-plus-function claiming. It also sig-nals to the court that the patentee haschosen to avail, or avoid, the benefits ofSection 112, ¶ 6. We recognize that thechoice to draft a claim in ‘‘broad structuralterms’’ rather than in a means-plus-func-tion format may render the claim morevulnerable to an invalidity attack. Id.Whether to draft a claim in broad structur-al terms is the claim drafter’s choice, andany resulting risk that emanates from that


choice is not a basis for the court to re-write a claim in means-plus-function for-mat. See id. By focusing on the claimterms the patentee chose, this presumptionalso reaffirms the primacy of the claimlanguage during claim construction, as out-lined in Phillips v. AWH Corp. 415 F.3d1303 (Fed.Cir.2005). Here, as in all as-pects of claim construction, ‘‘the name ofthe game is the claim.’’ In re Hiniker Co.,150 F.3d 1362, 1369 (Fed.Cir.1998) (quot-ing Giles Sutherland Rich, Extent of Pro-tection and Interpretation of Claims–American Perspectives, 21 Int’l Rev. In-dus. Prop. & Copyright L. 497, 499 (1990)).

In this case, Motorola bears the burdenof overcoming the presumption that Sec-tion 112, ¶ 6 does not apply by a prepon-derance of the evidence. See Apex Inc. v.Raritan Computer, Inc., 325 F.3d 1364,1372 (Fed.Cir.2003). The district courtmade several erroneous findings that led itto incorrectly conclude that Motorola re-butted this strong presumption. The dis-trict court misapplied our precedent byrequiring the claim limitations of the 8949patent themselves to disclose ‘‘a step-by-step algorithm as required by AristocratTechnologies.’’ Aristocrat and relatedcases hold that, if a patentee has invokedcomputer-implemented means-plus-func-tion claiming, the corresponding structurein the specification for the computer im-plemented function must be an algorithmunless a general purpose computer is suffi-cient for performing the function. Aristo-crat Techs. Austl. Pty Ltd. v. Int’l GameTech., 521 F.3d 1328, 1333 (Fed.Cir.2008)(requiring disclosure of an algorithm whenit is not disputed that claims were draftedin means-plus-function format); WMSGaming, Inc. v. Int’l Game Tech., 184F.3d 1339, 1349 (Fed.Cir.1999) (same);ePlus, Inc. v. Lawson Software, Inc., 700F.3d 509, 518 (Fed.Cir.2012); Typhoon

Touch Techs., Inc. v. Dell, Inc., 659 F.3d1376, 1385 (Fed.Cir.2011). But see In reKatz Interactive Call Processing PatentLitig., 639 F.3d 1303, 1316 (Fed.Cir.2011)(finding that disclosure of a general pur-pose computer is sufficient correspondingstructure for means-plus-function claims).

[10] In all these cases, the claims recit-ed the term ‘‘means,’’ thereby expresslyinvoking means-plus-function claiming. Inaddition, the parties in these cases did notdispute on appeal that these claims weredrafted in means-plus-function format.Hence, where a claim is not drafted inmeans-plus-function format, the reasoningin the Aristocrat line of cases does notautomatically apply, and an algorithm istherefore not necessarily required. Thecorrect inquiry, when ‘‘means’’ is absentfrom a limitation, is whether the limitation,read in light of the remaining claim lan-guage, specification, prosecution history,and relevant extrinsic evidence, has suffi-ciently definite structure to a person ofordinary skill in the art. Here, the answeris yes.

[11] ‘‘Structure’’ to a person of ordi-nary skill in the art of computer-imple-mented inventions may differ from moretraditional, mechanical structure. For ex-ample, looking for traditional ‘‘physicalstructure’’ in a computer software claim isfruitless because software does not containphysical structures. Indeed, the typicalphysical structure that implements soft-ware, a computer, cannot be relied upon toprovide sufficiently definite structure for asoftware claim lacking ‘‘means.’’ Rather,to one of skill in the art, the ‘‘structure’’ ofcomputer software is understood through,for example, an outline of an algorithm, aflowchart, or a specific set of instructionsor rules. See, e.g., Typhoon Touch, 659


F.3d at 1385 (‘‘[T]he patent need only dis-close sufficient structure for a person ofskill in the field to provide an operativesoftware program for the specified func-tion.’’); Finisar Corp. v. DirecTV Grp.,Inc., 523 F.3d 1323, 1340 (Fed.Cir.2008).1

Requiring traditional physical structure insoftware limitations lacking the termmeans would result in all of these limita-tions being construed as means-plus-func-tion limitations and subsequently beingfound indefinite.

[12] A limitation has sufficient struc-ture when it recites a claim term with astructural definition that is either providedin the specification or generally known inthe art. See, e.g., Flo Healthcare, 697F.3d at 1374 (‘‘We will not apply § 112, ¶ 6if the limitation contains a term that ‘isused in common parlance or by persons ofskill in the pertinent art to designatestructure.’ ’’) (quoting Lighting World, 382F.3d at 1359); Personalized Media, 161F.3d at 704–05. In Personalized Media,we found that the claim term ‘‘detector,’’by itself, connoted sufficient structure to aperson of ordinary skill in the art. 161F.3d at 704–05 (agreeing with ALJ that‘‘ ‘detector’ had a well-known meaning tothose of skill in the electrical arts connota-tive of structure’’). There, we contrastedthe structural term ‘‘detector’’ with gener-ic, non-structural, terms such as ‘‘means,’’‘‘element,’’ and ‘‘device.’’ Id. at 705; seealso Apex, 325 F.3d at 1373 (finding thatthe term ‘‘circuit,’’ coupled with identifierssuch as ‘‘interface,’’ ‘‘programming,’’ and‘‘logic,’’ connoted sufficient structure to aperson of ordinary skill in the art).

[13, 14] Structure may also be provid-ed by describing the claim limitation’s op-

eration, such as its input, output, or con-nections. The limitation’s operation ismore than just its function; it is how thefunction is achieved in the context of theinvention. For example, in Linear, wefound that the claim term ‘‘circuit’’ has aknown structural definition and that thepatent described the circuit’s operation,including its input, output, and objective.379 F.3d at 1320–21. Similarly, in Light-ing World, we found that ‘‘connector’’ hada known structural definition and that thespecification described its operational re-quirements, including which claim ele-ments it was connected to and how theywere connected. 382 F.3d at 1361–63. Inboth cases, we found the presumptionagainst means-plus-function claiming wasunrebutted.

[15] Even if a patentee elects to use a‘‘generic’’ claim term, such as ‘‘a nonceword or a verbal construct,’’ properly con-struing that term (in view of the specifica-tion, prosecution history, etc.) may stillprovide sufficient structure such that thepresumption against means-plus-functionclaiming remains intact. Id. at 1360; seealso Inventio, 649 F.3d at 1356–57(‘‘Claims are interpreted in light of thewritten description supporting them, andthat is true whether or not the claim con-struction involves interpreting a ‘means’clause.’’); MIT, 462 F.3d at 1354 (‘‘Thegeneric terms ‘mechanism,’ ‘means,’ ‘ele-ment,’ and ‘device,’ typically do not con-note sufficiently definite structure.’’). Forexample, in Inventio, the claim includedthe generic term ‘‘device.’’ 649 F.3d at1354 (reciting ‘‘at least one modernizingdevice and connecting the at least one

1. We cite these cases as examples of ‘‘struc-ture’’ to a person ordinarily skilled in the artof computer software. We do not cite thesecases for the principle that we must reviewthe specification, prosecution history, and rel-

evant external evidence when deciding if aclaim limitation connotes structure. See Dis-sent at fn. 2. As discussed herein, there isample support for that proposition elsewhere.See, e.g., Inventio, 649 F.3d at 1356–57.


modernizing device to said floor terminalsand said at least one computing unit.’’)(emphasis added). However, the specifica-tion described the modernizing device’s in-put, output, internal components, and howthe internal components were intercon-nected. Id. at 1358–59. As such, the pre-sumption against means-plus-functiontreatment was not overcome. See also FloHealthcare, 697 F.3d at 1374–75 (notingthat ‘‘mechanism’’ is a generic term, butthen looking to remaining claim languageand written description before finding thatthe full claim limitation connoted struc-ture). These cases teach that, if a limita-tion recites a term with a known structuralmeaning, or recites either a known or ge-neric term with a sufficient description ofits operation, the presumption againstmeans-plus-function claiming remains in-tact.

[16, 17] The limitation need not con-note a single, specific structure; rather, itmay describe a class of structures. See,e.g., Personalized Media Communica-tions, LLC v. Int’l Trade Comm’n, 161F.3d 696, 705 (Fed.Cir.1998) (‘‘Eventhough the term ‘detector’ does not specifi-cally evoke a particular structure, it doesconvey to one knowledgeable in the art avariety of structures known as ‘detec-tors.’ ’’); Flo Healthcare, 697 F.3d at1374–75 (finding that claim term ‘‘heightadjustment mechanism’’ designates ‘‘aclass of structures that are generally un-derstood to persons of skill in the art’’).Indeed, even if the patent describes allstructures that perform the recited func-tion, this, by itself, does not overcome thestrong presumption that means-plus-func-tion claiming does not apply when theterm ‘‘means’’ is not recited in the claim.Lighting World, 382 F.3d at 1361–62.

[18, 19] By contrast, if the claim mere-ly recites a generic nonce word and the

remaining claim language, specification,prosecution history, and relevant externalevidence provide no further structural de-scription to a person of ordinary skill inthe art, then the presumption againstmeans-plus-function claiming is rebutted.In MIT, for example, the claims recited a‘‘colorant selection mechanism.’’ 462 F.3dat 1353. As noted, ‘‘mechanism’’ by itselfdoes not connote sufficient structure, andthe term ‘‘colorant selection’’ was not de-fined in the specification or otherwiseknown to a person of ordinary skill in theart. Id. at 1353–55. Further, the paten-tee used the terms ‘‘mechanism’’ and‘‘means’’ interchangeably in the specifica-tion. Id. at 1354; see also Mas–HamiltonGrp. v. LaGard Inc., 156 F.3d 1206, 1214–16 (Fed.Cir.1998) (claim recited ‘‘element’’and ‘‘member’’ and patent provided no fur-ther structural description of these genericterms); Welker Bearing, 550 F.3d at 1096–97 (claim recited a ‘‘mechanism’’ withoutfurther structure described in specifica-tion). Thus, if a claim recites a genericterm that, properly construed in light ofthe specification, lacks sufficiently definitestructure to a person of ordinary skill inthe art, the presumption is overcome andthe patentee has invoked means-plus-func-tion claiming.

[20] With this precedent in mind, weturn to the claim limitations at issue in the8949 patent. We find that ‘‘heuristic’’ hasa known meaning and the 8949 patent alsodescribes the limitation’s operation, includ-ing its input, output, and how its outputmay be achieved. Accordingly, the heuris-tic claim limitations recited above have‘‘sufficiently definite structure,’’ to a per-son of ordinary skill in the art, for per-forming the recited functions.

Broadly speaking, the function of therecited limitations is to identify a com-


mand based upon particular finger con-tacts. To achieve this function, the patentdescribes ‘‘heuristics.’’ Depending uponthe circumstances, heuristic is not neces-sarily a generic, structureless ‘‘nonce wordor a verbal construct’’ without any mean-ing, such as ‘‘mechanism,’’ ‘‘means,’’ ‘‘ele-ment,’’ or ‘‘widget.’’ The district courtcorrectly determined that a person of ordi-nary skill in the art would understand‘‘heuristic’’ to mean ‘‘one or more rules tobe applied to data to assist in drawinginferences from that data.’’ In this sense,‘‘heuristic’’ is similar to words that define aclass of structures, such as ‘‘connector,’’‘‘circuit,’’ and ‘‘detector,’’ and it does notinclude all means for performing the recit-ed function. See, e.g., Flo Healthcare, 697F.3d at 1374 (‘‘We will not apply § 112, ¶ 6if the limitation contains a term that ‘isused in common parlance or by persons ofskill in the pertinent art to designatestructure’ ’’) (quoting Lighting World, 382F.3d at 1359); Personalized Media, 161F.3d at 704–05. The fact that heuristic isdefined partly in terms of its function doesnot detract ‘‘from the definiteness of [the]structure’’ it may connote. PersonalizedMedia, 161 F.3d at 703–05. Indeed,‘‘many devices take their names from thefunctions they perform.’’ Greenberg v.Ethicon Endo–Surgery, Inc., 91 F.3d 1580,1583 (Fed.Cir.1996); see also MIT, 462F.3d at 1354; Lighting World, 382 F.3d at1359–60.

We need not decide here whether theterm ‘‘heuristic,’’ by itself, connotes suffi-cient structure to maintain the presump-tion against means-plus-function claimingbecause, in this case, the claims do notnakedly recite heuristics without furtherdescription in the remaining claim lan-guage and specification. To the contrary,the claim language and specification dis-close the heuristics’ operation within the

context of the invention, including the in-puts, outputs, and how certain outputs areachieved.

In all cases, the claimed input is a fingercontact. The specification explains thatthe finger contacts may be taps, swipes,double taps, or finger rolling, and mayinvolve one or two fingers contacting thescreen at different initial angles. See, e.g.,8949 Patent at col. 19, ll. 30–46; col. 65, ll.21–24; col. 66, ll. 47–51. The claims reciteheuristics with varying objectives, includ-ing vertical screen scrolling, two-dimen-sional screen translation, moving to thenext item in a list, and translating contentwithin a frame. The claims also explainthat the invention differentiates betweenvertical scrolling and two-dimensionaltranslation based upon the angle of initialmovement of the finger contact.

The written description provides furtherdetails regarding the heuristics’ inputs andoutputs. Regarding one-dimensional verti-cal screen scrolling, the specification ex-plains that ‘‘in response to an upwardswipe gesture 3937 by the user that iswithin a predetermined angle (e.g., 27∞) ofbeing perfectly vertical, the web page mayscroll one-dimensionally upward in the ver-tical direction.’’ 8949 Patent at col. 64, ll.21–25. Regarding two-dimensional trans-lation, the specification discloses that ‘‘inresponse to an upward swipe gesture 3939(FIG.39C) by the user that is not within apredetermined angle (e.g., 27∞) of beingperfectly vertical, the web page may scrolltwo-dimensionally along the direction ofthe swipe.’’ Id. at col. 64, ll. 30–34. Thespecification defines two-dimensionalmovement as ‘‘simultaneous movement inboth the vertical and horizontal di-rections.’’ Id. The specification explainshow a user can move to the next item in alist via a finger tap gesture on the right


side of the screen, a right-to-left fingerswipe, or by tapping a next image icon.Id. at col. 30, ll. 42–67.

The specification also discusses thestructure behind translating ‘‘content with-in a frame rather than translating theentire page that includes the frame.’’ Id.at col. 123, ll. 6–8. For performing thisfunction, the specification describes an ‘‘M-finger translation gesture 4214,’’ where Mis a number different from the number offingers used to translate the entire page.Id. at col. 75, ll. 18–26. The specificationalso explains that the direction of transla-tion may be the direction of the ‘‘move-

ment of the M-finger translation gesture.’’Id. at col. 75, ll. 34–35. Alternatively, thedirection of translation may be determinedby the angle of the movement of the M-finger gesture, according to a particularrule, i.e. a specific, identifying heuristic.Id. at col. 75, ll. 39–44.

The figures in the 8949 patent providefurther structural details. Figs. 12A, 39C,42A, 42B, and 42C illustrate the fingercontacts described in the specification thatresult in vertical scrolling (3937), two-di-mensional translation (3939), turning to thenext item (1218, 1220, and 1212), or trans-lating within a frame (4214).


In sum, the claims at issue differ fromthose that simply recite a generic meansor mechanism, without further descriptionin the remaining claim language or thespecification. Instead, the claim languageand specification outline the rules that theheuristics follow, based upon, for example,the initial angle of a finger contact, thenumber of fingers making contact, the di-rection of movement of a finger contact, aspecific swiping gesture, taping a certainlocation on the screen, or the angle of

movement of a finger on the screen. SeeWelker Bearing, 550 F.3d at 1096–97; In-ventio, 649 F.3d at 1359 (‘‘This is not acase where a claim nakedly recites a ‘de-vice’ and the written description fails toplace clear structural limitations on the‘device.’ ’’). Thus, the 8949 patent recites aclaim term with a known meaning and alsodescribes its operation, including its input,output, and how its output may beachieved.


Accordingly, the heuristic claim limita-tions provide ‘‘sufficiently definite struc-ture,’’ to a person of ordinary skill in theart, for performing the recited function,and Motorola has not rebutted the strongpresumption against means-plus-functionclaiming. We reverse the district court’sconstruction that the ‘‘heuristic’’ claim limi-tations were drafted in means-plus-func-tion format and vacate its summary judg-ment of non-infringement.

Apple’s 8647 Patent

[21, 22] Regarding Apple’s 8647 patent,the parties dispute the meaning of theclaim terms ‘‘analyzer server’’ and ‘‘linkingactions to the detected structures.’’ Thedistrict court construed ‘‘analyzer server’’as ‘‘a server routine separate from a clientthat receives data having structures fromthe client’’ and ‘‘linking actions to the de-tected structures’’ as ‘‘creating a specifiedconnection between each detected struc-ture and at least one computer subroutinethat causes the CPU to perform a se-quence of operations on that detectedstructure.’’ Apple argues that both con-structions are erroneous. We disagreewith Apple and affirm the district court’sclaim construction.

The 8647 patent discloses a system forrecognizing certain structures (such as atelephone number) on a touchscreen andthen linking certain actions (such as callingthe telephone number) to the structure.For example, a user may be able to call orsave a phone number it has received viatext message or email simply by touchingthe number on the screen of its device.Claim 1 of the 8647 patent, with relevantclaim limitations emphasized, recites:

A computer-based system for detectingstructures in data and performing ac-tions on detected structures, comprising:

an input device for receiving data;

an output device for presenting the data;

a memory storing information includingprogram routines including an analyzerserver for detecting structures in thedata, and for linking actions to the de-tected structures;

a user interface enabling the selection ofa detected structure and a linked action;

and an action processor for performingthe selected action linked to the selectedstructure; and a processing unit coupledto the input device, the output device,and the memory for controlling the exe-cution of the program routines.

8647 patent at col. 7, ll. 9–24 (emphasisadded). The district court agreed withMotorola that ‘‘analyzer server’’ should beconstrued as ‘‘a server routine separatefrom a client that receives data havingstructures from the client.’’ Apple arguesthat the analyzer server need not be ‘‘sep-arate from a client.’’ Instead, Apple ar-gues that ‘‘analyzer server’’ should beconstrued as ‘‘a program routine(s) thatreceives data, uses patterns to detectstructures in the data, and links actions tothe detected structures.’’

We agree with the district court’s con-struction of ‘‘analyzer server.’’ As the dis-trict court recognized, the plain meaning of‘‘server,’’ when viewed from the perspec-tive of a person of ordinary skill in the art,entails a client-server relationship. Con-sistent with this perspective, the specifica-tion discloses an analyzer server that isseparate from the application it serves.The analyzer server is part of the ‘‘pro-gram 165 of the present invention.’’ 8647patent at col. 3, ll. 38–39. Fig. 1 shows theprogram 165 and the application 167 asseparate parts of a random-access memory(RAM):


Id. at Fig. 1. Further, the specificationstates that ‘‘the program 165 of the pres-ent invention is stored in RAM 170 andcauses CPU 120 to identify structures indata presented by the application 167.’’Id. at col. 3, ll. 37–41. Thus, the specifica-tion describes the analyzer server and theapplication, which it serves, as separatestructures.

Apple does not point to evidence sug-gesting a different ordinary meaning, nordo we discern such evidence in the recordbefore this court. Indeed, Apple’s pro-posed construction contradicts the claimlanguage because it reads ‘‘analyzer ser-ver’’ out of the claim. The claim recites‘‘routines including an analyzer server fordetecting structures in the data, and forlinking actions to the detected struc-tures.’’ Apple’s proposed construction re-cites program routines that detect struc-tures and links actions to the detectedstructures, without any mention of ‘‘ana-lyzer servers.’’ Apple’s construction es-sentially takes the claim text and removesthe ‘‘analyzer server,’’ leaving the rest ba-sically unchanged. Thus, Apple’s con-struction conflicts with the claim languageby ignoring the claim term ‘‘server.’’ See,e.g., Merck & Co. v. Teva Pharms. USA,Inc., 395 F.3d 1364, 1372 (Fed.Cir.2005)(‘‘A claim construction that gives meaningto all the terms of the claim is preferred

over one that does not do so.’’); PauseTech., LLC v. TiVo, Inc., 419 F.3d 1326,1334 (Fed.Cir.2005) (‘‘In construing claims,however, we must give each claim term therespect that it is due.’’); Strattec Sec.Corp. v. Gen. Auto. Specialty Co., 126 F.3d1411, 1417 (Fed.Cir.1997) (holding that itwas legal error for the district court toinstruct the jury that the claim term‘‘sheet’’ was not properly considered partof the claim); Exxon Chem. Patents, Inc.v. Lubrizol Corp., 64 F.3d 1553, 1557 (Fed.Cir.1995) (‘‘We must give meaning to allthe words in Exxon’s claims.’’). By con-trast, the district court’s construction com-ports with the ordinary meaning of ‘‘ser-ver’’ and is supported by the specification.Accordingly, we affirm the district court’sconstruction of ‘‘analyzer server.’’

The district court also agreed with Mo-torola that ‘‘linking actions to the detectedstructures’’ should be construed as ‘‘creat-ing a specified connection between eachdetected structure and at least one com-puter subroutine that causes the CPU toperform a sequence of operations on thatdetected structure.’’ Apple argues thatthe district court’s construction is errone-ous for two reasons. First, the districtcourt incorrectly added the ‘‘specified con-nection’’ limitation. Second, the claims re-quire linking multiple actions to eachstructure, rather than ‘‘at least one.’’ Ap-


ple contends that the correct constructionis ‘‘associating detected structures to com-puter subroutines that cause the CPU toperform a sequence of operations on theparticular structure to which they are as-sociated.’’

We agree with the district court. Appleargues that the claims require only ‘‘asso-ciating’’ between the structure and the su-broutines but ignores that the claims re-cite ‘‘linking.’’ From a general sense, theplain meaning of associating relates to amere commonality, while linking infers ajoining. Additionally, the specificationhere demonstrates that linking is morethan just associating. The patent consis-tently differentiates between associatingand linking and implies that linking is amore specific connection than merely asso-ciating. For example, the specification ex-plains that actions are ‘‘associated’’ withspecific ‘‘grammars’’ or ‘‘patterns,’’ andthat ‘‘linking’’ occurs only after thesegrammars or patterns are ‘‘detected.’’See, e.g., 8647 patent at col. 5, ll. 59–61(‘‘upon detection of a structure based on aparticular pattern, actions associated withthe particular pattern are linked 825 tothe detected structure’’); col. 7, ll. 38–39(‘‘wherein the analyzer server links to adetected structure the actions associatedwith the grammar’’); col. 3, ll. 65–67 (‘‘ana-lyzer server 220 links actions associatedwith the responsible pattern to the detect-ed structure, using conventional pointers’’);col. 5, ll. 31–33 (‘‘analyzer server 220 linksthe actions associated with grammars 410and strings 420 to these identified struc-tures’’) (emphases added).

Apple argues that requiring a ‘‘specifiedconnection’’ limits the claims to the use ofthe ‘‘pointers’’ described in the specifica-tion. The district court explained that apointer is ‘‘a term of art in computer engi-

neering’’ that ‘‘stores a computer memoryaddress.’’ The specification explains thatpointers may be used to link the associatedactions to the detected structures. 8647patent at col. 3, ll. 65–67 (‘‘upon detectionof a structure, analyzer server links ac-tions associated with the responsible pat-tern to the detected structure, using con-ventional pointers’’); col. 4, l. 64—col. 5, l.5 (‘‘[U]pon identification of a structure inthe text, parser links the actions associatedwith the grammar to the identified struc-ture. More particularly, parser retrievesfrom grammar file pointers attached to thegrammar and attaches the same pointersto the identified structure. These pointersdirect the system to the associated actionscontained in associated actions file. Thus,upon selection of the identified structure,user interface can locate the linked ac-tions.’’).

Although the district court stated thatthe specification ‘‘makes clear that linkingis accomplished through pointers,’’ it didnot, as Apple argues, actually limit theclaims to ‘‘pointers.’’ Rather, the courtinterpreted linking to require a ‘‘specifiedconnection,’’ not just a connection estab-lished with the use of pointers. The speci-fication explains that linking may be ac-complished through the use of pointers butdoes not require their use and neither didthe district court. Thus, the districtcourt’s construction comports with thespecification, including the repeated differ-entiation between linking and associatingand the pointers embodiment describedtherein.

Apple is also incorrect that the claimsrequire each structure to be linked withmultiple actions. Apple points to theclaim’s recitation of the plural ‘‘actions.’’See 8647 patent at col. 7, ll. 17–18 (‘‘ananalyzer server for detecting structures in


the data, and for linking actions to thedetected structures ’’) (emphasis added).The plain language of the claims does notrequire multiple actions for each structurebecause the claim recites linking multipleactions to multiple structures. As such,

the plural ‘‘actions’’ may be reasonablyread as at least one action per structure.In fact, Fig. 4 displays an example of theinvention with only one action linked to aspecific structure.

8647 patent at Fig. 4. In Fig. 4, the ‘‘dategrammar’’ structure only has one corre-sponding action, ‘‘put in electronic calen-dar.’’ This directly contradicts Apple’sproposal to require the claims to link mul-tiple actions to each structure. 8647 pat-ent at Fig. 4. Accordingly, we affirm thedistrict court’s construction of ‘‘linking ac-tions to the detected structures.’’


[23] The 8263 patent discloses a systemfor processing data in ‘‘realtime.’’ Theparties dispute whether the ‘‘realtime ap-plication program interface (API)’’ in claim1 must itself function in realtime or wheth-er it must just facilitate realtime process-ing by other subsystems. The districtcourt concluded that the API need justfacilitate realtime processing and con-strued ‘‘realtime API’’ as an ‘‘API thatallows realtime interaction between two ormore subsystems.’’ Motorola argues that

this construction reads ‘‘realtime’’ out ofthe claim. We disagree and affirm thedistrict court’s construction.

Claim 1, with the relevant limitation em-phasized, recites:

A signal processing system for providinga plurality of realtime services to andfrom a number of independent clientapplications and devices, said systemcomprising:

a subsystem comprising a host centralprocessing unit (CPU) operating in ac-cordance with at least one applicationprogram and a device handler program,said subsystem further comprising anadapter subsystem interoperating withsaid host CPU and said device;

a realtime signal processing subsystemfor performing a plurality of data trans-forms comprising a plurality of realtimesignal processing operations; and

at least one realtime application pro-gram interface (API ) coupled between


the subsystem and the realtime signalprocessing subsystem to allow the sub-system to interoperate with said real-time services.

8263 patent at col. 11, ll. 28–43 (emphasisadded). The district court noted that, gen-erally, to be realtime, a system ‘‘must sat-isfy explicitly (bounded) response-timeconstraints or risk severe consequences,’’such as degraded performance.

Motorola contends that the districtcourt’s construction reads ‘‘realtime’’ out ofthe claim because it does not require theAPI itself to function in realtime. Motoro-la points to independent claim 31, whichrecites an API without the ‘‘realtime’’ qua-lifier, and argues that, by including real-time in claim 1, the patentee intended thatthe API itself operate in realtime. SeePhillips, 415 F.3d at 1314 (finding that useof the word ‘‘steel’’ in the term ‘‘steelbaffles’’ ‘‘strongly implies’’ a difference be-tween steel baffles and non-steel baffles).

We agree with the district court. Thespecification describes the API as an inter-face that sends commands and parametersto the ‘‘real-time engine,’’ which actuallyperforms the realtime data processing.The specification does not describe theAPI itself as meeting any specific re-sponse-time constraints or otherwise need-ing realtime functionality. Instead, theAPI’s role is to send commands and pa-rameters to the real-time engine. See,e.g., 8263 patent at col. 6, ll. 33–38 (‘‘eachinterface receives commands from an ap-plication program, through the handler 44,and instructs the real-time engine to carryout the necessary transforms’’); col. 5, ll.22–25 (‘‘the particular transforms to beperformed are sent as commands to thereal-time engine from the adapter handler44 via suitable application programminginterfaces 48’’); col. 10, ll. 40–44 (‘‘in re-

sponse thereto, the API 48 which receivesthese commands supplies the real-time en-gine with the appropriate parameters forperforming the transforms in the requiredformat’’). By contrast, the real-time en-gine is described as performing the actualprocessing, such as ‘‘text-to-speech conver-sion’’ or ‘‘video processing.’’ 8263 patent atabst. (‘‘a data transmission system havinga real-time engine for processing isochro-nous streams of data’’); col. 10, ll. 16–18(‘‘the actual modulation and demodulationof the hardware interface adapter’s isoch-ronous PCM data stream is accomplishedentirely by the real-time engine’’); col. 9,ll. 65–67 (‘‘the handler has no involvementwith the isochronous data stream createdby the real-time engine’’). Thus, althoughthe API interacts with the real-time en-gine, it is the latter that actually performsthe time-constrained processing.

Contrary to Motorola’s argument, thedistrict court did not read ‘‘realtime’’ out ofthe claim. The API is an ‘‘interface.’’ Assuch, it communicates and interacts withother subsystems that process data in real-time without necessarily processing anydata itself. This is what the specificationdescribes and what the district court cor-rectly understood. Further, the districtcourt’s construction does not prevent theAPI from functioning in realtime, it justdoes not require the API to function inrealtime. This is consistent with theclaims and written description, which onlyrequire the API to facilitate the functional-ity of the real-time engine. We affirm thedistrict court’s claim construction.

Motorola’s 8559 Patent

[24] Turning to Motorola’s assertedpatents, the 8559 patent discloses a methodfor generating ‘‘preamble sequences,’’which are used in communications between


cell phones and base stations. The districtcourt construed claim 5 of the 8559 patentto require that the third step (‘‘multiplyingthe outer code by the inner code’’) takeplace only after the first two steps (‘‘form-ing an outer code’’ and ‘‘forming an innercode’’) are completed. Based upon thisconstruction, the court granted summaryjudgment of non-infringement to Apple.Motorola appeals only the claim construc-tion decision. We affirm.

Claim 5, with relevant limitations em-phasized, recites:

A method for generating preamble se-quences in a CDMA system, the methodcomprising the steps of:

forming an outer code in a mobile sta-tion;

forming an inner code in the mobilestation utilizing the following equation:

where Sj, j=0, 1, TTT, M–1 are a set oforthogonal codewords of length P, whereM and P are positive integers; and mul-tiplying the outer code by the inner codeto generate a preamble sequence.

8559 patent at col. 5, ll. 20–35 (emphasisadded).

[25] Steps in a method claim need notnecessarily be performed in the order theyare written. Altiris, Inc. v. SymantecCorp., 318 F.3d 1363, 1369 (Fed.Cir.2003).On the other hand, if grammar, logic, thespecification, or the prosecution history re-quire the steps to be performed sequen-tially, then the claims are so limited. Id.;Loral Fairchild Corp. v. Sony Corp., 181F.3d 1313, 1322 (Fed.Cir.1999) (‘‘Althoughnot every process claim is limited to the

performance of its steps in the order writ-ten, the language of the claim, the specifi-cation and the prosecution history supporta limiting construction in this case.’’). Thedistrict court correctly noted that, while ina preferred embodiment the inner and out-er codes are formed before the multiplica-tion step begins, this alone does not limitthe claims. See, e.g., 8559 patent at Fig. 4.The district court also acknowledged thatthe invention would likely function even ifthe multiplication step began before thefull inner and outer codes were formed.Notwithstanding the foregoing, the courtconstrued the claims to require that stepthree occur only after steps one and twoare completed.

We agree with the district court. Theclaims recite multiplying ‘‘the’’ inner codewith ‘‘the’’ outer code to create a preamble‘‘sequence.’’ Both the inner code and out-er code are sequences of numbers. Theplain meaning of multiplying ‘‘the’’ codestogether is that the entire sequences aremultiplied together after they have beenformed. If claim 17 was directed to amethod that multiplied only parts of theinner and outer code together, it would notrecite multiplying ‘‘the’’ codes together toform the preamble sequence. The morenatural reading of the claim language sup-ports the district court’s finding that theinner and outer codes must be fully formedbefore they are multiplied together.

The specification supports this reading.When discussing multiplying the inner andouter codes, the specification describesforming an inner and outer code and thenmultiplying the codes together:

The present invention provides a meth-od for generating preamble sequences ina CDMA communication system. Themethod comprises forming an outercode and an inner code at a mobile


station. The mobile station then multi-plies the outer code by the inner code togenerate a preamble sequence.

8559 patent at col. 2, ll. 52–57 (emphasisadded).

Because it is supported by the plainmeaning of the claim language and thespecification, we affirm the district court’sclaim construction.

Because this construction was the basisfor the district court’s grant of summaryjudgment of noninfringement, we need notreach the court’s construction of ‘‘a set oforthogonal codewords.’’ Accordingly, weaffirm the district court’s grant of Apple’smotion for summary judgment of non-in-fringement of the 8559 patent based uponits construction that the steps of claim 17be performed in the sequence describedabove.

Motorola’s 8712 patent

[26] The 8712 patent discloses a systemfor encrypting data communications. Thedistrict court found that the claimed‘‘transmit overflow sequence number,’’ orTOSN, is ‘‘never transmitted’’ to the re-ceiver in the claimed system. Because thecounterpart to the TOSN in the accusedproducts is transmitted to the receiver, thedistrict court granted Apple’s motion for

summary judgment of non-infringement.We agree with the district court’s con-struction and therefore affirm its grant ofsummary judgment.

Claim 17 of the 8712 patent, with rele-vant limitations emphasized, recites:

In a communication system having aphysical layer, data link layer, and anetwork layer, a method for providingcryptographic protection of a datastream, comprising:

(a) assigning a packet sequence numberto a packet derived from a data streamreceived from the network layer;

(b) updating a transmit overflow se-quence number as a function of thepacket sequence number; and

(c) encrypting, prior to communicatingthe packet and the packet sequencenumber on the physical layer, the packetas a function of the packet sequencenumber and the transmit overflow se-quence number.

8712 patent at col. 8, l. 65–col. 9, l. 12. Thespecification explains that, in order to en-crypt and decrypt the data being transmit-ted, the invention assigns each ‘‘packet’’ ofdata both a ‘‘packet sequence number’’ andan overflow number. See id. at abst; col.5, ll. 13–15.


The ‘‘key’’ used to encrypt and decryptthe data includes both the packet sequencenumber and the overflow number. Thepacket sequence numbers are assigned se-quentially up to a maximum. Once themaximum is reached, the packet sequencenumber count ‘‘rolls over’’ and beginsagain from number one. For example, ifthe maximum is 128, after 128 is assignedto a packet of data, the next packet isassigned a packet sequence number of one.Id. at col. 4, ll. 15–17. The overflow coun-ter counts the number of times the packetsequence number rolls over. Id. at col. 3,ll. 54–55, 65–68. Thus, as demonstrated inthe above illustration, each time the packetsequence number rolls over, the overflowsequence number increases by one. Id.

On the transmission side, the overflownumber is called the TOSN. On the receiv-ing side, it is called the receiving overflowsequence number (‘‘ROSN’’). In this man-ner, the key to encrypt and decrypt thedata includes both a packet sequence num-ber and an overflow number. When thepackets of data are transmitted, theclaimed system also transmits the packet

sequence number. Id. at col. 5, ll. 29–32;col. 9, ll. 7–9. The specification does notdescribe transmitting the TOSN or anyother overflow number. Id. Instead, thereceiver couples the packet sequence num-ber it receives with a ROSN it generatesand uses these two numbers to decrypt thedata packet.

The specification supports the districtcourt’s construction that the TOSN is nev-er transmitted. First, the claims recite aTOSN, not a generic overflow sequencenumber. The specification plainly de-scribes transmitting certain aspects of thesystem (the data packet, the packet se-quence number) but never suggests thatthe TOSN is part of that transmission.Nor is it necessary for the TOSN to betransmitted. If it were, there would be noneed for the receiver to generate a ROSN.Further, if both the TOSN and packetsequence number were transmitted, aneavesdropper could intercept the entirekey needed to decrypt the message. Thiswould defeat the invention’s purpose ofproviding increased security by not trans-mitting the entire key.


Statements made by Motorola duringprosecution of a related Japanese patentfurther support this construction. Beforethe Japan Patent Office, Motorola distin-guished the prior art by explaining, withthe included emphasis, that ‘‘the overflowsequence number is never transmitted.’’Motorola further explained that, becausethe TOSN is never transmitted, ‘‘there isno chance to intercept the overflow se-quence number; thus, [the invention] pro-vides a higher level of security.’’ Motorolamade this argument on more than oneoccasion. The Japanese application claimspriority to the PCT application that issuedin the United States as the 8712 patent.Both the Japanese application and the 8712patent have the same specification, and theJapanese application included an identicalclaim to claim 17 at the time of Motorola’sstatements.

[27] Our precedent does not preciselyaddress the impact of statements such asMotorola’s here. Motorola’s description ofthe TOSN came after the 8712 patent is-sued and was made in front of a foreignpatent office. This court has previouslyfound that statements made in related,later-prosecuted U.S. patents may informthe meaning of earlier issued claims. See,e.g., Microsoft Corp. v. Multi–Tech Sys.,Inc., 357 F.3d 1340, 1350 (Fed.Cir.2004).In Microsoft, the court noted that ‘‘anystatement of the patentee in the prosecu-tion of a related application as to the scopeof the invention would be relevant to claimconstruction, and the relevance of thestatement made in this instance is en-hanced by the fact that it was made in anofficial proceeding in which the patenteehad every incentive to exercise care incharacterizing the scope of its invention.’’Id. (emphasis added). Based upon thisreasoning, the Microsoft court concluded

that the patentee’s statements made dur-ing the prosecution of a later patent wererelevant to an earlier issued patent thatshared a common specification. Id. Ofcourse, statements made in unrelated ap-plications are not relevant to claim con-struction. See, e.g., Goldenberg v. Cyto-gen, Inc., 373 F.3d 1158, 1167 (Fed.Cir.2004) (explaining that an unrelated patentor application is one that does not ‘‘have afamilial relationship’’ with the patent atissue); Abbott Labs. v. Dey, L.P., 287 F.3d1097, 1105 (Fed.Cir.2002) (finding applica-tions unrelated when they had ‘‘no formalrelationship and were presented to thepatent office as patentably distinct inven-tions’’).

This court has also considered state-ments made before a foreign patent officewhen construing claims if they are rele-vant and not related to unique aspects offoreign patent law. See, e.g., Abbott Labs.v. Sandoz, Inc., 566 F.3d 1282, 1290 (Fed.Cir.2009) (‘‘While statements made duringprosecution of a foreign counterpart to aU.S. patent application have a narrow ap-plication to U.S. claim construction TTT inthis case the JP 8199 application is part ofthe prosecution history of the 8507 patentitself’’) (internal citations omitted); GilletteCo. v. Energizer Holdings, Inc., 405 F.3d1367, 1374 (Fed.Cir.2005) (considering thepatentee’s arguments before the EuropeanPatent Office (EPO) and concluding that a‘‘blatant admission by this same defendantbefore the EPO clearly support[ed]’’ thecourt’s construction); Tanabe Seiyaku Co.,v. U.S. Int’l Trade Comm’n, 109 F.3d 726,733 (Fed.Cir.1997); Caterpillar TractorCo. v. Berco, S.p.A., 714 F.2d 1110, 1116(Fed.Cir.1983). But see AIA Eng’g Ltd. v.Magotteaux Int’l S/A, 657 F.3d 1264, 1279(Fed.Cir.2011) (‘‘[O]ur precedent cautionsagainst indiscriminate reliance on the pros-ecution of corresponding foreign applica-


tions in the claim construction analysis.’’);Pfizer, Inc. v. Ranbaxy Labs. Ltd., 457F.3d 1284, 1290 (Fed.Cir.2006) (‘‘[S]tate-ments made during prosecution of foreigncounterparts to the 8893 patent are irrele-vant to claim construction because theywere made in response to patentabilityrequirements unique to Danish and Euro-pean law.’’).

[28] The principles illustrated in thesedecisions provide ample support for hold-ing Motorola to the statements made dur-ing Japanese prosecution. Motorola’sstatements that the TOSN ‘‘is never trans-mitted’’ to the receiver could not be clear-er. See Gillette, 405 F.3d at 1374 (holdingparty to ‘‘blatant admission’’ in argumentmade to EPO). Motorola also explained,more than once, that not transmitting theTOSN improves the security of theclaimed system. This explanation is con-sistent with the claims and the inventiondescribed in the specification. See AIA,657 F.3d at 1279 (noting concerns regard-ing differences in foreign patent law).Thus, as in Microsoft, the statements were‘‘made in an official proceeding in whichthe patentee had every incentive to exer-cise care in characterizing the scope of itsinvention.’’ 357 F.3d at 1350. Further,the two patents are related and share afamilial relationship. They both claim pri-ority to the same PCT application. Bothspecifications are the same. And, at thetime Motorola made its statements to theJapanese patent office, the Japanese appli-cation contained a claim identical to claim17. Thus, the construction supported bythe specification is also supported by Mo-torola’s statements before the Japan pat-ent office. We hold that it was not errorfor the district court to rely upon Motoro-la’s statements to the Japan Patent Office.As such, we affirm the district court’s con-struction.

The district court granted Apple’s mo-tion for summary judgment of non-in-fringement based upon its construction ofclaim 17. On appeal, Motorola does notdispute that Apple does not infringe underthis construction. Accordingly, we affirmthe district court’s grant of summary judg-ment of non-infringement of the 8712 pat-ent.

ADMISSIBILITY OF DAMAGES

EXPERT EVIDENCE

In preparation for trial on the patentinfringement claims still at issue, the par-ties presented expert testimony supportingtheir damages calculations. Both partieschallenged the admissibility of the otherside’s expert evidence. The district courtheld a Daubert hearing at which the ex-perts testified and the parties presentedoral arguments regarding admissibility.Subsequently, the district court excludednearly all of both parties’ expert evidencerelating to damages for infringement. Notall of these decisions are before us. Onappeal, Apple argues that the proposedtestimony of its damages expert regardingthe 8949 and 8263 patents is admissible.Motorola argues that the proposed testi-mony of its damages expert for the 8898patent is admissible. Because the districtcourt erred in its decision to exclude theparties’ expert evidence, we reverse.

The legal framework for admission ofexpert testimony is provided by the Feder-al Rules of Evidence, along with Daubertv. Merrell Dow Pharm., Inc., 509 U.S. 579,113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) andits progeny. See FED.R.EVID. 702, 703. InDaubert, the Court addressed the properstandard for admitting expert testimonyand emphasized that the focus ‘‘must besolely on principles and methodology, noton the conclusions that they generate.’’


509 U.S. at 595, 113 S.Ct. 2786. In Kum-ho Tire Co. v. Carmichael, 526 U.S. 137,119 S.Ct. 1167, 143 L.Ed.2d 238 (1999), theCourt clarified that the district court’s ‘‘ga-tekeeping obligation’’ applied to all experttestimony. 526 U.S. at 147, 119 S.Ct.1167. Subsequently, ‘‘Rule 702 wasamended in response to Daubert and casesapplying it, including Kumho Tire.’’ MicroChem., Inc. v. Lextron, Inc., 317 F.3d1387, 1391 (Fed.Cir.2003).

Rule 702 states:A witness who is qualified as an expertby knowledge, skill, experience, training,or education may testify in the form ofan opinion or otherwise if:(a) the expert’s scientific, technical, orother specialized knowledge will help thetrier of fact to understand the evidenceor to determine a fact in issue;(b) the testimony is based on sufficientfacts or data;(c) the testimony is the product of reli-able principles and methods; and(d) the expert has reliably applied theprinciples and methods to the facts ofthe case.

FED.R.EVID. 702. Rule 703 states:An expert may base an opinion on factsor data in the case that the expert hasbeen made aware of or personally ob-served. If experts in the particular fieldwould reasonably rely on those kinds offacts or data in forming an opinion onthe subject, they need not be admissiblefor the opinion to be admitted. But ifthe facts or data would otherwise beinadmissible, the proponent of the opin-ion may disclose them to the jury only iftheir probative value in helping the juryevaluate the opinion substantially out-weighs their prejudicial effect.

FED.R.EVID. 703.

[29] Under these rules and precedent,a district court judge, acting as a gatek-

eeper, may exclude evidence if it is basedupon unreliable principles or methods, orlegally insufficient facts and data. See,e.g., Smith v. Ford Motor Co., 215 F.3d713, 718 (7th Cir.2000) (‘‘We emphasizethat the court’s gatekeeping function fo-cuses on an examination of the expert’smethodology.’’); Daubert, 509 U.S. at 595,113 S.Ct. 2786 (‘‘The focus, of course, mustbe solely on principles and methodology,not on the conclusions that they gener-ate.’’); i4i Ltd. Partnership v. MicrosoftCorp., 598 F.3d 831, 854 (Fed.Cir.2010)(stating that ‘‘Daubert and Rule 702 aresafeguards against unreliable or irrelevantopinions, not guarantees of correctness’’)(applying Fifth Circuit law); Walker v. SooLine R. Co., 208 F.3d 581, 587 (7th Cir.2000) (‘‘The critical point is that Dr. Plis-kin employed a proper methodology to de-termine Mr. Walker’s pre-incident IQ.’’).

[30, 31] A judge must be cautious notto overstep its gatekeeping role and weighfacts, evaluate the correctness of conclu-sions, impose its own preferred methodolo-gy, or judge credibility, including the cred-ibility of one expert over another. Thesetasks are solely reserved for the fact find-er. See, e.g., Smith, 215 F.3d at 718 (‘‘Thesoundness of the factual underpinnings ofthe expert’s analysis and the correctnessof the expert’s conclusions based on thatanalysis are factual matters to be deter-mined by the trier of fact.’’); Stollings v.Ryobi Techs., Inc., 725 F.3d 753, 766 (7thCir.2013) (‘‘An expert may provide experttestimony based on a valid and properlyapplied methodology and still offer a con-clusion that is subject to doubt. It is therole of the jury to weigh these sources ofdoubt.’’). As the Seventh Circuit noted inStollings, ‘‘the jury must still be allowed toplay its essential role as the arbiter of theweight and credibility of expert testimo-


ny.’’ 725 F.3d at 765; see also Lees v.Carthage Coll., 714 F.3d 516, 526 (7th Cir.2013).

[32] That the gatekeeping role of thejudge is limited to excluding testimonybased on unreliable principles and methodsis particularly essential in the context ofpatent damages. This court has recog-nized that questions regarding which factsare most relevant or reliable to calculatinga reasonable royalty are ‘‘for the jury.’’ i4i,598 F.3d at 856 (‘‘When the methodology issound, and the evidence relied upon suffi-ciently related to the case at hand, dis-putes about the degree of relevance oraccuracy (above this minimum threshold)may go to the testimony’s weight, but notits admissibility.’’); see also Micro Chemi-cal, 317 F.3d at 1392.

[33] This court has also recognizedthat estimating a ‘‘reasonable royalty’’ isnot an exact science. As such, the recordmay support a range of ‘‘reasonable’’ roy-alties, rather than a single value. Like-wise, there may be more than one reliablemethod for estimating a reasonable royal-ty. See, e.g., In re Innovatio IP Ventures,LLC Patent Litig., MDL 2303, 2013 WL5593609, at *30–40 (N.D.Ill. Oct. 3, 2013)(undertaking a detailed evaluation of thedifferent methods proposed by the partiesof valuing the patents at issue). For ex-ample, a party may use the royalty ratefrom sufficiently comparable licenses, val-ue the infringed features based upon com-parable features in the marketplace, orestimate the value of the benefit providedby the infringed features by a comparingthe accused product to non-infringing al-ternatives. All approaches have certainstrengths and weaknesses and, dependingupon the facts, one or all may produceadmissible testimony in a single case. It is

common for parties to choose different,reliable approaches in a single case and,when they do, the relative strengths andweaknesses may be exposed at trial orattacked during cross-examination. Thatone approach may better account for oneaspect of a royalty estimation does notmake other approaches inadmissible. Forexample, actual royalties paid for technolo-gies similar to the claims at issue mayinherently account for available, non-in-fringing alternatives. On the other hand,an analytical comparison of the same non-infringing alternatives to the claims maymore directly account for this cost. Thefact that one of these methods may be saidto more accurately value this aspect of areasonable royalty calculation does not,necessarily, make the other approach inad-missible.

[34] With this legal framework inmind, we review the district court’s dam-ages rulings. In doing so, we apply Sev-enth Circuit law to the question of admis-sibility. See, e.g., Ethicon, Inc. v. U.S.Surgical Corp., 135 F.3d 1456, 1465 (Fed.Cir.1998). The Seventh Circuit reviews denovo whether the district court applied theproper legal framework; but reviews deci-sions to admit or exclude expert testimonyunder this framework for an abuse of dis-cretion. See United States v. Parra, 402F.3d 752, 758 (7th Cir.2005) (citing UnitedStates v. Allen, 269 F.3d 842, 845 (7thCir.2001)).


[35] As to the 8949 patent, we find thatthe district court improperly excluded Ap-ple’s proposed expert testimony on dam-ages. First, as discussed above, the dis-trict court based its damages analysis onan incorrect claim construction. This er-ror, alone, would require reversal and re-


mand because the erroneous claim con-struction here tainted the district court’sdamages analysis. See, e.g., CohesiveTechs., Inc. v. Waters Corp., 543 F.3d1351, 1373 (Fed.Cir.2008) (vacating rulingthat patentee was not entitled to damagesand remanding for reconsideration basedupon modified claim construction); TexasDigital Sys., Inc. v. Telegenix, Inc., 308F.3d 1193, 1216 (Fed.Cir.2002) (vacatingdamages award where district court pro-vided erroneous claim constructions to thejury). Second, the district court erred bynot considering the full scope of the assert-ed claims, questioning the conclusions ofApple’s expert, and substituting its ownopinion, rather than focusing on the relia-bility of the principles and methods usedor the sufficiency of the facts and datarelied upon. These errors also requirereversal and remand.

In order to estimate the value of theasserted claims of the 8949 patent, Apple’sexpert (Brian W. Napper) analyzed Ap-ple’s Magic Trackpad (‘‘Trackpad’’).2 TheTrackpad is a touchpad for use with Maccomputers in place of a mouse. Much likea touchscreen, it has a tactile sensor thatdetects finger contacts and translates theminto computer commands. A user touchesthe Trackpad with her fingers in order toscroll through documents or translate con-tent on the computer screen. The Track-pad recognizes twelve finger gestures, in-cluding a vertical scrolling gesture and ahorizontal scrolling gesture. The decisionto use the Trackpad as a comparable prod-uct was supported by the testimony of oneof Apple’s technical experts, Dr. StephenFeiner, who concluded that the touch ges-tures of the Trackpad are ‘‘comparable or

similar’’ to the claimed features of the 8949patent. Further, Napper contended thatthe Trackpad features most similar to theasserted claims (such as the vertical scroll-ing and horizontal scrolling features) arethe most used touch features of the Track-pad. Thus, Napper provided factual sup-port for his contention that a Trackpadincludes features comparable to theclaimed features.

In order to isolate the value of theTrackpad’s touch features, Napper beganby comparing the price of the Trackpad tothat of a traditional computer mouse.Napper opined that the price differencebetween these products could be attrib-uted to two differences: the Trackpad’stouch features and its wireless connectivi-ty, neither of which is present in a tradi-tional mouse. Because wireless connectiv-ity is not part of the asserted claims,Napper discounted his royalty based uponthe estimated value of wireless connectivi-ty in this context. Napper estimated thisvalue by comparing the price of two setsof wired and wireless mice and using themidpoint price difference as a discount.

Napper further discounted his royaltycalculation to account for Trackpad touchfeatures that are outside the scope of theasserted claims. Specifically, Napper con-cluded that a subset of the Trackpad touchfeatures were ‘‘similar or comparable’’ tothe asserted features. Thus, he reducedhis royalty estimate by a proportionalamount to account for the extra features.Following this procedure, Napper arrivedat a reasonable royalty.

Napper then compared his royalty calcu-lation with royalty rates paid by Motorola

2. Because much of Napper’s expert report isconfidential, we are limited to providing anoverview of his methodology and application

but cannot include many of the supportingdetails.


in licenses for related touch screen tech-nology. Napper concluded that, becausethe technology at issue in those licenseswas similar but inferior, and because theparties to the license were not direct com-petitors, Motorola would have been willingto pay a higher royalty for a license to theasserted claims of the 8949 patent than ithad paid in those licenses.

As a possible design alternative, Nappertestified that Motorola could have removedthe asserted functionality from its prod-ucts. Napper noted that this would createa less functional touchscreen due to specif-ic technical reasons discussed in Dr. Fein-er’s report. According to Napper, thiswould adversely impact the user’s experi-ence of the touchscreen and create userfrustration, thereby resulting in a less de-sirable product. Napper concluded thatthis alternative would not produce a rea-sonable estimate upon which to base dam-ages.

The district court reviewed Napper’s ex-pert report and excluded all of his pro-posed testimony for the 8949 patent. Inlarge part, the court’s analysis incorporat-ed, and was based upon, its incorrect claimconstruction, which limited the claims toproducts containing a preloaded Kindle ap-plication that uses a tap gesture to go tothe next page. See, e.g., Apple, Inc. v.Motorola, Inc., 1:11–CV–08540, 2012 WL1959560 at *7 (N.D.Ill. May 22, 2012) (‘‘Atthis point in the litigation the dispositiveelement of the 8949 patent is the use of atap on the righthand side of the screen toswitch to the next page of a Kindle bookthat has been loaded on the cell phone’’);see id. at *7–8 (discussing design alterna-tives only to the tap gesture). This overly

narrow view of the claims led the court toconclude, incorrectly, that much of Nap-per’s analysis was too far removed fromthe asserted claims.

Although it nearly exclusively relied onthis narrow claim construction, the districtcourt did recognize that Napper’s expertreport was focused on the full (and proper)scope of the claims. In light of this, thecourt considered whether Napper’s pro-posed testimony provided a reliable basisfor estimating the value of another singleclaim limitation, the ‘‘vertical scrolling fea-ture.’’ 3 The court found that it did not,concluding that ‘‘the fact that many con-sumers will pay more for a Magic Track-pad than for a mouse tells one nothingabout what they will pay to avoid occasion-ally swiping unsuccessfully because theirswiping finger wasn’t actually vertical tothe screen.’’ Thus, the court concludedthat any comparison to a Trackpad was an‘‘inadequate’’ method of estimating the val-ue of either the vertical scrolling feature orthe tap for the next page feature.

Aside from relying on an incorrect claimconstruction, the district court erred bynot considering the full scope of infringe-ment. At no point did the court ask orconsider whether Napper had used reliableprinciples and methods, or sufficient data,to value the entire scope of the assertedclaims. Consequently, the district courtincorrectly focused on individual claim lim-itations in isolation when evaluating thereliability of Napper’s methods. It is cer-tainly conceivable that an otherwise soundmethodology could appear unreliable (or,indeed, irrelevant) when applied to a singlelimitation, or a subset of limitations, ratherthan to the full set of infringed claims.

3. Hence, the court did not consider Napper’sexpert testimony in light of the full scope of

the asserted claims.


This is why the proper inquiry evaluatesthe expert’s methodology in view of thefull scope of the infringed claims. See 35U.S.C. § 284 (‘‘[D]amages adequate tocompensate for the infringement TTT areasonable royalty for the use made of theinvention by the infringer.’’) (emphasisadded); Gen. Motors Corp. v. Devex Corp.,461 U.S. 648, 655, 103 S.Ct. 2058, 76L.Ed.2d 211 (1983) (‘‘Congress expresslyprovided in § 284 that the court ‘shallaward the claimant damages adequate tocompensate for the infringement.’ ’’) (em-phasis in original); Robert A. Matthews,Jr., Annotated Patent Digest § 30:6 (2014)(‘‘All claim constructions define the boundsof what infringes, and therefore also definethe bounds of what activity infringementdamages are applicable to.’’). Thus, theproper question is not whether a compari-son to the Trackpad accurately values oneof the claimed finger gestures. Rather,we must ask, with the entire scope of theasserted claims and the correct claim con-struction in mind, whether Napper em-ployed reliable principles and methods, re-liably applied them, and relied upon legallysufficient facts or data. We find that hedid.

Contrary to the district court’s determi-nation, Napper’s methodology is not inher-ently unsound; rather it is ‘‘the product ofreliable principles and methods.’’ FED.R. EVID. 702. This court has upheld theuse of similar methods involving compara-ble benchmark products in the past. See,e.g., i4i, 598 F.3d at 853–54; see also Vir-netX Inc. v. Apple Inc., 925 F.Supp.2d 816,839 (E.D.Tex.2013). Napper’s methodolo-gy focused on the claimed technology. Asdiscussed, Napper began with an existingproduct containing features he contendedwere similar to the asserted features (theTrackpad). Next, Napper attempted toisolate the value of these similar features

by valuing other, non-claimed features ofthe Trackpad and subtracting this value.To reach a reasonable royalty, Napperthen compared his resulting royalty to re-lated licenses and rationalized the pricedifferences. Napper also addressed thepossibility of removing the asserted fea-tures from the accused devices and arguedthat this would frustrate consumers anddevalue the products. In all, Napper’stestimony was the product of reliable prin-ciples and methods.

Napper also applied these principles andmethods in a reliable way and supportedthem with legally sufficient facts and data.See FED.R.EVID. 702. Napper’s applicationwas straightforward and adequately sup-ported by Apple’s technical experts andNapper’s own experience and expertise,the latter of which is not disputed. Forexample, Napper’s assertion that theTrackpad’s features are comparable to theasserted features is the result of reliableapplication. Like the asserted claims, theTrackpad translates finger contacts intocomputer commands, including some of thesame finger contacts and computer com-mands asserted. Napper also properly re-lied upon Apple’s technical expert to factu-ally support this comparison. To accountfor the differences between the Trackpad’sfeatures and those of the asserted claims,Napper addressed these differences anddiscounted his royalty estimate.

In excluding Napper’s testimony, thedistrict court reasoned, ‘‘that a consumerwill pay something for gestural controldoes not enable an estimation of how muchhe will pay for a particular improvement ina system of such control.’’ The court alsofound that there was a more preferablemethod for valuing the asserted claimsthat, it believed, would be used in a non-litigation context. Contrary to the district


court’s assertion, Napper’s estimate wasnot directed to the value of all ‘‘gesturalcontrol’’ but to the value of the assertedclaims because Napper focused on the as-serted features, as described above. Seesupra at 39–41 (describing Napper’s iden-tification of similar features, isolation oftheir value by discounting for unclaimedfeatures, and comparison of royalty rate tosimilar license agreements). While thedistrict court summarily concluded therewas a better way to calculate damages, itdid not point to any specific flaws in Nap-per’s Trackpad comparison. For example,the court did not question the testimony ofApple’s technical expert that the Track-pad’s features are ‘‘comparable or similar’’to the asserted features. Nor did thecourt point to any flaws in Napper’s meth-od of apportioning the cost of the relevantfeatures in the Trackpad or explain why itspreferred method of valuation was superi-or. Simply because other reliable methodsof estimating a reasonable royalty mayexist does not, by itself, render Napper’sapproach inadmissible.

The district court also faulted Napperfor not directly estimating what a consum-er ‘‘will pay to avoid occasionally swipingunsuccessfully because their swiping fingerwasn’t actually vertical to the screen.’’But Napper’s methodology did estimatewhat consumers will pay for the infringedfeatures by evaluating what consumershave actually paid for comparable features.More generally, the value a consumer at-tributes to the infringing feature may bean important data point for estimating aroyalty, but it is not a required piece of

information in all cases. Here, the ab-sence of a direct measurement of this val-ue does not topple the entirety of theprinciples and methods employed by Nap-per. The statute requires determinationof a ‘‘reasonable royalty,’’ not a reasonableconsumer price.4 As noted, there are mul-tiple reasonable methods for calculating aroyalty, and directly estimating the value aconsumer places on the infringing featureis not a requirement of admissibility.

Factually, if the Trackpad is not an ac-curate benchmark, Motorola is free tochallenge the benchmark or argue for amore accurate benchmark. But such anargument goes to evidentiary weight, notadmissibility, especially when, as here, anexpert has applied reliable methods todemonstrate a relationship between thebenchmark and the infringed claims. SeeLees, 714 F.3d at 526 (‘‘Carthage is free toargue [for] TTT a preferable benchmark,but that again is a matter of evidentiaryweight, not admissibility.’’). Motorola mayaddress any technical differences betweenthe Trackpad and the asserted features,Napper’s analysis thereof, or the correct-ness of Napper’s conclusions, during cross-examination. See Smith v. Ford MotorCo., 215 F.3d 713, 718 (7th Cir.2000) (ex-plaining that disagreements about the fac-tual underpinnings of an expert’s analysisgo to weight, not admissibility). But thedistrict court’s decision on the value esti-mated by Napper short-circuited Motoro-la’s opportunity to rebut Napper’s opinionthrough presentation of its own evidence.See Gayton v. McCoy, 593 F.3d 610, 616(7th Cir.2010) (‘‘Determinations on admis-

4. Depending upon the case, the price a con-sumer would pay for an infringing featuremay not, by itself, shed much light on what aninfringer would pay to license that feature.For example, if we assume that a consumer iswilling to pay $5.00 for a certain feature, an

infringer may want to retain some of that asprofit (and pay only a $3.00 royalty), but,depending upon the circumstances, it may bewilling to sacrifice all of that gain (by payinga $5.00 royalty) and make its profit else-where.


sibility should not supplant the adversarialprocess; ‘shaky’ expert testimony may beadmissible, assailable by its opponentthrough cross-examination.’’) (citing Dau-bert, 509 U.S. at 596, 113 S.Ct. 2786).Whether Napper’s testimony was theproduct of reliable principles and methodsis the focus of admissibility; whether thetestimony produced a correct degree ofestimation of the value of the 8949 patentis a factual consideration reserved for thefact finder. Here, the district court re-solved admissibility based upon its ownview on the correct estimate of value forthe 8949 patent, a question that shouldhave been reserved for the jury.

In sum, the majority of the districtcourt’s analysis relied upon its incorrectclaim construction and is therefore in er-ror. The district court also erred by notconsidering the full scope of infringement,and by questioning the factual underpin-nings and correctness of Napper’s testimo-ny, rather than the reliability of his princi-ples and method and the sufficiency of thedata upon which he relied. The districtcourt’s exclusion of Napper’s expert testi-mony regarding the 8949 patent is there-fore reversed.


The district court also excluded Apple’sdamages evidence related to the 8263 pat-ent because Apple’s expert relied uponinformation provided by a technical experthired by Apple. The district court subse-quently granted Motorola’s motion forsummary judgment that Apple was enti-tled to no damages for infringement of the8263 patent. We reverse the districtcourt’s decision to exclude Apple’s experttestimony and therefore vacate the court’sgrant of summary judgment regarding the8263 patent.

[36] Apple’s damages expert, again Dr.Napper, opined that there were two ap-proaches for designing around the assert-ed claims of the 8263 patent. The firstapproach would require Motorola to rede-sign the computer chip in its phones andalso require the application providers (thecompanies behind products such as Face-book and YouTube) to redesign the appli-cations that utilize streaming video or au-dio content. Napper estimated the cost ofthis approach at the time of infringementbut concluded that this design aroundwould have been practically unworkablefor technical reasons unimportant here.The other design around option identifiedby Napper involved Motorola replacing achip on its phones or adding an additionalchip. To estimate this cost, Napper used achip identified by one of Apple’s technicalexperts, Dr. Nathan Polish. Based uponfacts and data made available to Napperby Polish, Napper concluded that it wouldhave been more expensive to designaround via this second option. Faced withthese choices, Napper arrived at a setfigure which he concluded Motorola wouldhave been willing to pay as a reasonableroyalty because it represented the lesserof the two options.

The district court excluded Napper’sproposed testimony for the sole reasonthat Napper relied upon an expert hiredby Apple, Dr. Polish, for the identity of thereplacement chip. The district court con-cluded that, had Polish been the only per-son competent to opine on the proper sub-stitute chip, then Napper’s testimonywould have been admissible. The courtreasoned:

imagine [a] TTT conversation betweenNapper and Motorola, which I’ll pretendhired Napper to advise on how at lowestcost to duplicate the patent’s functionali-


ty without infringement: Motorola:‘What will it cost us to invent around,for that will place a ceiling on the royal-ty we’ll pay Apple?’ Napper: ‘Braceyourself: $35 million greenbacks.’ Mo-torola: ‘That sounds high; where didyou get the figure?’ Napper: ‘I askedan engineer who works for Apple.’ Mo-torola: ‘Dummkopf! You’re fired.’

Apple, Inc., 2012 WL 1959560 at *9. Basedon this hypothetical conversation, the dis-trict court excluded all of Napper’s pro-posed testimony. Id.

The district court’s exclusion of Nap-per’s proposed testimony was erroneous.The district court’s decision states a rulethat neither exists nor is correct. Expertsroutinely rely upon other experts hired bythe party they represent for expertise out-side of their field. See, e.g., Dura Auto-motive Sys. of Ind., Inc. v. CTS Corp., 285F.3d 609, 609–13 (7th Cir.2002) (‘‘[I]t iscommon in technical fields for an expert tobase an opinion in part on what a differentexpert believes on the basis of expertknowledge not possessed by the first ex-pert.’’); Monsanto Co. v. David, 516 F.3d1009, 1015 (Fed.Cir.2008) (‘‘Numerouscourts have held that reliance on scientifictest results prepared by others may consti-tute the type of evidence that is reasonablyrelied upon by experts for purposes ofRule of Evidence 703.’’). Rule 703 explicit-ly allows an expert to rely on informationhe has been made aware of ‘‘if experts inthe particular field would reasonably relyon those kinds of facts or data in formingan opinion on the subject.’’ FED.R.EVID.

703; see also Monsanto, 516 F.3d at 1016.This Rule does not predicate admissibilityon the source of the facts or data or, inparticular, on whether the source is em-ployed by either of the parties.

Consistent with Rule 703, patent dam-ages experts often rely on technical exper-

tise outside of their field when evaluatingdesign around options or valuing the im-portance of the specific, infringing featuresin a complex device. Monsanto, 516 F.3dat 1015. For example, a party may wantto explain, from a technical perspective,why one potential design alternative is lessexpensive in order to justify a lower royal-ty calculation. Patent damages calcula-tions are often intertwined with highlytechnical issues precisely because damagesmust be based on the scope of infringe-ment, often an involved technical question.Despite this precedent, the district courterroneously found that Apple cannot sup-port its damages calculation with informa-tion from a technical expert it has hired.

In particular, the district court ex-pressed concern that Dr. Polish’s technicaladvice was incurably biased because hewas hired by Apple. See Apple, Inc., 2012WL 1959560 at *9 (Napper ‘‘obtained es-sential information, namely the identity ofthe chip that would avoid infringement,from an agent of the party rather than adisinterested source.’’). While it may betrue that the potential for bias is an inher-ent concern with respect to all hired ex-perts, this concern is addressed by theweight given to the expert’s testimony, notits admissibility. See, e.g., Tagatz v. Mar-quette Univ., 861 F.2d 1040, 1042 (7thCir.1988) (‘‘Hired experts, who generallyare highly compensated—and by the partyon whose behalf they are testifying—arenot notably disinterested.’’); Ethicon, 135F.3d at 1465 (‘‘Furthermore, a witness’specuniary interest in the outcome of a casegoes to the probative weight of testimony,not its admissibility.’’). To the extent biasexists, cross-examination and the testimo-ny of an opposing expert may be used to‘‘expose’’ it. Id.; see also Annotated Pat-ent Digest, at § 44:43.150 (‘‘That an expertmay have a bias or self interest towards a


party on whose behalf he or she is offeringtestimony, without more, generally doesnot require disqualifying the expert solong as the expert appears to be providingan objective analysis. Bias goes to theweight of the testimony, not necessarily itsadmissibility.’’). In Tagatz, the court rec-ognized that, in the context of a partytestifying as his own expert, ‘‘the trier offact should be able to discount for so obvi-ous a conflict of interest.’’ 861 F.2d at1042. Here, if the chip Dr. Polish hassuggested is not the best or cheapest alter-native, Motorola is free to address thisissue on cross examination and throughthe testimony of its own expert witness.See, e.g., Walker, 208 F.3d at 589–90;Stollings, 725 F.3d at 765; Gayton, 593F.3d at 616. The issue is one of evidentia-ry weight and not admissibility.

The district court also implied that Dr.Polish’s recommendation of a particularreplacement chip is not the type of infor-mation that an expert in the field wouldreasonably rely upon outside the context oflitigation. See Apple, Inc., 2012 WL1959560 at *9–10; see also Kumho, 526U.S. at 152, 119 S.Ct. 1167 (noting that anexpert must employ ‘‘in the courtroom thesame level of intellectual rigor that charac-terizes the practice of an expert in therelevant field.’’). We disagree. As notedabove, Rule 703 requires that the experttestimony be tied to the facts of the case.See id. at 153–54, 119 S.Ct. 1167. If Applehired Napper to value the asserted claimsof the 8263 patent outside of litigation andthere was technical issue Napper did notunderstand, it would be reasonable forNapper to ask a technical expert hired byApple. For example, it would be reason-able for an Apple expert to explain thebackground of Apple’s patent, how theclaimed invention fits into a larger prod-uct, or the key advances over the prior art.

All of this information would be outside ofNapper’s expertise but helpful, if not criti-cal, to his task of valuing the patent. In-deed, if the technical information neededwas within the expertise of Apple employ-ees, it would be wasteful for Apple, if nototherwise disadvantageous, not to provideit and, in certain circumstances, even de-ceitful. Overall, outside of litigation, itwould be reasonable, and quite common,for Napper to rely on technical informationprovided by Apple or one of its experts inorder to value the cost to design aroundApple’s technology. Indeed, such an ap-proach would carry the same intellectualrigor as the approach employed in thecourtroom in this case.

A rule that would exclude Apple’s dam-ages evidence simply because it relies uponinformation from an Apple technical expertis unreasonable and contrary to Rules 702and 703 and controlling precedent. As aresult, we reverse the district court’s ex-clusion of Napper’s testimony regardingthe 8263 patent and vacate its grant ofsummary judgment regarding damages.

Motorola’s 8898 Patent

The district court excluded all of Moto-rola’s proposed testimony related to dam-ages for infringement of the 8898 patentand subsequently granted Apple’s motionfor summary judgment that Motorola wasentitled to no damages. We affirm theexclusion of the testimony that relied onMotorola’s licensing expert, Charles Dono-hoe. The district court erred, however,when it excluded the remainder of Motoro-la’s proposed expert testimony becauseMotorola’s analysis of comparable licensesused ‘‘reliable principles and methods,’’and Motorola reliably applied them to suf-ficient facts and data to estimate the over-all value of the 8898 patent. Accordingly,


we reverse the court’s exclusion of theremainder of Motorola’s damages experttestimony and vacate the court’s grant ofsummary judgment.

[37] Motorola’s damages expert wasCarla S. Mulhern. Mulhern’s proposedtestimony included an analysis of licenseagreements between Motorola and ‘‘all ofthe major’’ cellular phone makers in theUnited States, except for Apple. Whilethe terms of each agreement vary, all gen-erally cover Motorola’s entire standard-essential patent (‘‘SEP’’) portfolio (hun-dreds of patents), of which the 8898 patentis a very small part. Some of the licenseswere cross-licenses, wherein Motorola ob-tained a license to third parties’ SEPs. Theroyalty base in each of these licenses wasthe sale price of a cell phone. Accordingto Mulhern, these licenses show that Moto-rola has previously received a royalty rateof approximately 2.25% for a license to itsentire SEP portfolio. Mulhern also exam-ined licenses for SEP portfolios covering‘‘cellular communications technology’’ be-tween Apple and third parties. These li-censes also typically included cross-licenseagreements. Apple’s royalties under theseagreements were in a similar range.

Because Apple is not alleged to infringeall of the patents in Motorola’s SEP port-folio, Mulhern attempted to separate outthe value of the 8898 patent from the totalvalue of the portfolio. Mulhern recognizedthat individual patents in a SEP portfoliomay have different values, based upon, forexample, the relative importance of thepatent to the standard or to the technologyas a whole. Specifically, Mulhern opinedthat economic studies have shown that pat-ent values are skewed with many patentshaving little or no value and a small minor-ity having significant value. Mulhern alsonoted that, during a hypothetical negotia-

tion, a licensee would expect to pay less fora license to a few patents than for anentire portfolio of standard-essential pat-ents. Mulhern concluded, based upon theexpert reports of Motorola’s technical ex-perts, that the 8898 patent represented animportant innovation and was ‘‘dispropor-tionately valuable in the context of Motoro-la’s overall portfolio.’’

Mulhern opined that licensing one pat-ent from a large SEP portfolio was not atypical industry practice because, in orderto use a standard, a license to all essentialpatents would be needed. In the ‘‘realworld,’’ according to Mulhern, SEPs areonly licensed in large portfolios. Mulherncited proposed testimony of other expertswith experience in the licensing industry tosupport this claim. Despite this industrypractice, Mulhern recognized that the‘‘hypothetical negotiation construct wouldforce Motorola to enter into a license forjust a subset of its standard-essential pat-ents.’’ Such a license, according to Mul-hern, would capture a ‘‘nonlinear’’ share ofthe portfolio rate because it would forceMotorola to, inefficiently, license its pat-ents a few at a time. In other words, thefirst few patents licensed from the portfo-lio would command a higher royalty ratethan those that followed.

Based upon this reasoning, Mulhernopined that the first few patents wouldtypically command 40%–50% of the entireportfolio rate. Mulhern supported this 40& % estimate with: the proposed testimo-ny of another of Motorola’s licensing ex-perts, Charles Donohoe; the statements ofa Motorola licensing executive; the state-ments of a former Ericsson licensing exec-utive; and her knowledge of ‘‘IBM’s well-known licensing policy in the 1980s andearly 1990s.’’ According to Mulhern, IBMwould offer a license to a single patent in


its portfolio, which totaled over 10,000 pat-ents, at a royalty rate of 1% of net saleprice. Each successive patent would costan additional 1%, up to the fifth patentlicensed. Beyond five patents, or a 5%royalty, the rate would not increase.Based upon this analysis, Mulhern con-cluded, as one estimate, that the 8898 pat-ent was worth 40%–50% of Motorola’sSEP portfolio rate.

As an alternative and more conservativeestimate, Mulhern opined that the 8898patent represented 5% of the ‘‘actually-essential’’ patents in Motorola’s portfolio.This estimate did not rely upon Mulhern’s‘‘non-linear’’ valuation theory, but ratherwas based upon a linear ‘‘modified patentcounting method.’’ Under this approach,Mulhern concluded that Apple would havebeen willing to pay 5% of the entire SEPportfolio rate for a license to the 8898patent.

Mulhern also recognized that Applecould have avoided infringement by intro-ducing an iPhone only on the Verizon net-work, which did not use the GMS/UTMSnetworks. Mulhern addressed Apple’s de-sign around cost at the time of the hypo-thetical negotiation by first concluding thatthe 8898 patent was essential to the GMS/UTMS mobile networks and that, there-fore, Apple could not use these networkswithout a license. Mulhern outlined pur-ported impracticalities with the Verizonoption, including confidential details we de-cline to recite here. Stopping there, Mul-hern did not specifically estimate the costto Apple of pursuing a Verizon phone inplace of an AT & T phone.

[38] The district court assumed thatthe royalty rates proposed by Motorolasatisfied its fair, reasonable and nondis-criminatory (‘‘FRAND’’) licensing obli-

gations but, nevertheless, excluded all ofMulhern’s proposed testimony because she‘‘failed to consider the range of plausiblealternatives TTT facing Apple.’’ Specifical-ly, the district court concluded that Mul-hern’s ‘‘failure to analyze’’ the possibility ofApple contracting with Verizon made hermethod unreliable. It appears the districtcourt was concerned with Mulhern’s fail-ure to specifically value the cost of con-tracting with Verizon in place of AT & T.See Apple, Inc., 2012 WL 1959560 at *11(‘‘Mulhern offers no evidence that it wouldhave been $347 million more costly to Ap-ple’’ to contract with Verizon over AT &T.). Regarding Donohoe’s proposed testi-mony, the district court noted that hisdeclaration did not offer any basis for se-lecting the correct nonlinear royalty rate.In addition, Donohoe ‘‘admitted that heknows nothing about the portfolio that in-cludes the 8898 patent; his 40–to–50 per-cent figure is a statement about portfoliosof standards-essential telecommunicationspatents in general.’’

We agree with the district court thatDonohoe’s testimony is insufficient to sup-port an award of damages. Indeed, thedistrict court would have been justified inexcluding Donohoe’s testimony as inher-ently unreliable based on his failure to tiethe 40%–50% rate to the technological con-tribution of the patent to the standard-essential patent portfolio. We have consis-tently explained that proof of damagesmust be carefully tied to the claimed in-vention itself. See ResQNet.com, Inc. v.Lansa, Inc., 594 F.3d 860, 869 (Fed.Cir.2010); Riles v. Shell Exploration & Prod.Co., 298 F.3d 1302, 1312 (Fed.Cir.2002)(excluding patentee’s damages model be-cause the expert ‘‘[did] not associate hisproposed royalty with the value of thepatented method at all’’). Here, whereDonohoe admitted that he knew nothing


about the SEP portfolio at issue and didnot even purport to link the 40%–50% rateto the claimed invention of the 8898 patent,there can be little doubt whether his testi-mony was tied to the facts of this case.5

Mulhern incorporated Donohoe’s testi-mony into her own when she relied on hisproposed testimony to opine that the firstfew patents from a given portfolio wouldtypically command 40%–50% of the entireportfolio rate. The substance of Dono-hoe’s testimony is no more reliable whenadmitted through Mulhern than throughDonohoe himself. We note that the gener-al theory that Donohoe and Mulhern reliedupon, that the first patent from a largerportfolio may, in practice, garner a largerroyalty than later patents from the sameportfolio, is not inherently unreliable.Testimony based upon this theory may beadmissible if properly tied to the claimedinvention. But here, where a potentiallyreliable theory is not tied to the facts ofthe case, the expert testimony is inadmissi-ble. See, e.g., Kumho, 526 U.S. at 154–55,119 S.Ct. 1167. For these reasons, weaffirm the district court’s exclusion of thatportion of Mulhern’s testimony that wasbased on Donohoe’s declaration.

[39] We disagree with the districtcourt regarding the remainder of Mul-hern’s testimony. Although Mulhern didnot directly estimate the cost to Apple ofchoosing AT & T over Verizon, Mulherndid ‘‘consider’’ and ‘‘analyze’’ the possibilityof Apple overcoming the difficulties of con-tracting with Verizon and releasing aniPhone on a different network. While thedistrict court may disagree with Mulhern’sconclusions that contracting with Verizonwas not desirable or practical and that,

therefore, this option need not be specifi-cally valued, this is an issue that Applecould address at trial via cross-examina-tion and presentation of its own expertevidence. See, e.g., Walker, 208 F.3d at589–90; Gayton, 593 F.3d at 616. That aparty may choose to pursue one course ofproving damages over another does notrender its expert’s damages testimony in-admissible. Nor is there a requirementthat a patentee value every potential non-infringing alternative in order for its dam-ages testimony to be admissible.

[40] The district court failed to recog-nize that Mulhern did construct a costestimate typically relied upon when calcu-lating patent damages—the cost to licensethe technology. See, e.g., Riles, 298 F.3dat 1313; Studiengesellschaft Kohle, m.b.H.v. Dart Indus., Inc., 862 F.2d 1564, 1568(Fed.Cir.1988) (‘‘The patentee’s usual li-censing approach should be considered inassessing a reasonable royalty.’’). As wehave held many times, using sufficientlycomparable licenses is a generally reliablemethod of estimating the value of a patent.See, e.g., ActiveVideo Networks, Inc. v.Verizon Commc’ns, Inc., 694 F.3d 1312,1333 (Fed.Cir.2012); Lucent Techs., Inc. v.Gateway, Inc., 580 F.3d 1301, 1325 (Fed.Cir.2009) (‘‘The second Georgia–Pacificfactor is ‘[t]he rates paid by the licenseefor the use of other patents comparable tothe patent in suit.’ ’’) (citing Georgia–Pac.Corp. v. U.S. Plywood Corp., 318 F.Supp.1116, 1120 (S.D.N.Y.1970)); Maxwell v. J.Baker, Inc., 86 F.3d 1098, 1110 (Fed.Cir.1996) (‘‘This rate is supported by evidencethat Maxwell entered into agreements withother licensees at a royalty rate of $.10 perpair of shoes.’’); Nickson Indus., Inc. v.

5. Indeed, as Apple points out, Motorola didnot adjust its 40%–50% rate when the 8559patent dropped out of the case on summary

judgment, further suggesting that the rate wasnever tied to the specific patents at issue. SeeApple Reply Br. 38–39.


Rol Mfg. Co., 847 F.2d 795, 798 (Fed.Cir.1988) (‘‘Where an established royalty ex-ists, it will usually be the best measure ofwhat is a ‘reasonable’ royalty.’’).

This approach is generally reliable be-cause the royalty that a similarly-situatedparty pays inherently accounts for marketconditions at the time of the hypotheticalnegotiation, including a number of factorsthat are difficult to value, such as the costof available, non-infringing alternatives.See LaserDynamics, Inc. v. Quanta Com-puter, Inc., 694 F.3d 51, 79 (Fed.Cir.2012)(‘‘Actual licenses to the patented technolo-gy are highly probative as to what consti-tutes a reasonable royalty for those patentrights because such actual licenses mostclearly reflect the economic value of thepatented technology in the marketplace.’’).For example, in this case, Motorola’s otherlicensees would have had the option of onlyreleasing a phone on the Verizon network.Thus, the royalty rate agreed to in theselicenses, assuming the overall licensing sit-uation is factually comparable, would nec-essarily account for the cost of this non-infringing alternative.

Here, whether these licenses are suffi-ciently comparable such that Motorola’scalculation is a reasonable royalty goes tothe weight of the evidence, not its admissi-bility. See ActiveVideo, 694 F.3d at 1333.Mulhern’s analysis of comparable licensesused ‘‘reliable principles and methods,’’and reliably applied them to sufficientfacts and data on the record to calculatethe overall value of the 8898 patent, includ-ing the cost of non-infringing alternatives.FED.R.EVID. 702. With the aid of Apple’scross-examination and expert testimony,the jury is capable of assigning the appro-priate weight to Mulhern’s testimonybased upon their judgment of her credibili-ty, factual analysis, and conclusions.

In sum, with the exception of the testi-mony that relied upon Donohoe’s declara-tion, we reverse the district court’s exclu-sion of Mulhern’s proposed testimony.Consequently, we vacate the court’s grantof summary judgment regarding damagesfor infringement.

SUMMARY JUDGMENT

After the district court excluded the ma-jority of the damages expert testimony,both parties moved for summary judgmentregarding damages and injunctive relief.Having excluded their damages evidence,the court found that neither party wasentitled to any damages or an injunctionand granted the motions. As noted above,on the basis of our reversal of the districtcourt’s admissibility decisions, we vacatethe court’s grant of summary judgmentregarding damages for Apple’s 8949 and8263 patents and for Motorola’s 8898 pat-ent. All that remains is the district court’sgrant of summary judgment that Applewas not entitled to any damages for in-fringement of the 8647 patent and thatneither party was entitled to an injunction.As discussed below, we reverse the court’sdecision that Apple was not entitled to anydamages for infringement of the 8647 pat-ent, vacate its decision that Apple was notentitled to an injunction, and affirm itsdecision that Motorola is not entitled to aninjunction.

Summary judgment is appropriate ‘‘ifthe pleadings, depositions, answers to in-terrogatories, and admissions on file, to-gether with the affidavits, if any, show thatthere is no genuine issue as to any materi-al fact and that the moving party is enti-tled to a judgment as a matter of law.’’FED.R.CIV.P. 56(c). We review the districtcourt’s grant of summary judgment underregional circuit law, which is that of the


Seventh Circuit. See, e.g., MicroStrategy,Inc. v. Bus. Objects, S.A., 429 F.3d 1344,1349 (Fed.Cir.2005). The Seventh Circuitreviews a grant of summary judgment denovo. See, e.g., Feliberty v. Kemper Corp.,98 F.3d 274, 276 (7th Cir.1996) (‘‘When wereview a district court’s grant of summaryjudgment, we consider the record accord-ing to the same standards employed by thedistrict court.’’).


Motorola moved for summary judgmentthat, even assuming the 8647 patent wasinfringed, Apple was not entitled to anydamages, including a nominal reasonableroyalty. Motorola argued that Apple wasnot entitled to damages because there was‘‘no evidence upon which Apple may rely toreliably establish or measure any amountof damages’’ for infringement of the 8647patent.

[41] The district court agreed and con-cluded that Apple was not entitled to anymeasure of damages because Apple hadfailed to show that its measure of damageswas correct. We disagree and hold that afinding that a royalty estimate may sufferfrom factual flaws does not, by itself, sup-port the legal conclusion that zero is areasonable royalty. See Dow Chem. Co. v.Mee Indus., Inc., 341 F.3d 1370, 1382, n. 4(Fed.Cir.2003) (‘‘However, the fact that theagreement did not support the specificamount requested by Dow does not there-by mean that it does not support anyaward at all.’’).

Due to the procedural posture in thiscase, we must assume that the patents atissue are valid and infringed. With in-

fringement assumed, the statute requiresthe court to award damages ‘‘in no eventless than a reasonable royalty.’’ 35 U.S.C.§ 284; see also Dow Chemical Co., 341F.3d at 1381–82 (noting that a districtcourt is obligated to award damages onceinfringement is found, and reversing afinding of no damages that was based upona lack of expert evidence). Because noless than a reasonable royalty is required,the fact finder must determine what royal-ty is supported by the record. See NorianCorp. v. Stryker Corp., 363 F.3d 1321, 1333(Fed.Cir.2004) (holding that ‘‘the jury’sfinding of no damages cannot be sup-ported’’ because ‘‘the statute requires thatdamages to a successful claimant in a pat-ent infringement suit shall not be less thana reasonable royalty’’); Riles, 298 F.3d at1311 (‘‘The statute guarantees patentees areasonable royalty even when they areunable to prove entitlement to lost profitsor an established royalty rate.’’); Linde-mann Maschinenfabrik GmbH v. Am.Hoist & Derrick Co., 895 F.2d 1403, 1406(Fed.Cir.1990) (‘‘In patent law, the fact ofinfringement establishes the fact of dam-age because the patentee’s right to excludehas been violated.’’); Annotated PatentDigest, § 30:7 (‘‘When a patentee showsinfringement, a presumption arises thatthe patentee is entitled to some form ofdamages. The Federal Circuit has ex-plained that this presumption arises fromthe statute once infringement is admittedor proven.’’).

[42, 43] If a patentee’s evidence failsto support its specific royalty estimate, thefact finder is still required to determinewhat royalty is supported by the record.6

See Dow Chem., 341 F.3d at 1381–82

6. By contrast, if a patentee seeks lost profitsand fails to meet its burden, then an award ofno lost profits is justified. See R.R. Dynamics,

Inc. v. A. Stucki Co., 727 F.2d 1506, 1520(Fed.Cir.1984) (affirming district court deci-sion awarding no lost profits, due to lack of


(‘‘The statute is unequivocal that the dis-trict court must award damages in anamount no less than a reasonable royal-ty.’’); Norian, 363 F.3d at 1333; Linde-mann Maschinenfabrik, 895 F.2d at 1406;Del Mar Avionics, Inc. v. Quinton Instru-ment Co., 836 F.2d 1320, 1327 (Fed.Cir.1987) (‘‘The requirement to determine ac-tual damages is not diminished by difficul-ty of determination.’’). Indeed, if the rec-ord evidence does not fully support eitherparty’s royalty estimate, the fact findermust still determine what constitutes areasonable royalty from the record evi-dence. See SmithKline Diagnostics, Inc.v. Helena Labs. Corp., 926 F.2d 1161,1167–68 (Fed.Cir.1991) (‘‘[T]he factual de-termination of a reasonable royalty, how-ever, need not be supported, and indeed,frequently is not supported by the specificfigures advanced by either partyTTTT

[T]he district court may reject the ex-treme figures proffered by the litigants asincredible and substitute an intermediatefigure as a matter of its judgment from allof the evidence.’’). Certainly, if the paten-tee’s proof is weak, the court is free toaward a low, perhaps nominal, royalty, aslong as that royalty is supported by therecord. See Lindemann Maschinenfa-brik, 895 F.2d at 1407–08 (affirming$10,000 reasonable royalty award becausepatentee failed to provide sufficient evi-dence to support a greater award).

[44] Thus, a fact finder may award nodamages only when the record supports azero royalty award. For example, in acase completely lacking any evidence onwhich to base a damages award, the recordmay well support a zero royalty award.Also, a record could demonstrate that, atthe time of infringement, the defendantconsidered the patent valueless and thepatentee would have accepted no paymentfor the defendant’s infringement. Ofcourse, it seems unlikely that a willinglicensor and willing licensee would agree toa zero royalty payment in a hypotheticalnegotiation, where both infringement andvalidity are assumed.7

[45, 46] At summary judgment, as isthe case here, a judge may only award azero royalty for infringement if there is nogenuine issue of material fact that zero isthe only reasonable royalty. Thus, if apatentee raises a factual issue regardingwhether it is due any non-zero royalty,summary judgment must be denied. Inany event, simply because a patentee failsto show that its royalty estimate is correctdoes not, by itself, justify awarding a roy-alty of zero at summary judgment, as thedistrict court did here. See Dow Chem.,341 F.3d at 1382 n. 4.

[47] On appeal, Apple points out thatnot all of its damages expert testimony for

proof, but awarding reasonable royalty); Gus-tafson, Inc. v. Intersystem Indus. Prods., Inc.,897 F.2d 508, 509–10 (1990) (finding, in thelost profits context, ‘‘no reversible error in thedistrict court’s TTT awarding no damages toGustafson because none were proven’’). Thisis because a lost profits award is above andbeyond the reasonable royalty floor requiredby the statute.

7. We note that we know of no case where wefound that the record supported an infringe-ment award of a zero royalty. Id.; see alsoDePuy Spine, Inc. v. Medtronic Sofamor Da-nek, Inc., 567 F.3d 1314, 1334 (Fed.Cir.2009)

(finding 0% jury royalty award inconsistentwith finding of infringement and jury instruc-tion that the jury choose a rate between 6%and 15%). But see Devex Corp. v. Gen. MotorsCorp., 667 F.2d 347, 361 (3d Cir.1981) (‘‘Inthe absence of any evidence as to what wouldconstitute a reasonable royalty in a givencase, a fact finder would have no means ofarriving at a reasonable royalty, and nonecould be awarded.’’), aff’d on other grounds,461 U.S. 648, 103 S.Ct. 2058, 76 L.Ed.2d 211(1983).


the 8647 patent was excluded. Specifically,Apple’s expert report, prepared by Nap-per, estimated a royalty based upon thetime it allegedly took another smartphonemanufacturer, HTC Corporation (‘‘HTC’’),to design around the same patent. TheUnited States International Trade Com-mission (‘‘ITC’’) issued an exclusion orderagainst HTC based upon infringement ofthe 8647 patent, the same patent Appleasserts here. The ITC gave HTC a four-month grace period before the exclusionorder took effect. Based upon this, Nap-per concluded that four months was a rele-vant timeframe for designing around the8647 patent. Napper noted that HTC andMotorola would likely have similar re-sources to develop a design around. Nap-per estimated that Motorola would lose$52 million in operating profit based uponthe removal of its accused products fromthe market for this four-month period.Accordingly, Napper concluded that Moto-rola would be willing to pay a $52 millionroyalty.

The district court noted that it did notexclude this aspect of Napper’s proposedtestimony. Nevertheless, the court dis-credited the testimony because the ITC’sclaim construction differed from thecourt’s and because Napper failed to suffi-ciently address any differences betweenthe two companies, HTC and Motorola.The court did not, however, identify anydifferences between HTC and Motorolathat undermined Napper’s royalty esti-mate. The court determined that Applewas not entitled to any measure of dam-ages because it had not proven its HTC-based royalty estimate was correct. SeeApple, Inc. v. Motorola, Inc., 869

F.Supp.2d 901, 911 (N.D.Ill.2012) (‘‘Applehas not presented admissible evidence thatthe Georgia–Pacific factors support itsdamages claim.’’).

The record before us does not supportgranting summary judgment of no dam-ages. Motorola has not met its burden ofdemonstrating, as the party who moved forsummary judgment, that the record is un-controverted that zero is the only reason-able royalty.

Other than incorrectly asserting thatApple has no admissible expert evidence,Motorola itself provides no evidence orargument to support a zero royalty. Forexample, there is nothing in the recordsuggesting that Apple would have beenwilling to accept no payment for Motoro-la’s infringement. Nor is there any evi-dence that, at the time of infringement,Motorola concluded that the 8647 patenthad no value. To the contrary, Apple hasexpert evidence on record indicating that itwould take four months to design aroundthe 8647 patent. Apple’s expert estimatedthe profit lost by Motorola during thisdesign around period based upon Motoro-la’s sales figures for the accused devices.Apple also presented evidence, in arguingfor an injunction, that the claimed featuresof the 8647 patent drive consumer demandfor its products. The specific reasonableroyalty this—and any other record evi-dence—supports is a factual question re-served for the jury.

The problems that the district courtidentifies with Apple’s HTC-based royaltyestimate (a different claim construction be-fore the ITC and potential differences be-tween HTC and Motorola) only serve toillustrate that factual disputes remain.8

Hence, the district court’s conclusion that

8. Motorola tacitly acknowledges this by argu-ing on appeal that Apple’s evidence is inad-

missible, not that it is factually flawed.


Apple’s HTC-based estimate was flaweddoes not establish, as a matter of law, azero royalty.9 See Norian, 363 F.3d at1333; Dow Chem., 341 F.3d at 1382 n. 4;Del Mar, 836 F.2d at 1327 (‘‘The trialcourt is required to approximate, if neces-sary, the amount to which the patent own-er is entitled.’’). Even if Apple had notsubmitted expert evidence, this alonewould not support a finding that zero is areasonable royalty. See Dow Chem., 341F.3d at 1381–82 (reversing award of nodamages, which was based upon a lack ofadmissible expert evidence, and remandingfor determination of royalty); 7 Donald S.Chisum, Chisum on Patents § 20.07[3][a](2011) (‘‘In considering the question of evi-dence and certainty of proof required tojustify an award of a reasonable royalty,one should distinguish between (1) the ex-istence of damage to a patentee which willsupport any award and (2) the evidentiarysupport for a particular rate or quantity asa reasonable royalty. As to the first issue,no specific proof should be required. Thepremise of the reasonable royalty measureis that a holder of a valid and infringedpatent has inherently suffered legal dam-age at least to the extent of a lost licenseroyalty opportunity.’’); Annotated PatentDigest, § 44:79 (‘‘The use of expert testi-mony is permissive. Thus, if a patentee’sdamage expert is excluded, that fact doesnot automatically deny a patentee a rightto recover damages.’’); FED.R.EVID. 702(‘‘A witness who is qualified as an expertby knowledge, skill, experience, training,or education may testify.’’) (emphasis add-ed); 35 U.S.C. § 284 (‘‘The court mayreceive expert testimony as an aid to thedetermination of damages or of what roy-alty would be reasonable under the cir-cumstances.’’) (emphasis added).

In sum, Motorola has not demonstratedthat there is no genuine issue of materialfact regarding whether zero is a reason-able royalty for infringement of the 8647patent. Motorola presented no evidencethat zero was a reasonable royalty. Incontrast, Apple presented admissible evi-dence that it is entitled to a non-zero roy-alty. That Apple’s royalty estimate maysuffer from factual flaws does not, by it-self, support the legal conclusion that zerois a reasonable royalty. Accordingly, wereverse the district court’s grant of sum-mary judgment.

Apple’s Request for an Injunction

[48] The district court granted Moto-rola’s motion for summary judgment thatApple was not entitled to an injunction forinfringement of the 8949, 8263 and 8647patents because Apple had failed to show acausal nexus between any alleged irrepara-ble harm and Motorola’s infringement.The district court’s analysis necessarily re-lied upon its overly narrow construction ofthe claims of the 8949 patent. We havereversed the court’s claim construction de-cision, thus altering the potential scope ofinfringement underlying Apple’s injunctionrequest. The scope of infringement per-meates an injunction analysis. See, e.g.,Apple Inc. v. Samsung Elecs. Co., 735 F.3d1352, 1363 (Fed.Cir.2013) (‘‘[T]he purposeof the causal nexus requirement is to showthat the patentee is irreparably harmed bythe infringement.’’) (emphasis in original).For example, when considering whether toenjoin a product, it is proper for the courtto consider the aggregate harm caused byall of the infringing features, rather thanrequiring a patentee to address each pat-

9. We would be answering a different questionif Motorola had moved that Napper’s HTC-based calculation was incorrect as a matter of

law. In that situation, pointing to flaws inNapper’s analysis could be enough to justifysummary judgment.


ent or claim individually. See id. at 1365(‘‘We believe there may be circumstanceswhere it is logical and equitable to viewpatents in the aggregate.’’).

Infringement of multiple patents by asingle device may strengthen a patentee’sargument for an injunction by, for exam-ple, supporting its argument that the in-fringed features drive consumer demandor are causing irreparable harm. Id. Bythe same token, we also consider the im-pact on the general public of an injunctionon a product with many non-infringingfeatures. See id. at 1372–73 (‘‘We see noproblem with the district court’s decision,in determining whether an injunctionwould disserve the public interest, to con-sider the scope of Apple’s requested in-junction relative to the scope of the pat-ented features and the prospect that aninjunction would have the effect of depriv-ing the public of access to a large numberof non-infringing features.’’).

Accordingly, because we have reversedthe district court’s construction of the 8949patent, we vacate its grant of summaryjudgment regarding Apple’s request for aninjunction.

Motorola’s Request for an Injunction

[49] Apple moved for summary judg-ment that Motorola was not entitled to aninjunction for infringement of the 8898 pat-ent. While we review the district court’sdecision to grant or deny an injunction foran abuse of discretion, Lab. Corp. of Am.Holdings v. Chiron Corp., 384 F.3d 1326,1331 (Fed.Cir.2004), we review the districtcourt’s grant of summary judgment denovo, Feliberty, 98 F.3d at 276. The Su-preme Court in eBay Inc. v. MercEx-change, L.L.C. outlined the factors a dis-trict court should consider before issuing

an permanent injunction, stating that ‘‘aplaintiff must demonstrate: (1) that it hassuffered an irreparable injury; (2) thatremedies available at law, such as mone-tary damages, are inadequate to compen-sate for that injury; (3) that, consideringthe balance of hardships between theplaintiff and defendant, a remedy in equityis warranted; and (4) that the public inter-est would not be disserved by a permanentinjunction.’’ 547 U.S. 388, 391, 126 S.Ct.1837, 164 L.Ed.2d 641 (2006).

The 8898 patent is a SEP and, thus,Motorola has agreed to license it on fair,reasonable, and nondiscriminatory licens-ing (‘‘FRAND’’) terms. The district courtgranted Apple’s motion, stating:

I don’t see how, given FRAND, I wouldbe justified in enjoining Apple from in-fringing the 8898 unless Apple refuses topay a royalty that meets the FRANDrequirement. By committing to licenseits patents on FRAND terms, Motorolacommitted to license the 8898 to anyonewilling to pay a FRAND royalty andthus implicitly acknowledged that a roy-alty is adequate compensation for a li-cense to use that patent. How could itdo otherwise? How could it be permit-ted to enjoin Apple from using an inven-tion that it contends Apple must use if itwants to make a cell phone with UMTStelecommunications capability—withoutwhich it would not be a cell phone.

Apple, Inc., 869 F.Supp.2d at 913–14 (em-phases in original).

To the extent that the district courtapplied a per se rule that injunctions areunavailable for SEPs, it erred. While Mo-torola’s FRAND commitments are certain-ly criteria relevant to its entitlement to aninjunction, we see no reason to create, assome amici urge, a separate rule or ana-


lytical framework for addressing injunc-tions for FRAND-committed patents. Theframework laid out by the Supreme Courtin eBay, as interpreted by subsequent de-cisions of this court, provides amplestrength and flexibility for addressing theunique aspects of FRAND committed pat-ents and industry standards in general.547 U.S. at 391–94, 126 S.Ct. 1837. Apatentee subject to FRAND commitmentsmay have difficulty establishing irrepara-ble harm. On the other hand, an injunc-tion may be justified where an infringerunilaterally refuses a FRAND royalty orunreasonably delays negotiations to thesame effect. See, e.g., U.S. Dep’t of Jus-tice and U.S. Patent and Trademark Of-fice, Policy Statement on Remedies forStandard–Essential Patents Subject toVoluntary F/RAND Commitments, at 7–8(Jan. 8, 2013). To be clear, this does notmean that an alleged infringer’s refusal toaccept any license offer necessarily justi-fies issuing an injunction. For example,the license offered may not be on FRANDterms. In addition, the public has an in-terest in encouraging participation in stan-dard-setting organizations but also in en-suring that SEPs are not overvalued.While these are important concerns, thedistrict courts are more than capable ofconsidering these factual issues when de-ciding whether to issue an injunction underthe principles in eBay.

Applying those principles here, we agreewith the district court that Motorola is notentitled to an injunction for infringementof the 8898 patent. Motorola’s FRANDcommitments, which have yielded manylicense agreements encompassing the 8898patent, strongly suggest that money dam-ages are adequate to fully compensate Mo-torola for any infringement. Similarly,Motorola has not demonstrated that Ap-ple’s infringement has caused it irrepara-

ble harm. Considering the large numberof industry participants that are alreadyusing the system claimed in the 8898 pat-ent, including competitors, Motorola hasnot provided any evidence that adding onemore user would create such harm.Again, Motorola has agreed to add asmany market participants as are willing topay a FRAND royalty. Motorola arguesthat Apple has refused to accept its initiallicensing offer and stalled negotiations.However, the record reflects that negotia-tions have been ongoing, and there is noevidence that Apple has been, for example,unilaterally refusing to agree to a deal.Consequently, we affirm the districtcourt’s grant of summary judgment thatMotorola is not entitled to an injunctionfor infringement of the 8898 patent.

AFFIRMED–IN–PART, REVERSED–IN–PART, VACATED–IN–PART, ANDREMANDED

COSTS

Each party shall bear its own costs.

RADER, Chief Judge, dissenting-in-part.

I join the court’s opinion in its entirety,except for the affirmance of the districtcourt’s denial of Motorola’s request for aninjunction. To my eyes, the record con-tains sufficient evidence to create a genu-ine dispute of material fact on Apple’sposture as an unwilling licensee whosecontinued infringement of the 8898 patentcaused irreparable harm. Because of theunique and intensely factual circumstancessurrounding patents adopted as industrystandards, I believe the district court im-properly granted summary judgment.Therefore, on this narrow point, I respect-fully dissent in part.


At the outset, a patent adopted as astandard undoubtedly gains value by vir-tue of that adoption. This enhancementcomplicates the evaluation of the technolo-gy independent of the standardization. Bythe same token, the standardization deci-sion may also simply reflect and validatethe inherent value of the technology ad-vance accomplished by the patent. Untan-gling these value components (at the heartof deciding whether a putative licensee was‘‘unwilling’’ to license, and thus irreparableharm and other injunction factors) re-quires intense economic analysis of com-plex facts. In sum, right from the theoret-ical outset, this question is not likely to besusceptible to summary adjudication.

In reciting the legal principles for aninjunction, this court accurately states theinquiries. Those principles supply no perse rule either favoring or proscribing in-junctions for patents in any setting, letalone the heightened complexity of stan-dardized technology. This court notesthat a patent owner in a standard context‘‘may have difficulty establishing irrepara-ble harm TTTT [but] an injunction may bejustified where an infringer unilaterally re-fuses a FRAND royalty or unreasonablydelays negotiations to the same effect.’’Majority Op. 1332 (citing U.S. Dep’t ofJustice and U.S. Patent and TrademarkOffice, Policy Statement on Remedies forStandard Essential Patents Subject toVoluntary F/RAND Commitments, at 7–8(Jan. 8, 2013)).

Market analysts will no doubt observethat a ‘‘hold out’’ (i.e., an unwilling licenseeof an SEP seeking to avoid a license basedon the value that the technological advancecontributed to the prior art) is equally aslikely and disruptive as a ‘‘hold up’’ (i.e., anSEP owner demanding unjustified royal-ties based solely on value contributed by

the standardization). These same complexfactual questions regarding ‘‘hold up’’ and‘‘hold out’’ are highly relevant to an injunc-tion request. In sum, differentiating ‘‘holdup’’ from ‘‘hold out’’ requires some factualanalysis of the sources of value—the inven-tive advance or the standardization.

The record in this case shows evidencethat Apple may have been a hold out. See,e.g., Apple, Inc. v. Motorola, Inc., 869F.Supp.2d 901, 914 (N.D.Ill.2012); Appel-lees’ Br. 64–65, 72–73; Appellees’ ReplyBr. 26–27; J.A. 118884–86. This evidencealone would create a dispute of materialfact.

More important, the district court madeno effort to differentiate the value due toinventive contribution from the value dueto standardization. Without some atten-tion to that perhaps dispositive question,the trial court was adrift without a map,let alone a compass or GPS system. Infact, without that critical inquiry, the dis-trict court could not have properly appliedthe eBay test as it should have.

Instead of a proper injunction analysis,the district court effectively consideredMotorola’s FRAND commitment as dis-positive by itself: ‘‘Motorola committed tolicense the 8898 to anyone willing to pay aFRAND royalty and thus implicitly ac-knowledged that a royalty is adequatecompensation for a license to use that pat-ent. How could it do otherwise?’’ Apple,869 F.Supp.2d at 914. To the contrary,Motorola committed to offer a FRANDlicense, which begs the question: What isa ‘‘fair’’ and ‘‘reasonable’’ royalty? If Mo-torola was offering a fair and reasonableroyalty, then Apple was likely ‘‘refus[ing] aFRAND royalty or unreasonably de-lay[ing] negotiations.’’ See Majority Op.1332. In sum, the district court could notduck the question that it did not address;


was Motorola’s FRAND offer actuallyFRAND?

Furthermore, the district court acknowl-edged the conflicting evidence about Ap-ple’s willingness to license the 8898 patent:‘‘Apple’s refusal to negotiate for a license(if it did refuse—the parties offer compet-ing accounts, unnecessary for me to re-solve, of why negotiations broke down) wasnot a defense to a claim by Motorola for aFRAND royalty.’’ Apple, 869 F.Supp.2dat 914. Yet this scenario, adequately pre-sented by this record, is precisely one thatthe court today acknowledges may justifyan injunction.

In my opinion, the court should haveallowed Motorola to prove that Apple wasan unwilling licensee, which would stronglysupport its injunction request. The courtstates that ‘‘the record reflects that negoti-ations have been ongoing,’’ Majority Op.1332; but, as the district court even ac-knowledged, Motorola asserts otherwise—that Apple for years refused to negotiatewhile nevertheless infringing the 8898 pat-ent, see, e.g., Appellees’ Br. 64–65, 72–73;Appellees’ Reply Br. 26–27. Motorolashould have had the opportunity to proveits case that Apple’s alleged unwillingnessto license or even negotiate supports ashowing that money damages are inade-quate and that it suffered irreparableharm. The district court refused to devel-op the facts necessary to apply eBay as itshould have. Consequently, the caseshould be remanded to develop that rec-ord. For these reasons, I respectfully dis-sent in part.

PROST, Circuit Judge, concurring-in-part and dissenting-in-part.

I join the majority opinion with respectto many of the issues discussed therein.

However, I respectfully dissent with re-spect to the proper construction of the‘‘heuristic’’ claim terms in the 8949 patent.Additionally, I dissent from the majority’sdecision to vacate the district court’s grantof summary judgment regarding Apple’srequest for an injunction, as I would affirmthat decision. As I also depart from themajority’s reasoning while reaching thesame result on several other issues, I writeseparately to explain my views.

I. CLAIM CONSTRUCTION

OF THE 8949 PATENT

A. ‘‘Means–Plus–Function’’

The majority concludes that the ‘‘heuris-tic’’ claim terms in the 8949 patent havesufficiently definite structure to avoidmeans-plus-function treatment. MajorityOp. 1304. I disagree.

As an initial matter, the majority mis-states our law on means-plus-functionclaiming. Generally speaking, a means-plus-function analysis proceeds in twophases: first, the court must determinewhether the claim term is drafted inmeans-plus-function format such that 35U.S.C. § 112 ¶ 6 applies. See KemcoSales, Inc. v. Control Papers Co., 208 F.3d1352, 1361 (Fed.Cir.2000) (‘‘Before a courtattempts to analyze what appears to be ameans-plus-function claim limitation, itmust first assure itself that such a claimlimitation is at issue.’’). Only then shouldthe court undertake to construe the disput-ed claim term by identifying the ‘‘corre-sponding structure, material, or acts de-scribed in the specification’’ to which theclaim term will be limited. Welker Bear-ing Co. v. PHD, Inc., 550 F.3d 1090, 1097(Fed.Cir.2008) (‘‘Because ‘mechanism formoving said finger’ is a means-plus-func-tion limitation, this court must next exam-


ine the trial court’s identification of ‘thecorresponding structure, material, or actsdescribed in the specification and equiva-lents thereof.’ ’’).

Although the majority accurately statesthese legal principles, see Majority Op.1296, it fails to faithfully apply them.Rather, the majority’s analysis collapsesthese two steps into one, and in doing so, iteffectively renders the category of non-indefinite means-plus-function claim termsa null set.

The majority correctly states that theproper inquiry in determining whether aclaim term is drafted in means-plus-func-tion format is whether the limitation has‘‘sufficiently definite structure.’’ MajorityOp. 1297–98. The majority then identifiesseveral ways in which a term may bedeemed to have such structure, such aswhen it has ‘‘a structural definition that iseither provided in the specification or gen-erally known in the art,’’ when it ‘‘out-lin[es] an algorithm, a flowchart, or a spe-cific set of instructions or rules,’’ or whenthe specification ‘‘describ[es] the claim lim-itation’s operation, such as its input, out-

put, or connections.’’ Id. at 1298–99. Bycontrast, the majority states, ‘‘if the claimmerely recites a generic nonce word andthe remaining claim language, specifica-tion, prosecution history, and relevant ex-ternal evidence provide no further struc-tural description to a person of ordinaryskill in the art, then the presumptionagainst means-plus-function claiming is re-butted.’’ Id. at 1300.

In effect, what the majority has done isimported the second step of the analysis(where you define the scope of a means-plus-function claim term based on the cor-responding structure in the specification)into the first step (where you identifywhether the term is drafted in means-plus-function format). The majority’s analysisimplies that so long as a claim term hascorresponding structure in the specifica-tion, it is not a means-plus-function limita-tion.1 But such a rule would render everymeans-plus-function claim term indefinite.Under the majority’s approach, a termwould only be deemed a means-plus-func-tion limitation if it has no correspondingstructure—an absurd result that wouldeviscerate means-plus-function claiming.2

1. Admittedly, the majority suggests that ‘‘it ispossible to find that a claim limitation doesnot connote sufficiently definite structure de-spite the presence of some correspondingstructure in the specification.’’ Majority Op.1297. But the majority’s actual analysis ofhow to identify a sufficiently definite structureto avoid means-plus-function treatment sug-gests that the opposite is true.

2. Nothing so clearly demonstrates the majori-ty’s confusion on this issue as the cases onwhich it relies. In support of its view that acourt must scour the specification for corre-sponding structure in order to save a claimterm from means-plus-function treatment, themajority cites several cases that were attempt-ing to identify the corresponding structure ofundisputedly means-plus-function claimterms. For example, the majority cites Typ-hoon Touch Technologies, Inc. v. Dell, Inc.,

659 F.3d 1376, 1385 (Fed.Cir.2011), for theproposition that a claim may contain suffi-cient structure to avoid means-plus-functionclaiming by reciting an algorithm. MajorityOp. 1298–99. However, the claim term atissue in that case was ‘‘means for cross-refer-encing,’’ which no one disputed was a means-plus-function limitation. This court’s searchfor an algorithm to support that claim termwas merely an attempt to save the means-plus-function term from indefiniteness. SeeTyphoon Touch, 659 F.3d at 1385–86. Thesame is true of Finisar Corp. v. DirecTVGroup, Inc., 523 F.3d 1323 (Fed.Cir.2008), onwhich the majority also relies. Id. at 1340(evaluating whether ‘‘computer-implementedmeans-plus-function claims’’ disclosedenough of an algorithm to provide the neces-sary structure under § 112, ¶ 6). The majori-ty’s misreading of these cases is all the morestriking because it is the same error the ma-


Applying the proper legal analysis, therecan be little doubt that the heuristic limita-tions are means-plus-function limitations.

To begin with, it is true that the absenceof the word ‘‘means’’ in the disputed claimterms creates a presumption that theseare not means-plus-function limitations.See Inventio AG v. ThyssenKrupp Eleva-tor Ams. Corp., 649 F.3d 1350, 1356 (Fed.Cir.2011). However, it is undisputed thatthe heuristics limitations recite functionsperformed by the heuristics (e.g., ‘‘deter-mining that the one or more finger con-tacts correspond to a command to transi-tion from displaying a respective item in aset of items to displaying a next item inthe set of items’’). The relevant questiontherefore is whether the claim fails to re-cite sufficient structure for performingthose functions, in which case the pre-sumption against means-plus-functiontreatment would be overcome. See id.

As we have previously explained:

What is important is whether the termis one that is understood to describestructure, as opposed to a term that issimply a nonce word or a verbal con-struct that is not recognized as the nameof structure and is simply a substitutefor the term ‘‘means for.’’ The court inPersonalized Media Communicationsdrew the pertinent distinction in holdingthat the term ‘‘detector,’’ althoughbroad, is still structural for purposes ofsection 112 ¶ 6 because it ‘‘is not a ge-neric structural term such as ‘means,’‘element,’ or ‘device’; nor is it a coinedterm lacking a clear meaning such as‘widget’ or ‘ram-a-fram.’ ‘‘

Lighting World, Inc. v. Birchwood Light-ing, Inc., 382 F.3d 1354, 1360 (Fed.Cir.

2004) (quoting Personalized MediaCommc’ns, LLC v. Int’l Trade Comm’n,161 F.3d 696, 704 (Fed.Cir.1998)). Follow-ing this, we have found that terms such as‘‘computing unit,’’ Inventio, 649 F.3d at1359–60, and ‘‘soft start circuit,’’ PowerIntegrations, Inc. v. Fairchild Semicon-ductor International, Inc., 711 F.3d 1348,1365 (Fed.Cir.2013), connote sufficientstructure to avoid means-plus-functiontreatment. On the other hand, terms suchas ‘‘colorant selection mechanism,’’ Massa-chusetts Institute of Technology v. AbacusSoftware, 462 F.3d 1344, 1353–54 (Fed.Cir.2006), and ‘‘mechanism for moving saidfinger,’’ Welker Bearing, 550 F.3d at 1095–97, have failed to recite sufficient structureto avoid means-plus-function treatment.

The term heuristic, which the districtcourt construed as ‘‘rules to be applied todata to assist in drawing inferences fromthat data,’’ is a vague concept that does notconnote known, physical structure in thesame way as the terms ‘‘computing unit’’or ‘‘circuit.’’ Rather, the term heuristic ismuch more similar to the imprecise terms‘‘mechanism,’’ ‘‘means,’’ and ‘‘element.’’Therefore, I would affirm the districtcourt’s conclusion that the heuristic limita-tions are subject to means-plus-functiontreatment.

If, as the majority concludes—in myview incorrectly—our precedent dictatesthat the ‘‘heuristic’’ limitations in the 8949patent are not subject to means-plus-func-tion treatment, then perhaps it is time torevisit the issue of when claim languageinvokes § 112 ¶ 6, particularly for comput-er-implemented inventions. As explainedabove, I see no real difference between thefollowing two claim limitations:

jority accuses the district court of committingin its reading of Aristocrat Technologies Aus-tralia Pty Ltd. v. International Game Technolo-

gy, 521 F.3d 1328, 1333 (Fed.Cir.2008). Ma-jority Op. 1298.


a next item means for determining thatthe one or more finger contacts corre-spond to a command to transition fromdisplaying a respective item in a set ofitems to displaying a next item in the setof items; and

a next item heuristic for determiningthat the one or more finger contactscorrespond to a command to transitionfrom displaying a respective item in aset of items to displaying a next item inthe set of items.

The first limitation would surely be con-strued as a means-plus-function limitation.Yet, although the second limitation pro-vides no more real structure than the first,the majority concludes that it is outsidethe realm of § 112 ¶ 6.

That one minor drafting decision greatlyexpands the scope of the claim limitationbecause the claim is not limited to thecorresponding structure disclosed in thepatent specification. Indeed, under themajority’s view, this case provides a starkexample of how patent applicants are ableto claim broad functionality without beingsubject to the restraints imposed by § 112¶ 6. As interpreted by the majority, the‘‘next item heuristic’’ limitation in claim 1covers any heuristic used by a touch-screen device to determine that ‘‘one ormore finger contacts correspond to a com-mand to transition from displaying a re-spective item in a set of items to displayinga next item in the set of items.’’ And yet, itis undisputed that the patent specificationdiscloses only two such heuristics: a swipefrom right to left and a tap on the rightside of the screen. In fact, the majorityrelied on exactly those two heuristics as

the necessary ‘‘structure’’ that removedthis claim term from the means-plus-func-tion realm altogether. Majority Op. 1301–02.3

Professor Lemley has recently writtenon the issue of functional claiming in soft-ware patents. Mark A. Lemley, SoftwarePatents and the Return of FunctionalClaiming, 2013 WIS. L.REV. 905 (2013). Heacknowledges this court’s recent case lawin which we have been ‘‘vigilant in limitingsoftware patentees who write claims inmeans-plus-function format to the particu-lar algorithms that implement thoseclaims.’’ Id. at 926. However, he notesthat in many cases involving software pat-ents, we have not treated claims as means-plus-function claims at all, leading to prob-lems of overbroad patents:

The presence of structure in the form of‘‘a computer’’ or ‘‘a processor’’ or even‘‘the Internet’’ has led the Federal Cir-cuit to give these claims control over theclaimed function however implemented.As a result, software patents have cir-cumvented the limits the 1952 Act placeson functional claiming. The result hasbeen a plethora of software patentsclaimed not on the basis of the technolo-gy the patentee actually developed, buton the basis of the function that technol-ogy performs. Those claims aren’t lim-ited to or commensurate with what thepatentee invented, and they are accord-ingly the ones that patent plaintiffs tendto assert against defendants whose sys-tems bear little resemblance to what thepatentee actually inventedTTTT [U]nderthis functional claiming rubric the soft-ware patents with the least actual tech-nical content end up with the broadest

3. Oddly, in defining ‘‘heuristic,’’ the majoritystates that it ‘‘does not include all means forperforming the recited function.’’ MajorityOp. 1301. But of course, once this court has

ruled that the term is not a means-plus-func-tion limitation, that term can indeed be con-strued so broadly.


claims: ‘‘Its monopoly breadth is a func-tion of its lack of technical specification.’’

Id. at 926–28 (footnotes omitted) (quotingChristina Bohannan & Herbert Hoven-kamp, Creation Without Restraint: Pro-moting Liberty and Rivalry in Innovation125 (2012)); see also id. at 905 (‘‘[P]aten-tees claim to own not a particular machine,or even a particular series of steps forachieving a goal, but the goal itself. Theresulting overbroad patents overlap andcreate patent thickets.’’). I believe Profes-sor Lemley raises valid concerns aboutthis court’s means-plus-function jurispru-dence as it relates to computer-implement-ed inventions. Although the majority heredid not base its conclusion on such genericstructural terms as ‘‘a computer’’ or ‘‘aprocessor,’’ the end result is the same:Apple’s 8949 patent is construed broadly tocover the function performed by the heu-ristics, not the specific heuristics disclosedin the specification. This outcome shouldcompel our court to reconsider when wetreat functional claims as means-plus-func-tion claims.

B. Corresponding Structure

Because I would affirm the districtcourt’s conclusion that the heuristics limi-tations are means-plus-function limitations,I would therefore reach Apple’s alternativeargument regarding the correspondingstructure for the term ‘‘next item heuris-tic.’’

The 8949 patent specification containstwo examples of a next item heuristic: atap on the right side of the touchscreenand a right-to-left horizontal swipe. See,e.g., 8949 patent col. 34 ll. 12–16 (‘‘In someembodiments, the user can also initiateviewing of the next image by making a tapgesture 1620 on the right side of the image.In some embodiments, the user can also

initiate viewing of the next image by mak-ing a swipe gesture 1616 from right to lefton the image.’’ (emphases added)). Never-theless, the district court rejected the hori-zontal swipe as corresponding structurefor the claimed next item heuristic. Thedistrict court reasoned that a horizontalfinger swipe fell within the scope of anoth-er claimed heuristic namely the ‘‘two-di-mensional screen translation heuristic’’ inthe prior claim limitation. The districtcourt did not see how ‘‘the same userfinger movement [could be] understood tocommunicate two separate commands.’’J.A. 93. Therefore, the court limited thenext item heuristic’s corresponding struc-ture to ‘‘a heuristic that uses as one inputa user’s finger tap on the right side of thedevice’s touch screen.’’ J.A. 94.

Apple argues—and I agree—that thedistrict court failed to consider that theheuristics might apply in different scenar-ios. For example, in describing the photoalbum application depicted in Figure 16Aof the 8949 patent, the specification ex-plains that the user can ‘‘initiate viewing ofthe next image by making a swipe gesture1616 from right to left on the image.’’8949 patent col. 34 ll. 14–16. The specifica-tion goes on to explain, however, that ‘‘ifjust a portion of image 1606 is displayed,in response to detecting a finger drag orswipe (e.g., 1626), the displayed portion ofthe image is translated in accordance withthe direction of the drag or swipe gesture(e.g., vertical, horizontal, or diagonal trans-lation).’’ Id. at col. 35 ll. 19–24. In otherwords, a horizontal swipe will be treateddifferently depending on whether a fullimage is displayed or the user has zoomedin on a portion of the image. There isnothing in the claim language that requiresthat the ‘‘two-dimensional screen transla-tion heuristic’’ and ‘‘next item heuristic’’apply in the same context. Therefore, I


would reverse the district court’s decisionto limit the corresponding structure to afinger tap on the right side of the screen,and would reverse the grant of summaryjudgment of non-infringement for the ac-cused products that use a ‘‘swipe’’ gestureas a next item heuristic.

II. ADMISSIBILITY OF EXPERT TESTIMONY

RELATING TO THE 8949 PATENT

While I concur in the judgment revers-ing and remanding the district court’s ex-clusion of the testimony of Apple’s expert,Brian Napper, based on his erroneousclaim construction, I write separately tonote my agreement with the district courtthat Napper’s reliance on the MagicTrackpad was inherently unreliable.

The majority concludes that the districtcourt erred in excluding the testimony ofApple’s expert, Brian Napper, for two rea-sons. First, the district court’s analysiswas based on an incorrect claim construc-tion. Majority Op. 1315–16. Specifically,the district court criticized Napper for fail-ing to isolate the value to consumers of the‘‘tap for next item’’ functionality, and forfailing to consider alternatives to a $35million royalty that would enable Motorolato provide this functionality. See J.A. 112–14. These criticisms were based on thedistrict court’s conclusion that the 8949patent’s claimed ‘‘next item heuristic’’ wasa means-plus-function claim that was limit-ed to a tap on the right-hand side of thescreen to turn to the next item. See J.A.112. The majority determined that thisclaim construction was erroneous becausethe ‘‘next item heuristic’’ was not a means-plus-function limitation. Majority Op.1304. As explained above, I would affirmthe district court’s construction of the‘‘heuristic’’ claim terms as means-plus-function limitations, but would reverse the

court’s ruling that the scope of the corre-sponding structure for the ‘‘next item heu-ristic’’ is limited to a tap on the right-handside of the screen. Thus, although I arriveat the outcome in a different way, I concurin the majority’s conclusion that the dis-trict court’s exclusion of Napper’s testimo-ny must be reversed and remanded on thebasis of its erroneous claim construction.

However, unlike the majority, I do notbelieve the district court abused its discre-tion in excluding Napper’s testimony asunreliable. The district court determinedthat Napper’s reliance on the MagicTrackpad to calculate the value of theclaimed functionality was unreliable be-cause that product was not a sufficientlycomparable benchmark. Specifically, thecourt said:

Napper’s proposed testimony does notprovide a reliable basis for inferring thevalue even of the vertical scrolling fea-ture. The fact that many consumerswill pay more for a Magic Trackpadthan for a mouse tells one nothing aboutwhat they will pay to avoid occasionallyswiping unsuccessfully because theirswiping finger wasn’t actually vertical tothe screen. Maybe consumers wouldpay $2, but there is no evidence theywould, or at least none furnished byNapper.

J.A. 115.

It is true that our precedent supportslooking to the cost of benchmark commer-cial products in determining the value of adefendant’s infringement. See, e.g., i4iLtd. P’ship v. Microsoft Corp., 598 F.3d831, 853–56 (Fed.Cir.2010). However, aseven Apple concedes, the Trackpad ‘‘con-tains none of the function asserted fromthe 8949 patent.’’ Motorola Response Br.39; see also Apple Reply Br. 70 (acknowl-


edging that ‘‘[t]hat is true’’). Nappertherefore began his analysis from a highlyquestionable starting point. And becausethe Trackpad does not contain any of theclaimed functionality, the discounts Nap-per applied to get from $20 to $2 (suppos-edly because the Trackpad contains morefeatures than those claimed by the 8949patent) appear to be completely arbitrary.Accordingly, I do not believe the districtcourt abused its discretion in concludingthat Napper’s proposed testimony failed toprovide a reliable basis for inferring thevalue of the claimed functionality. On re-mand, I do not think it would be reversibleerror for the district court to again prohib-it Apple from relying on such unreliabletestimony.

III. APPLE’S REQUEST FOR

INJUNCTIVE RELIEF

The district court granted summaryjudgment that Apple was not entitled to aninjunction for infringement of the 8949,8263, and 8647 patents. The majority va-cates the grant of summary judgment—apparently for all three patents—becauseof its reversal of the claim construction forthe 8949 patent. Majority Op. 1330–31. Irespectfully dissent, and would affirm thegrant of summary judgment for all threepatents.

Apple argues that it presented ampleevidence with respect to the eBay factorsoutlined above to survive summary judg-ment. I will only address the first injunc-tion factor—irreparable harm—because Iview it as dispositive.

Apple argues that it is being irreparablyharmed by Motorola’s infringement be-cause sales of Motorola’s infringing prod-ucts are causing Apple to lose marketshare and downstream sales.4 But in or-der to rely on lost market share and down-stream sales to show irreparable harm,Apple must provide more than evidenceshowing merely that Motorola is takingmarket share from Apple. Rather, Applemust be able to show a causal nexus be-tween the inclusion of the allegedly in-fringing features in Motorola’s phones andthe alleged harm to Apple. See Apple III,735 F.3d at 1360–61; Apple Inc. v. Sam-sung Elecs., 678 F.3d 1314, 1324 (Fed.Cir.2012) (Apple I ) (‘‘Sales lost to an infring-ing product cannot irreparably harm a pat-entee if consumers buy that product forreasons other than the patented feature.If the patented feature does not drive thedemand for the product, sales would belost even if the offending feature wereabsent from the accused product.’’).

The district court found that Apple hadno evidence linking Apple’s lost marketshare and downstream sales to the inclu-sion of the allegedly infringing features inMotorola’s phones. See J.A. 152 (‘‘Apple’s‘feel good’ theory does not indicate thatinfringement of these claims (if they wereinfringed) reduced Apple’s sales or marketshare, or impaired consumer goodwill to-ward Apple products.’’). Rather, the dis-trict court found that ‘‘Apple is complain-ing that Motorola’s phones as a wholeripped off the iPhone as a whole,’’ whichthe court explained was insufficient be-

4. Apple also argues that a finding of irrepara-ble harm is supported by evidence showingthat Apple has a policy against licensing com-petitors to practice the three asserted patents.Apple’s willingness to license the asserted pat-ents is, of course, relevant to the second in-junction factor—the inadequacy of legal rem-

edies to compensate for irreparable harm.See, e.g., Apple Inc. v. Samsung Elecs. Co., 735F.3d 1352, 1369–71 (Fed.Cir.2013) (AppleIII ). However, I am not aware of any casesestablishing that a policy against licensingpatents can show irreparable harm in the firstplace.


cause ‘‘Motorola’s desire to sell productsthat compete with the iPhone is a separateharm—and a perfectly legal one—fromany harm caused by patent infringement.’’Id. at 153.

Apple contends that its evidence raises agenuine issue as to whether the allegedlyinfringing features are drivers of consumerdemand for Motorola’s products. I dis-agree. First, with respect to the 8949patent, Apple cites consumer survey evi-dence purporting to show that ‘‘having asuperior touchscreen interface—as op-posed to a physical keyboard—drives con-sumers demand for smartphones.’’ AppleBr. 65. This evidence, however, says noth-ing about the specific features claimed inthe 8949 patent. See, e.g., Apple Inc. v.Samsung Elecs. Co., 695 F.3d 1370, 1376(Fed.Cir.2012) (Apple II ). The only evi-dence Apple cites that allegedly relatesspecifically to ‘‘the scrolling heuristicsclaimed by the 8949 patent’’ is testimonyfrom Motorola executives about what theythought would be important to consumers.Apple Br. 66. However, we found similarevidence, standing alone, to be insufficientto establish a causal nexus in Apple I,explaining:

While the evidence that Samsung’s em-ployees believed it to be important toincorporate the patented feature intoSamsung’s products is certainly relevantto the issue of nexus between the patentand market harm, it is not dispositive.That is because the relevant inquiry fo-cuses on the objective reasons as to whythe patentee lost sales, not on the in-fringer’s subjective beliefs as to why itgained them (or would be likely to gainthem).

678 F.3d at 1327–28. Accordingly, I agreewith the district court that Apple’s evi-dence fails to raise a genuine issue as to

whether the features accused of infringingthe 8949 patent drive consumer demand forMotorola’s phones.

As for the 8263 patent, Apple cites stud-ies and surveys purporting to show thatthe claimed invention facilitated the devel-opment of popular apps for the iPhone andiPad, which in turn helped make Apple’sdevices so popular. But Apple does notcite any similar studies for consumers ofMotorola’s products. Even assuming theevidence shows that the 8263 patent is adriver of demand for Apple’s products,that does not mean the 8263 patent is adriver of demand for Motorola’s products.See Apple II, 695 F.3d at 1376. Apple’sonly other evidence for the 8263 patentallegedly shows Motorola’s subjective be-liefs that using ‘‘Apple’s simplification ofstreaming technology’’ would help it gainsales. Apple Br. 67. Again, however, thisevidence, standing alone, is insufficient toestablish a causal nexus between Motoro-la’s incorporation of the allegedly infring-ing feature and the alleged harm sufferedby Apple. See Apple I, 678 F.3d at 1327–28.

Similarly, with respect to the 8647 pat-ent, Apple cites evidence that ‘‘Motorolaitself’’ identified the patent’s ‘‘structure-detection feature’’ as ‘‘a high priority and a‘differentiating’ feature.’’ Apple Br. 68(quoting J.A. 29,857). Apple also cites anewspaper article describing the feature asa ‘‘cool[ ] new feature[ ]’’ in Apple’s prod-ucts. J.A. 29,893. As with the 8263 pat-ent, this evidence is insufficient to estab-lish the requisite nexus.

For these reasons, I agree with the dis-trict court that Apple’s evidence fails toraise a genuine issue as to whether theallegedly infringing features are drivers ofconsumer demand for Motorola’s products.As a result, Apple cannot show that Moto-


rola’s infringement has caused it irrepara-ble harm. Apple therefore cannot meetthe eBay standard for injunctive relief.Accordingly, I would affirm the districtcourt’s grant of summary judgment of noinjunctive relief.

IV. MOTOROLA’S REQUEST FOR INJUNCTIVE

RELIEF RELATING TO THE 8898 PATENT

I concur in the majority’s judgment thatMotorola is not entitled to an injunctionfor infringement of the 8898 patent. Ma-jority Op. 1331–32. However, I write sep-arately to note my disagreement with themajority’s suggestion that an alleged in-fringer’s refusal to negotiate a license jus-tifies the issuance of an injunction after afinding of infringement.

As an initial matter, I agree with themajority that there is no need to create acategorical rule that a patentee can neverobtain an injunction on a FRAND-commit-ted patent.5 Id. at 71–72. Rather,FRAND commitment should simply befactored into the consideration of the eBayframework. Moreover, I agree that astraightforward application of the eBayfactors does not necessarily mean that in-junctive relief would never be available fora FRAND-committed patent. However, Idisagree as to the circumstances underwhich an injunction might be appropriate.

Motorola argues—and the majorityagrees—that an injunction might be appro-priate where an alleged infringer ‘‘unilater-ally refuses a FRAND royalty or unrea-sonably delays negotiations to the sameeffect.’’ Id. Motorola insists that in theabsence of the threat of an injunction, an

infringer would have no incentive to nego-tiate a license because the worst-case sce-nario from a patent infringement lawsuit isthat it would have to pay the same amountit would have paid earlier for a license.

I disagree that an alleged infringer’srefusal to enter into a licensing agreementjustifies entering an injunction against itsconduct, for several reasons. First, asApple points out, an alleged infringer isfully entitled to challenge the validity of aFRAND-committed patent before agree-ing to pay a license on that patent, and soshould not necessarily be punished forless than eager negotiations. Second,there are many reasons an alleged in-fringer might prefer to pay a FRANDlicense rather than undergoing extensivelitigation, including litigation expenses, thepossibility of paying treble damages or at-torney’s fees if they are found liable forwillful infringement, and the risk that thefact-finder may award damages in anamount higher than the FRAND rates.Indeed, as Motorola itself pointed out, wehave previously acknowledged that a trialcourt may award an amount of damagesgreater than a reasonable royalty if neces-sary ‘‘to compensate for the infringe-ment.’’ Stickle v. Heublein, Inc., 716 F.2d1550, 1563 (Fed.Cir.1983). Thus, if a trialcourt believes that an infringer previouslyengaged in bad faith negotiations, it is en-titled to increase the damages to accountfor any harm to the patentee as a resultof that behavior.

But regardless, none of these consider-ations alters the fact that monetary dam-ages are likely adequate to compensate for

5. For what it’s worth, I would note that thedistrict court did not apply a per se rule thatinjunctions are unavailable for SEPs. Rather,Judge Posner expressly noted that injunctiverelief might have been appropriate if Apple

had ‘‘refuse[d] to pay a royalty that meets theFRAND requirement.’’ J.A. 140. Thus, themajority need not have suggested that thedistrict court erred insofar as it applied sucha categorical rule. See Majority Op. 1331–32.

1343IN RE DINSMORECite as 757 F.3d 1343 (Fed. Cir. 2014)

a FRAND patentee’s injuries. I see noreason, therefore, why a party’s pre-litiga-tion conduct in license negotiations shouldaffect the availability of injunctive relief.

Instead, an injunction might be appro-priate where, although monetary damagescould compensate for the patentee’s inju-ries, the patentee is unable to collect thedamages to which it is entitled. For ex-ample, if an alleged infringer were judg-ment-proof, a damages award would likelybe an inadequate remedy. Or, if a defen-dant refused to pay a court-ordered dam-ages award after being found to infringe avalid FRAND patent, a court might bejustified in including an injunction as partof an award of sanctions.

But regardless, these circumstances arenot present in this case, and I agree withthe district court that under the facts here,Motorola cannot show either irreparableharm or inadequacy of damages. I wouldtherefore affirm the district court’s denialof Motorola’s claim for injunctive relief forthe 8898 patent.

CONCLUSION

For the reasons above, I concur in partand dissent in part.

,

In re Mark T. DINSMOREand David J. Caruso.

No. 2013–1637.

United States Court of Appeals,Federal Circuit.

June 10, 2014.

Rehearing and Rehearing En BancDenied Aug. 27, 2014.

Background: Applicants sought to modifyrecorded terminal disclaimer to delete

common-ownership provision, and to modi-fy claims that gave rise to disclaimer, onbasis that the two patents were not andnever had been commonly owned. Patentexaminer rejected owners’ reissue applica-tion. Owners appealed. The Patent Trialand Appeal Board, 2013 WL 5274029, af-firmed. Owners appealed.

Holding: The Court of Appeals, Taranto,Circuit Judge, held that reissue applicationdid not meet ‘‘error’’ requirement.

Affirmed.

1. Patents O136, 155

Applicants did not have sufficient ba-sis for reissue on basis that terminal dis-claimer was ‘‘ineffective’’ because it did notproduce a patent they could actually en-force by themselves without common own-ership and that they erred in producingthat result by choosing to file terminaldisclaimer in order to include claims inpatent, in their final form, that led toobviousness-type double-patenting rejec-tion; even if condition of present inabilityto enforce constituted inoperativeness, itultimately was no more than statement ofsubsequently-regretted choice because ap-plicants did not identify any cognizablefalse or deficient understanding of fact orlaw that underlay that choice. 35U.S.C.(2006 Ed.) § 251.

2. Patents O140

Determining whether applicant forreissue patent has met statutory require-ments is question of law, which Court ofAppeals reviews de novo. 35 U.S.C.(2006Ed.) § 251.

3. Patents O136

The patent reissue provision requiresnot only that the original patent be ‘‘wholly

1286 757 federal reporter, 3d series - … › wp-content › uploads › 2016 › 05 › ...1286...

Documents