From Casetext: Smarter Legal Research

Apple, Inc. v. Samsung Electronics Co.

UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA SAN JOSE DIVISION
Feb 15, 2018
Case No. 12-CV-00630-LHK (N.D. Cal. Feb. 15, 2018)

Summary

applying test to defendant's products that were allegedly "designed around" the infringed patent

Summary of this case from Opticurrent, LLC v. Power Integrations, Inc.

Opinion

Case No. 12-CV-00630-LHK

02-15-2018

APPLE INC., Plaintiff, v. SAMSUNG ELECTRONICS CO., LTD., et al., Defendants.


ORDER GRANTING IN PART AND DENYING IN PART MOTION FOR ONGOING ROYALTIES

Re: Dkt. No. 2217

Before the Court is Apple's motion for ongoing royalties on Samsung's post-judgment sales of products with features that the jury found infringed U.S. Patent No. 5,946,647. ECF No. 2239-5. Samsung stipulates that it owes $6,494,252 plus interest in ongoing royalties for these sales. Apple also argues that it is entitled to ongoing royalties on products containing design-arounds that Apple alleges are not colorably different from the products that the jury found to infringe the '647 patent. Having considered the law, the record, and the parties' arguments from their briefs and from the January 11, 2018 hearing, the Court GRANTS Apple's motion only as to the stipulated amount of $6,494,252 plus interest. The Court DENIES Apple's motion for ongoing royalties for the products containing design-arounds.

Samsung unsealed the total amount of its stipulated ongoing royalties. See ECF No. 2239-5 at 24; ECF No. 2239-6 at 2.

I. BACKGROUND

On May 5, 2014, a jury reached a verdict that found that Samsung infringed Apple's U.S. Patent Nos. 5,946,647 (the "'647 patent"), 8,046,721 (the "'721 patent"), and 8,074,172 (the "'172 patent"). ECF No. 1884. Only the '647 patent is at issue in the instant motion. The Court first provides an overview of the relevant claim of the '647 patent, including the construction of several claim terms. The Court next summarizes the jury's verdict and the relevant post-trial proceedings.

A. The '647 Patent

The '647 patent, entitled "System and Method for Performing an Action on a Structure in Computer-Generated Data," was filed on February 1, 1996 and issued on August 31, 1999. The '647 patent discloses a "system and a method [that] causes a computer to detect and perform actions on structures identified in computer data." '647 patent. In other words, "[t]he '647 patent discloses a system for recognizing certain structures (such as a telephone number) on a touchscreen and then linking certain actions (such as calling the telephone number) to the structure." Apple Inc. v. Motorola, Inc., 757 F.3d 1286, 1304 (Fed. Cir. 2014), overruled in part on other grounds by Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc). For example, the system may scan a Microsoft Word document and recognize when phone numbers or email addresses appear in the document. See '647 patent at col. 1:24-35; see also id. at col. 2:42-53. Then, the system may link actions to these structures and allow the user to select an action. Id. As an example, when an email address is detected in a document, the system may automatically give the user the options to send an email to the identified address or to store the email address in an electronic address book. Id. at col. 5:5-18. As another example, when a phone number is detected in a document, the system may give the user the option to place a call to that phone number or to place the number in an electronic contact list. Id.

Claim 1 of the '647 patent recites:

A computer-based system for detecting structures in data and performing actions on detected structures, comprising:
an input device for receiving data;

an output device for presenting the data;

a memory storing information including program routines including

an analyzer server for detecting structures in the data, and for linking actions to the detected structures;

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

an action processor for performing the selected action linked to the selected structure; and

a processing unit coupled to the input device, the output device, and the memory for controlling the execution of the program routines.
'647 patent at col. 7:9-24. Asserted claim 9, which depends from claim 1, claims "[t]he system recited in claim 1, wherein the user interface enables selection of an action by causing the output devices to display a pop-up menu of the linked actions." Id. at col. 7:53-55.

1. Claim Construction

The Court construed "action processor" to mean "program routine(s) that perform the selected action on the detected structure." ECF No. 447 at 20. In its summary judgment order, the Court noted that "[t]he parties agree that the plain and ordinary meaning of the claim limitation 'a user interface enabling the selection of a detected structure' requires 'the user interface to enable selection of a structure, by the user, after the structure has already been detected.'" ECF No. 1150 at 20 (quoting Reply Declaration of Dr. Todd C. Mowry Concerning U.S. Patent No. 5,946,647 at ¶ 217 (ECF No. 805-14) (emphasis in original)).

In addition, at the preliminary injunction and summary judgment stages of the litigation, the Court adopted Judge Richard A. Posner's construction of "analyzer server" from Apple, Inc. v. Motorola, Inc., No. 1:11-cv-8540 (N.D. Ill. March 19, 2012), which construed "analyzer server" to mean "a server routine separate from a client that receives data having structures from the client." ECF No. 1150 at 17. On the last day of presenting evidence at trial in the instant case, the Federal Circuit affirmed Judge Posner's construction of "analyzer server" in the Motorola case. Motorola, 757 F.3d at 1304; see Apple Inc. v. Samsung Elecs. Co., Ltd., 839 F.3d 1034, 1041 (Fed. Cir. 2016) (en banc) ("Federal Circuit En Banc Opinion") (describing sequence of events). The Federal Circuit also construed "linking actions to the detected structures" to mean "creating a specified connection between each detected structure and at least one computer subroutine that causes the CPU to perform a sequence of operations on that detected structure." Motorola, 757 F.3d at 1305-06. With the agreement of the parties in the instant case, this Court instructed the jury in the instant case on the Motorola constructions of "analyzer server" and "linking actions to the detected structures." Federal Circuit En Banc Opinion, 839 F.3d at 1041; ECF No. 1848 at 30. Neither party appealed these claim constructions. See Federal Circuit En Banc Opinion, 839 F.3d at 1043 ("Claim construction was not appealed . . . .").

B. The Jury Verdict and Relevant Post-Trial Proceedings

On May 5, 2014, the jury found that claim 9 was not invalid and was infringed by all nine adjudicated products: the Admire, Galaxy Nexus, Galaxy Note, Galaxy Note II, Galaxy SII, Galaxy SII Epic 4G Touch, Galaxy SII Skyrocket, Galaxy SIII, and Galaxy Stratosphere. ECF No. 1884 at 2, 7. On May 23, 2014, both parties filed motions for judgment as a matter of law. See ECF Nos. 1896-3, 1897-3. The Court subsequently resolved the parties' motions for judgment as a matter of law, including by denying Samsung's motion for judgment as a matter of law on claim 9 of the '647 patent. ECF Nos. 1963, 1965.

On September 3, 2014, Apple filed a motion seeking ongoing royalties for any future infringement by Samsung. ECF No. 1958. Apple demanded royalties from Samsung for any continuing sales of the products for which the jury found infringement and for any Samsung products "not more than colorably different" from the adjudicated products. ECF No. 1959 (Apple's Proposed Order). The Court ordered the parties to brief whether Apple was entitled to ongoing royalties at all, as well as the proper amount of any such royalties. ECF No. 1978. After receiving briefing from the parties on both issues, see ECF Nos. 1985-3, 1986-3, 2001, 2015-2, 2046-3, 2050, the Court on November 25, 2014 granted Apple's motion for ongoing royalties for "products adjudicated to infringe the '647, '172, or '721 Patents, and to products 'not more than colorably different therefrom,'" ECF No. 2075 at 36.

Specifically, the Court ruled that Apple was entitled to ongoing royalties beginning after entry of final judgment. ECF No. 2075 at 22. After analyzing the factors from Georgia-Pacific Corp. v. U.S. Plywood Corp., 318 F. Supp. 1116, 1120 (S.D.N.Y. 1970), the Court determined that the following ongoing royalty rates for the '647 patent were appropriate based on the jury's verdict:

Product

'647 Patent Ongoing Royalty Rate

Admire

[Redacted]

Galaxy Nexus

[Redacted]

Galaxy Note

[Redacted]

Galaxy Note II

[Redacted]

Galaxy S II (AT&T / T-Mobile)

[Redacted]

Galaxy S II Epic 4G Touch

[Redacted]

Galaxy S II Skyrocket

[Redacted]

Galaxy S III

[Redacted]

Stratosphere

[Redacted]

Id. at 36. The Court also found that "[f]or any unadjudicated products 'not more than colorably different,' the rate[] shall be [Redacted] for the '647 patent." Id. However, the Court did not determine which unadjudicated products, if any, were subject to ongoing royalties. See id. at 18-21. The Court entered final judgment on November 25, 2014. ECF No. 2076.

Samsung filed a notice of appeal of the judgment to the Federal Circuit that same day. ECF No. 2077. On December 5, 2014, Apple filed a notice of cross-appeal. ECF No. 2079. On February 26, 2016, a three judge panel of the Federal Circuit reversed this Court's order denying judgment as a matter of law to Samsung on claim 9 of the '647 patent. Apple Inc. v. Samsung Elecs. Co., 816 F.3d 788, 797 (Fed. Cir. 2016). On October 7, 2016, the Federal Circuit, sitting en banc, reversed the Federal Circuit three judge panel's decision and affirmed this Court's denial of Samsung's motion for judgment as a matter of law as to claim 9 of the '647 patent. Federal Circuit En Banc Opinion, 839 F.3d at 1040-47. On November 6, 2017, the United States Supreme Court denied Samsung's petition for certiorari. 138 S. Ct. 420 (Nov. 6, 2017); ECF No. 2216.

On November 13, 2017, Apple filed the instant motion regarding ongoing royalties. ECF Nos. 2217-2, 2239-5 ("Mot."). On December 6, 2017, Samsung filed an opposition. ECF Nos. 2220-3, 2239-16 ("Opp'n"). On December 13, 2017, Apple filed a reply. ECF Nos. 2222-2, 2239-17 ("Reply").

Samsung concedes that it owes $6,494,252 (plus interest) for post-judgment sales of [Redacted] products between November 26, 2014, the day after judgment was entered, through January 31, 2016, the day before the '647 patent expired, that contained the identical features found to infringe at trial. See Mot. at 4; Opp'n at 9 n.6; ECF No. 2197-2 at 1; ECF No. 2217-4 ("Robinson Decl.") at ¶ 6. The products covered by this ongoing royalty amount include: Galaxy Centura, Galaxy Mega 6.3", Galaxy Rugby Pro, Galaxy S3 Mini, Galaxy S4 Mini, Samsung ATIV SE, Galaxy Note Edge, Galaxy Note3, Galaxy Note4, Galaxy S4, Galaxy S5, Galaxy Trend Style, Galaxy Core Prime, Galaxy S5 Active, Galaxy Core LTE, Galaxy Mega 2, Galaxy Exhibit, Galaxy Ace Style, and Galaxy Alpha. Robinson Decl. Exh. 3B. In Samsung's brief opposing Apple's motion for ongoing royalties, Samsung argued that setting an ongoing royalty rate would be a waste of judicial time and resources because such an ongoing royalty rate would never be applied given "the absence of continued sales of the products with software held to infringe." ECF No. 1986-3 at 15. Samsung's concession that it owes nearly $6.5 million in ongoing royalties for products containing the code found to infringe is not consistent with its previous representation that sales of infringing products had stopped.

The parties disagree about whether Samsung owes ongoing royalties on products that include one or more of three design-arounds.

II. LEGAL STANDARD

An ongoing royalty permits an adjudged infringer to continue using a patented invention for a price. See Paice LLC v. Toyota Motor Corp., 504 F.3d 1293, 1313 n.13 (Fed. Cir. 2007) (defining an ongoing royalty and distinguishing a compulsory license). The Federal Circuit has identified 35 U.S.C. § 283, which authorizes "injunctions in accordance with the principles of equity," as statutory authority for awarding ongoing royalties. See id. at 1314 (citing § 283 and stating that "[u]nder some circumstances, awarding an ongoing royalty for patent infringement in lieu of an injunction may be appropriate"); see also Mark A. Lemley, The Ongoing Confusion Over Ongoing Royalties, 76 Mo. L. Rev. 695, 695- 99 (2001) (analyzing authority for ongoing royalties under §§ 283 and 284). Accordingly, while this remedy involves monetary relief, there is no Seventh Amendment right to jury trial for ongoing royalties. See Paice, 504 F.3d at 1315-16 ("[T]he fact that monetary relief is at issue in this case does not, standing alone, warrant a jury trial.").

The Federal Circuit has not squarely addressed the standard for evaluating claims for ongoing royalties on newly accused products. However, the Federal Circuit has articulated a test in the closely related context of "evaluating whether an injunction against continued infringement has been violated by a newly accused product." Proveris Scientific Corp. v. Innovasystems, Inc., 739 F.3d 1367, 1370 (Fed. Cir. 2014). In the injunction context, "courts must follow the two-step test outlined in TiVo Inc. v. EchoStar Corp., 646 F.3d 869 (Fed. Cir. 2011) (en banc)." Proveris, 739 F.3d at 1370.

Under the first step of TiVo, "a party seeking to enforce an injunction must show that 'the newly accused product is not more than colorably different from the product found to infringe.'" Id. (quoting TiVo, 646 F.3d at 882). "The analysis must focus not on differences between randomly chosen features of the product found to infringe in the earlier infringement trial and the newly accused product, but on those aspects of the accused product that were previously alleged to be, and were a basis for, the prior finding of infringement, and the modified features of the newly accused product." TiVo, 646 F.3d at 882 (internal citation omitted). Specifically, the Federal Circuit instructed courts to "focus on those elements of the adjudged infringing products that the patentee previously contended, and proved, satisfy specific limitations of the asserted claims." Id. "Where one or more of the elements previously found to infringe has been modified or removed, the court must determine whether that modification is significant." Proveris, 739 F.3d at 1370-71. If the modification or removal is significant, then the newly accused product is more than colorably different from the infringing product, and the inquiry ends. Id. at 1371. At that point, "whether the newly accused product actually infringes is irrelevant," because "'the modifying party generally deserves the opportunity to litigate the infringement questions at a new trial.'" TiVo, 646 F.3d at 882 (quoting Arbek Mfg., Inc. v. Moazzam, 55 F.3d 1567, 1570 (Fed. Cir. 1995)); see also Proveris, 739 F.3d at 1371 (stating that if the newly accused product is more than colorably different, "a new infringement action must be brought regarding the newly accused product").

However, if a court determines that the newly accused product is not more than colorably different from the infringing product, "the court must then go on to the second step and determine whether the newly accused product in fact infringes the relevant claims." Proveris, 739 F.3d at 1371. "Thus, the court is required to evaluate the modified elements of the newly accused product against the asserted claim, on a limitation by limitation basis, to ensure that each limitation continues to be met." TiVo, 646 F.3d at 883. "In making this infringement evaluation, out of fairness, the district court is bound by any prior claim construction that it had performed in the case." Id.

As this Court previously observed, see ECF No. 2075 at 20, district courts have applied the "colorably different" standard in the ongoing royalty context. See, e.g., VirnetX Inc. v. Apple Inc., No. 6:13-CV-211, 2014 WL 12672822, at *5 (E.D. Tex. Mar. 6, 2014) (awarding ongoing royalty in patent infringement case based on sales of "adjudicated products and products not colorably different from those adjudicated at trial"). In Creative Internet Advertising Corp. v. Yahoo! Inc., 674 F. Supp. 2d 847 (E.D. Tex. 2009), the court recognized that "[t]he issue presented here—determining an ongoing royalty when Defendant argues that infringement has stopped—appears to be one of first impression." Id. at 853. The Creative Internet court looked to the injunction context for guidance and applied the "colorably different" standard as a result. Id. at 853-55; see also KFx Med. Corp. v. Arthrex Inc., No. 11cv1698 DMS (BLM), 2014 WL 11961953, at *3 (S.D. Cal. Feb. 18, 2014) (citing Creative Internet and applying "colorably different" standard to motion for ongoing royalty).

As another example, in Bianco v. Globus Medical, Inc., 53 F. Supp. 3d 929 (E.D. Tex. 2014), Federal Circuit Judge William Bryson (sitting by designation) ordered ongoing royalties for trade secret misappropriation. Judge Bryson noted that "[a]n order basing ongoing royalty payments on future sales of those three products implicitly extends to any products that are not colorably different from those products." Id. at 942. Judge Bryson continued, "[The defendant] cannot avoid its royalty obligations simply by . . . making some trivial and immaterial change in the products." Id. In addition, at least one other district court has explicitly applied TiVo's two-step test to a motion for ongoing royalties. See XpertUniverse, Inc. v. Cisco Systems, Inc., No. 09-157-RGA, 2013 WL 6118447, at *10-11 (D. Del. Nov. 20, 2013).

Moreover, here, both parties in their briefing on the instant motion and at the hearing asserted that TiVo provides the relevant rule. See Mot. at 5-6; Opp'n at 3-4. Accordingly, the Court will apply TiVo's two-step test in analyzing Apple's claim for ongoing royalties on the non-adjudicated products.

III. DISCUSSION

The Court first describes the adjudicated products and, where relevant, the theory of infringement that Apple relied on at trial. The Court then summarizes the design-arounds at issue. The Court then applies the TiVo two-step test to the non-adjudicated products.

A. The Adjudicated Products and Apple's Theories of Infringement at Trial

At trial, Apple asserted claim 9 of the '647 patent against Samsung's web browser and messenger applications in nine Samsung devices. Among these nine devices, three versions of the Android operating system were at issue: Gingerbread, Ice Cream Sandwich, and Jelly Bean. See Rebuttal Expert Report of Dr. Kevin Jeffay in Connection with Ongoing Royalties of U.S. Patent No. 5,946,647 ("Jeffay Rep."), ECF Nos. 2217-10, 2239-11 at ¶¶ 75-76. The Court first discusses the adjudicated browsers and then discusses the adjudicated messengers.

1. The Adjudicated Browsers

a. Gingerbread and Ice Cream Sandwich Browsers

In the Gingerbread and Ice Cream Sandwich browsers, structures such as phone numbers and e-mail addresses are automatically detected when a web page loads. Jeffay Rep. at ¶¶ 82-83. These structures are not displayed to the user any differently than any other text on the page. Id. For each structure, "there are two possible gestures that a user can perform: short tap and long press." Id. ¶ 79. If the user performs a short tap, the browser performs a default action. Id. ¶ 80. For example, if the user short-taps a phone number, the browser initiates the dialer application to dial the phone number. Id. If the user performs a long press—that is, if the user presses the structure for more than one second—then a "context menu" appears. Id. ¶ 84. The context menu lists different actions that a user can choose to take for the selected structure. Id. For example, if the user long-presses a phone number, the resulting context menu would provide the user the option to dial that number or add the phone number to a contacts list, among other options. See id. ¶¶ 84-85.

Depending on the user's choice, the system may also display another menu, which the parties refer to as a "resolver activity menu," for choosing which among the same type of applications should be used to perform the chosen action. For example, if the user long-presses on an email address and chooses from the context menu to send an email, a resolver activity menu might ask if the user would like to send the email using the default email program or Gmail. Id. ¶ 118.

b. Jelly Bean Browser

In the Jelly Bean browser, unlike in the Gingerbread and Ice Cream Sandwich browsers, structures are not automatically detected when a web page loads. Id. ¶ 86. Instead, "[t]he process of content detection does not begin until after a user interacts with a displayed web page by putting a finger down on the screen." Id. However, like in the Gingerbread and Ice Cream Sandwich browsers, if a user in the Jelly Bean browser long-presses a structure, then a context menu appears. Id. ¶¶ 88-89. Depending on the user's selection, a resolver activity menu may also appear.

c. Relevant Infringement Arguments at Trial

As is relevant here, Samsung argued that the Jelly Bean browser did not infringe claim 9 because it did not practice the limitation in claim 1 requiring a "user interface enabling the selection of a detected structure." Specifically, Samsung argued that structures in the Jelly Bean browser are not automatically detected. Rather, a structure is only detected after a user touches the structure. Thus, according to Samsung, "[b]ecause a user can only select text that has not been detected, the Accused Jelly Bean Browser Products do not infringe claim 1 of the '647 Patent or any of its dependents because they do not provide a 'user interface enabling the selection of a detected structure and a linked action.'" Samsung's Motion for Summary Judgment, ECF No. 805-4 at 3 (emphasis in original).

Apple's expert, Dr. Todd Mowry, testified at trial that "before the user touches the screen the structures, like phone numbers, have not yet been detected." ECF No. 1624 ("Trial Tr.") at 867:14-16. However, "[w]hen the user's finger first touches the screen, it immediately begins detecting structures." Trial Tr. 867:17-18. The detection occurs and the result of the detection is stored in memory while the user is performing the long press. See Trial Tr. 867:17-868:11; 868:23-870:8. Dr. Mowry testified that "[m]eanwhile, time goes by. The user eventually is holding down long enough that it becomes a selection through a press and hold." Trial Tr. 869:13-15. In other words, "As soon as they touch the screen, the software sets a timer, and when this timer goes off, if they've continued to hold their finger down in place, more or less, when the timer goes off, it says, okay, you've held it long enough that you've actually made a selection through press and hold." Trial Tr. 869:22-870:2. At that point, the context menu appears. See Trial Tr. 872:16-25. Thus, under Apple's theory of infringement for the Jelly Bean browser, the limitation requiring "selection of a detected structure" is satisfied because the detection of the structure occurs before the user's selection of that structure is complete. See Apple's Opposition to Samsung's Motion for Summary Judgment, ECF No. 853-4 at 5; Initial Expert Report of Dr. Todd C. Mowry in Connection with Ongoing Royalties ("Mowry 2017 Initial Report"), ECF No. 2217-9, at ¶¶ 42-43.

The other aspect of Apple's theory of infringement at trial that is relevant to the instant motion is that Dr. Mowry testified that the limitation in Claim 9 requiring "a pop-up menu of the linked actions" was satisfied by the context menu. See Mowry 2017 Initial Report ¶ 45; Trial Tr. 874:7-875:4. Before and at trial, Apple did not argue that the resolver activity menu satisfied the pop-up limitation in claim 9. See Mot. at 19 n.7 (arguing that "demonstrating infringement by the first pop-up menu in the Adjudicated Products was sufficient to prove Apple's case"); Opp'n at 23 (arguing that the resolver activity menu was always present but not previously accused); Mowry 2017 Initial Rep. ¶ 86; Jeffay Rep. ¶¶ 98, 111.

2. The Adjudicated Messengers

In all versions of the adjudicated messenger, structures are automatically detected upon receipt of a text message and displayed in blue underlined font. Jeffay Rep. ¶ 91. In Ice Cream Sandwich and Jelly Bean messenger, a short tap on a structure produces the context menu. Id. ¶¶ 92-93. In the Gingerbread messenger, a long press on a structure produces the context menu. Id. ¶¶ 94-95. Like in the adjudicated browser, if more than one application is capable of performing the selected action, a resolver activity menu appears. Id. ¶ 107. If the user long-presses in the Ice Cream Sandwich or Jelly Bean messenger, then a generic pop-up menu appears that includes options such as "delete message," or "copy message text." Id. ¶ 97. Apple did not accuse the generic pop-up menu of infringement. Id.

B. Samsung's Design-Arounds

Samsung implemented three changes that it alleges design around the '647 patent. Samsung refers to these changes as DA1-A, DA1-B, and DA2. In each device, Samsung implemented one design-around for the browser and one design-around for the messenger. Jeffay Rep. ¶ 147. Thus, depending on the device, the browser received either the DA1-A or DA1-B design-around, and the messenger received either the DA1-A or the DA2 design-around. Id. The Court describes the DA1-A, DA1-B, and DA2 design-arounds in turn.

1. DA1-A

Samsung implemented the DA1-A change in the browser and messenger applications. DA1-A makes four primary changes that Samsung contends design around the '647 patent by changing the products so they no longer practice the limitation in claim 1 requiring "a user interface enabling the selection of a detected structure and a linked action." See Mowry 2017 Initial Rep. ¶ 46. First, in devices implementing DA1-A, structures are not automatically detected. Mowry 2017 Initial Rep. ¶¶ 57, 63; Jeffay Rep. ¶¶ 150, 165. This is a change from the Gingerbread and Ice Cream Sandwich browsers and all versions of messenger, in which structures were automatically detected before any user interaction. Second, in devices implementing DA1-A, a long press in the browser produces a generic pop-up menu, which was not accused, rather than the accused context menu. Mowry 2017 Initial Rep. ¶ 64; Jeffay Rep. ¶¶ 152-53. This is a change from all versions of the browser and the Gingerbread messenger, in which a long press produced the context menu. Third, in devices implementing DA1-A, a short tap on a structure produces a context menu. Mowry 2017 Initial Rep. ¶¶ 58, 65; Jeffay Rep. ¶ 164. This is a change from all versions of the browser and the Gingerbread messenger, in which a long press produced the context menu. Fourth, in devices implementing DA1-A, the detection process begins at a different point in the user's interaction with the structure. Specifically, Samsung altered the software code so that the detection of the structure occurs after a user lifts his or her finger up from the screen after performing a short tap, as opposed to occurring when the user put his or her finger down onto the screen, as was the case in the adjudicated products. Mowry 2017 Initial Rep. ¶ 95; Jeffay Rep. ¶¶ 150, 155, 165, 170-71. This is a change from all versions of the browser and messenger. Dr. Mowry contends that this last change does not make a difference from the user's perspective because detection occurs instantaneously. Mowry 2017 Initial Rep. ¶ 99.

As a result of these changes, Samsung argues that the newly accused DA1-A devices do not satisfy the limitation in claim 1 requiring "a user interface enabling the selection of a detected structure and a linked action" because detection does not occur until the user lifts his or her finger up from the screen after performing a short tap. As a result, Samsung argues that no structure is detected until after the user's selection of the structure is complete. Thus, a user cannot select a detected structure because the plain and ordinary meaning of "selection of a detected structure" requires detection to occur before selection. See Opp'n at 16-17.

2. DA1-B

Samsung implemented DA1-B only in the browser application. Mowry 2017 Initial Rep. ¶ 74. Like DA1-A, DA1-B makes three primary changes that Samsung contends are relevant in bringing the DA1-B products outside the scope of claim 1 of the '647 patent. The first two of these changes are the same as in DA1-A, and the third differs somewhat. Namely, first, like in DA1-A, structures in DA1-B are not automatically detected. Mowry 2017 Initial Rep. ¶ 75; Jeffay Rep. ¶¶ 173, 183. Second, like in DA1-A, a long press in a DA1-B browser produces a generic pop-up menu, rather than a context menu. Mowry 2017 Initial Rep. ¶ 75; Jeffay Rep. ¶¶ 176-77, 181, 237. Third, like in DA1-A, a short tap on a structure in a DA1-B browser produces a context menu. Mowry 2017 Initial Rep. ¶¶ 76, 101; Jeffay Rep. ¶ 182. However, the timing of structure detection differs between DA1-A and DA1-B. Whereas detection of structures occurs in DA1-A when a user lifts his or her finger up from the screen, detection in DA1-B occurs when a user puts his or her finger down on the screen. Mowry 2017 Initial Rep. ¶¶ 76, 101; Jeffay Rep. ¶ 184. As a result, the sequence of events in DA1-B is more like the adjudicated Jelly Bean browser, where detection occurs when the user puts his or her finger down on the screen, than like DA1-A, where detection occurs when the user lifts his or her finger up from the screen. Mowry 2017 Initial Rep. ¶ 101.

Samsung contends that DA1-B devices do not satisfy the limitation in Claim 1 requiring "a user interface enabling the selection of a detected structure and a linked action" because detection does not begin until a user interacts with a structure. Opp'n at 20. Although detection in some of the infringing devices also began after user interaction, Samsung distinguishes the DA1-B devices by pointing out that the infringing devices required a long press. Apple argued at trial that this long press meant that detection completed before the long press completed, which meant that the structure was detected by the time the user's selection of the structure was complete. See id. Samsung contends that Apple has offered no proof that detection in DA1-B devices completes before selection is complete. As a result, under any burden of proof, Samsung argues that Apple has not carried its burden to prove that the DA1-B products infringe. Id.

3. DA2

Finally, Samsung implemented DA2 only in the messenger application. Mowry 2017 Initial Rep. ¶ 50. In devices implementing DA2, structures are still automatically detected and displayed in highlighted font. Id. ¶ 81. However, in devices implementing DA2, the context menu is removed entirely. Id. ¶ 51; Jeffay Rep. ¶¶ 200-01, 247. Instead, a short tap on a structure in a DA2 messenger performs a predetermined action associated with that structure. Mowry 2017 Initial Rep. ¶ 51; Jeffay Rep. ¶¶ 202, 247-48. For example, a short tap on a phone number in a DA2 messenger will dial that phone number. However, if there are multiple applications capable of carrying out the predetermined action, then a resolver activity menu will appear for the user to select which program to use. Mowry 2017 Initial Rep. ¶¶ 52-53, 82; Jeffay Rep. ¶¶ 215-16. For instance, if a user short-taps on an email address, a resolver activity menu would appear if the user has more than one email program installed. The resolver activity menu would then allow the user to choose whether to send the email using Gmail or the default email program, for example. Mowry 2017 Initial Rep. ¶ 53. A long press in a DA2 messenger produces a generic pop-up menu, which Apple has not accused of infringement. Jeffay Rep. ¶¶ 203-04.

Samsung contends that DA2 products do not satisfy the limitation in claim 9 requiring a user interface that "display[s] a pop-up menu of the linked actions" for several reasons. Opp'n at 25. First, Samsung argues that DA2 products remove the context menu altogether. Second, Samsung argues that Apple previously admitted that a single-action-per-structure functionality did not infringe. See id. Third, Samsung argues that the plain language of claim 9 requires multiple linked actions, but the resolver activity menu offers different methods of performing only one action. See id. at 25-26. C. TiVo Analysis of Design-Arounds

The Court now applies TiVo's two-step analysis to each design-around. As explained above, the Court first must determine whether the newly accused product is "more than colorably different from the product found to infringe." TiVo, 646 F.3d at 882. This inquiry "focus[es] on those elements of the adjudged infringing products that the patentee previously contended, and proved, satisfy specific limitations of the asserted claims." Id. In other words, this analysis "focuses on how the patentee in fact proved infringement, not what the claims require." Ncube Corp. v. SeaChange Int'l, Inc., 732 F.3d 1346, 1351 (Fed. Cir. 2013). "Where one or more of the elements previously found to infringe has been modified or removed, the court must determine whether that modification is significant." Proveris, 739 F.3d at 1370-71. If the modification or removal is significant, then the newly accused product is more than colorably different from the infringing product, and the inquiry ends. Id. at 1371.

In TiVo, the Federal Circuit highlighted several considerations it saw as important to the "colorably different" standard. The Federal Circuit observed that "[t]he significance of the differences between the two products is much dependent on the nature of the products at issue." 646 F.3d at 882. "The court must also look to the relevant prior art, if any is available, to determine if the modification merely employs or combines elements already known in the prior art in a manner that would have been obvious to a person of ordinary skill in the art at the time the modification was made." Id. The Federal Circuit explained, "A nonobvious modification may well result in a finding of more than a colorable difference." Id. at 882-83. The Federal Circuit also stated that "[t]he analysis may also take account of the policy that legitimate design-around efforts should always be encouraged as a path to spur further innovation. But an assertion that one has permissibly designed around a patent should not be used to mask continued infringement." Id. at 883.

With these considerations in mind, the Court next considers the parties' arguments about whether Samsung's design-arounds render the newly accused DA1-A, DA1-B, and DA2 products more than colorably different from the adjudicated products. If the Court finds that a newly accused product is not more than colorably different from the adjudicated products, the Court then must determine whether the newly accused product continues to infringe.

1. DA1-A

a. DA1-A Browser Products

i. TiVo Step One: Is the Newly Accused Product More than Colorably Different?

1. Which Features Are the Focus of the Comparison?

As an initial matter, the parties appear to dispute which aspects of the newly accused browser products are relevant for purposes of the colorably different inquiry. As TiVo makes clear, the Court must compare "those elements of the adjudged infringing products that the patentee previously contended, and proved, satisfy specific limitations of the asserted claims," TiVo, 646 F.3d at 882, with whatever feature(s) replaced those elements, id. at 884. Apple contends that the feature in the newly accused DA1-A products that should be the subject of the colorably different inquiry is the context menu that is produced when a user short-taps a structure. Reply at 4. Samsung argues that "the relevant accused feature [in the adjudicated browser products] was a context menu of options that is generated following a long press gesture on structured text." Opp'n at 11. Samsung suggests in passing that because Samsung changed the result of a long press from a context menu to a generic pop-up menu, the generic pop-up menu should be the focus of the colorably different analysis. Opp'n at 11.

Samsung's argument is not persuasive. As Samsung repeatedly points out in its opposition, Apple did not previously accuse the generic pop-up menu or the resolver activity menu of infringing claim 9 of the '647 patent. See, e.g., id. at 2, 6, 7 & n.2, 8, 9. The context menu—whether produced as the result of a short tap or a long press—was the focus of Apple's claim 9 infringement theory. In DA1-A (and DA1-B), Samsung retained the context menu but, depending on the product, made changes to the user action that produces the context menu and/or the timing of the structure detection. See Mowry Reply Rep. ¶ 27 ("Samsung did not remove th[e] [context menu] functionality, but simply moved it so that the same contextual pop-up menu now appears on a short tap, rather than a long press."). Accordingly, in DA1-A (and DA1-B), the newly accused products' context menu is the relevant subject of the colorably different analysis. See Ncube, 732 F.3d at 1350 (holding that the ClientID remained the relevant feature for comparison in newly accused products where the ClientID was moved but still performed the same relevant function as in the infringing products).

The Court must thus determine whether the context menu feature in the DA1-A browser products is more than colorably different from the context menu feature in the adjudicated browser products.

2. DA1-A Browser Products Are Not More Than Colorably Different From the Adjudicated Products

After considering the parties' briefing, the expert reports, the parties' arguments during the January 11, 2018 hearing, and the videos that Apple presented at the January 11, 2018 hearing, the Court concludes that the DA1-A browser products are not more than colorably different from the adjudicated browser products. First, from a technical perspective, the code modification necessary to change from a long press to a short tap "is trivial" and "amounts to little more than adjusting the calls to the functions to generate and display the contextual pop-up menu from a timer-based determination that there has been a long press to a timer-based determination that there has been a short tap." Mowry Reply Rep. ¶ 27; see also Mowry 2017 Initial Report ¶ 98 [Redacted] . Similarly, the code modification to change the timing of structure detection from "finger down" to "finger up" "is minor." Mowry 2017 Initial Report ¶ 98. In addition, the code used to perform the structure detection [Redacted] . Jeffay Dep. 143:18. Samsung does not dispute that the changes in its code were relatively simple.

Relatedly, under TiVo, the Court "must also look to the relevant prior art, if any is available, to determine if the modification merely employs or combines elements already known in the prior art in a manner that would have been obvious to a person of ordinary skill in the art at the time the modification was made." 646 F.3d at 882. When the Court asked the parties at the January 11, 2018 hearing to identify the prior art relevant to the TiVo obviousness inquiry, neither party identified any specific prior art. Transcript of January 11, 2018 Hearing ("Hearing Tr."), ECF No. 2238 at 50:2-5; id. at 67:1-2. Rather, Apple argued generally that the modifications were obvious. Id. at 48:11-49:14. Similarly, Apple's expert, Dr. Mowry, opines that "the minor technical modifications Samsung implemented in its design-arounds would have been obvious to a person of ordinary skill in the art at least as early as May 5, 2014, when the Infringing Products were found to infringe the '647 patent." Mowry 2017 Initial Rep. ¶ 106. Samsung did not dispute at the hearing or in its opposition that the modifications were obvious. Hearing Tr. at 66:17-67:2; Opp'n at 3-4. Dr. Jeffay contends that because Dr. Mowry did not provide a basis for his opinion on obviousness, Dr. Jeffay is "not in a position to rebut his unfounded opinion." Jeffay Rep. ¶ 226 n.6. However, Dr. Jeffay does not offer a separate opinion on the obviousness of the changes. See Jeffay Dep. 34:4-35:23.

Rather than dispute the obviousness of its changes, Samsung argues that a simple or obvious change can nonetheless render a newly accused product more than colorably different from an adjudicated product. Opp'n at 3-4. As a result, Samsung contends that Apple misreads TiVo by asserting that an obvious modification is unlikely to be significant. Id. The Court agrees that the obviousness of a change is not necessarily dispositive of whether that change represents more than a colorable difference from an adjudicated product. See TiVo, 646 F.3d at 882 n.1 ("We do not suggest that the law on obviousness is binding in contempt proceedings, where, in most cases, a single limitation that has been modified by an infringer is at issue."). That said, it is clear that the Federal Circuit views the "innovative significance of the modification" as a potentially relevant factor in the colorably different analysis. See id. In addition, the Federal Circuit observed that a non-obvious change "may well result in a finding of more than a colorable difference," but it did not say that the reverse is necessarily true. TiVo, 646 F.3d at 882-83. Indeed, as Part III.C.3 of this order demonstrates, the total removal of an infringing element may be an obvious change, but also a significant one. Instead, the appropriate inference to be drawn from a finding that a change is obvious likely depends on the circumstances of the case. Here, the Court finds that the obviousness of the DA1-A changes are neutral.

In addition, the Court finds that the DA1-A browser changes are not significant from a functional perspective or a user's perspective. From both a functional and a user's perspective, it is undisputed that the context menu itself is essentially the same in the DA1-A browser and adjudicated browser products and provides the user with the same functionality. With regard to the difference between a short tap and a long press, Apple compares this difference to the difference between a right click or left clink on a computer mouse. Mot. at 9. "In both instances, the user is making an intentional selection and indicating through the choice of input the desired result." Id. Thus, both a short tap and a long press serve the same function of selecting a structure. Id.

With respect to the timing of structure detection, there is no noticeable difference from the user's perspective whether the structure is detected on "finger down" or "finger up." "Because of the speed with which structures are detected in the devices, 'the display of the contextual pop-up menu in the DA1-A products is as instantaneous, from a users' perspective, as in the infringing Galaxy S III.'" Mot. at 10 (quoting Mowry 2017 Initial Rep. ¶ 99). Indeed, at the hearing on January 11, 2018, Apple presented videos of a user selecting a phone number in the browser of the adjudicated Galaxy S III and the newly accused Galaxy S 5, which includes the DA1-A changes. See Hearing Tr. at 8:16-9:24; Apple Ongoing Royalties Hearing Slidedeck at Slide 25. The Court can discern no difference between the two phones in the amount of time that elapses between the selection of the phone number and the appearance of the context menu.

Samsung has not disputed the accuracy of these videos.

Samsung argues that the functional perspective and user's perspective do not play a role in the colorably different analysis. Opp'n at 13 (citing TiVo, 646 F.3d at 882); Hearing Tr. 43:3-9 ("And whether it's perceptible to the user is not relevant. Can you imagine if we decided utility patent code design arounds based on the user experience?"). Of course, two products could be more than colorably different even though a user perceives no difference between the products. See Ncube, 732 F.3d at 1350-51. However, Federal Circuit precedent suggests that whether a change is significant from a functional perspective or the user's perspective may inform the colorably different analysis, even if such perspectives are not dispositive.

For example, the Federal Circuit looked to whether two ingredients functioned interchangeably in its colorably different analysis in Merial Ltd. v. Cipla Ltd., 681 F.3d 1283 (Fed. Cir. 2012), which dealt with "patent compositions for protecting domestic dogs and cats from infestation with ectoparasites, e.g., fleas and ticks." Id. at 1288. The claim at issue in Merial required "a spot-on pest control composition comprising (1) a synergistic effective amount of fipronil, (2) a synergistic effective amount of an [insect growth regulator], and (3) at least one customary spot-on formulation adjuvant." Id. at 1300. The infringing product contained 9.7% fipronil, 11.8% methoprene, and at least one customary spot-on adjuvant. Id. "By comparison, the district court found during the contempt proceedings that [the newly accused product] contain[ed] 9.8% fipronil, 11.8% methoprene, and at least one customary spot-on adjuvant." Id. "Furthermore, the court credited Merial's expert testimony that any two spot-on adjuvants would function interchangeably and that replacing one for another in any such pest control composition would not amount to a colorable difference." Id. at 1300-01. The Federal Circuit concluded that the "district court thus had ample basis" to conclude that the newly accused product was not more than colorably different from the infringing product. Id. at 1301.

Relatedly, in Proveris, the Federal Circuit considered both functional equivalence and the user's perspective in analyzing whether an allegedly redesigned mechanism for evaluating aerosol spray plumes was more than colorably different from the infringing mechanism. 739 F.3d at 1369. The defendant argued that the infringing product "allowed a user to identify what range of images he or she wanted to analyze before activating the spray plume, while the [newly accused] device requires the user to first activate the spray plume and then later determine what range of images he or she would like to analyze." Id. at 1370. The defendant "contend[ed] that this is a significant modification" because the preamble of the relevant claim "specifie[d] that the image data may be captured 'at a predetermined instant in time.'" Id. The district court found that the newly accused device was not more than colorably different from the infringing device, and the Federal Circuit affirmed. Id. at 1370. Specifically, the Federal Circuit observed that "it is not at all clear from the record whether [the defendant's] purported change actually had any effect. In fact, the User Manuals for both products appear to instruct the user to select the range of images to be analyzed after the actual spray event takes place." Id. at 1371. The Federal Circuit went on, "even if [the defendant] did make some small changes to the product's software, a comparison of the User Manuals demonstrates that the two products are functionally identical." Id. Thus, the Federal Circuit agreed with the district court that the newly accused product was not more than colorably different from the infringing product based on differences that were indiscernible from a functional perspective and a user's perspective. Id.

Thus, under Merial and Proveris, both the functional effect of any changes and the significance of those changes from a user's perspective inform the Court's colorably different analysis. Cf. TiVo, 646 F.3d at 878-79 (noting but not reaching district court's finding that newly accused product's automatic flow control feature was not more than colorably different from infringing product based on district court's analysis of the functional effect of the change). Here, the Court finds that both the functional effect of the DA1-A changes and their significance from a user's perspective show that the DA1-A browser is not more than colorably different from the adjudicated products. Neither change affects the functionality that the context menu provides the user, and neither change affects the user's experience of the context menu feature.

Indeed, Samsung's argument heavily depends on the proposition that a change that brings a product outside the literal scope of the patent is by definition a significant change. Opp'n at 13-14. Specifically, Samsung argues that Apple's reliance on the doctrine of equivalents at TiVo's second step shows that the DA1-A changes bring DA1-A products outside the literal scope of the '647 patent. Id. Such a change, Samsung argues, must be significant. However, Samsung cites no authority for such a per se rule. Indeed, the Federal Circuit has stated—and Samsung argued at the hearing—that the colorably different analysis compares the newly accused product to the adjudicated product, not the newly accused product to the claim. See Ncube, 732 F.3d at 1351; Hearing Tr. 55:13-19. Using the claim to determine whether a change renders a newly accused product more than colorably different would "collapse the [two prong] test of TiVo—that is, the only inquiry would be whether the newly accused products infringe the asserted patent." Tinnus Enters., LLC v. Telebrands Corp., No. 6:15-cv-551 RWS-JDL, 2016 WL 1084800, at *3 (E.D. Tex. Feb. 23, 2016). Because the Federal Circuit specifically said that it "reject[ed] [an] infringement-based understanding of the colorably different test," TiVo, 646 F.3d at 882, Samsung's argument fails.

Accordingly, the Court finds that the DA1-A browser products are not more than colorably different from the adjudicated products. As a result, the Court must next analyze whether the DA1-A browser products infringe claim 9 of the '647 patent.

ii. TiVo Step Two: Does It Infringe?

Apple argues that the DA1-A products infringe Claim 9 of the '647 patent under the doctrine of equivalents. Mot. at 12-13. "Under the doctrine of equivalents, 'a product or process that does not literally infringe upon the express terms of a patent claim may nonetheless be found to infringe if there is equivalence between the elements of the accused product or process and the claimed elements of the patented invention.'" DePuy Spine, Inc. v. Medtronic Sofamor Danek, Inc., 469 F.3d 1005, 1016 (Fed. Cir. 2006) (quoting Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 21 (1997)).

Samsung contends that its DA1-A products do not infringe because they do not satisfy the limitation in claim 1 that requires "a user interface enabling the selection of a detected structure and a linked action." Opp'n at 16. As explained above, Samsung's theory is that to satisfy that user interface limitation, detection of a structure must occur before selection of the structure. Id. at 16-17. Because detection occurs on "finger up" rather than "finger down," Samsung argues that selection is complete in DA1-A products before detection occurs. Opp'n at 17. Apple effectively concedes that the DA1-A products do not literally satisfy the user interface limitation by only arguing that the DA1-A products infringe under the doctrine of equivalents. See Mot. at 12.

Samsung objects to Apple's reliance on the doctrine of equivalents at this stage in the proceedings. Specifically, Samsung argues that Apple waived any reliance on the doctrine of equivalents by withdrawing the portions of Dr. Mowry's report relating to the doctrine of equivalents before trial. Opp'n at 15 (citing ECF No. 1056 at 3 n.4). In addition, Samsung argues that Apple did not assert any doctrine of equivalents theory at trial. Id. Finally, Samsung argues that allowing Apple to raise the doctrine of equivalents only after trial would undercut the Local Patent Rules and prejudice Samsung by depriving Samsung of its opportunity to present its defenses to the jury. Id. at 16. Thus, the Court must determine whether a party may rely on the doctrine of equivalents to prove infringement in post-trial proceedings if that party relied only on a literal infringement theory at trial.

1. Can Apple Rely on Doctrine of Equivalents?

Samsung's contention that a party is limited in post-trial proceedings to the theory of infringement it used at trial is not persuasive. Samsung cites no on-point authority for its position that a party may only prove infringement in post-trial contempt or ongoing royalty proceedings using the theory of infringement that the party relied upon at trial. At least two courts have applied the doctrine of equivalents during contempt proceedings for violation of a preliminary injunction that was based on literal infringement. See Chamberlain Grp., Inc. v. Techtronic Indus. Co., No. 16 C 6097, 2017 WL 368027, at *4-6 (N.D. Ill. Jan. 23, 2017); Aevoe Corp. v. AE Tech Co., No. 2:12-cv-53-GMN-RJJ, 2012 WL 1559768, at *3 (D. Nev. May 2, 2012). Moreover, in Bass Pro Trademarks, L.L.C. v. Cabela's, Inc., 485 F.3d 1364, 1368 (Fed. Cir. 2007), the Federal Circuit stated that "the grant of a contempt order for violation, by a modified device, of an injunction against infringement requires that the modified device infringes the patent, either literally or by application of the doctrine of equivalents," which suggests that both avenues are available to prove infringement in post-trial proceedings.

Furthermore, it makes little sense to say that Apple's actions before and at trial waived theories of infringement for design-around products that did not yet exist at the time of trial. Indeed, had Apple attempted to introduce evidence at trial to prove that hypothetical design-around products infringed the patent under any theory of infringement, the Court would have excluded such evidence under Federal Rule of Evidence 403 as confusing the issues, misleading the jury, causing undue delay, and wasting time. As to Samsung's notice and trial right argument, the Federal Circuit stated in Ncube that TiVo's colorably different step "preserves values of notice and preservation of trial rights by keeping contempt suitably limited." 732 F.3d at 1351. In other words, the Federal Circuit has decided that the first TiVo step protects litigants' rights by significantly restricting the range of newly accused products that are subject to a post-trial infringement analysis to those newly accused products that are less than colorably different from the adjudicated products. Thus, Apple may rely on the doctrine of equivalents even though Apple did not assert the doctrine of equivalents at trial. The Court next analyzes whether the DA1-A browser products infringe under the doctrine of equivalents.

2. Does DA1-A Infringe Under the Doctrine of Equivalents?

"The doctrine of equivalents prohibits one from avoiding infringement liability by making only 'insubstantial changes and substitutions . . . which, though adding nothing, would be enough to take the copied matter outside the claim, and hence outside the reach of law.'" Siemens Med. Sols. USCA, Inc. v. Saint-Gobain Ceramics & Plastics, Inc., 637 F.3d 1269, 1279 (Fed. Cir. 2011) (quoting Graver Tank & Mfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 607 (1950)). "[H]owever, the doctrine of equivalents is not a license to ignore or 'erase . . . structural and functional limitations of the claim,' limitations 'on which the public is entitled to rely in avoiding infringement.'" Athletic Alternatives, Inc. v. Prince Mfg., Inc., 73 F.3d 1573, 1582 (Fed. Cir. 1996) (quoting Perkin-Elmer Corp. v. Westinghouse Elec. Corp., 822 F.2d 1528, 1532 (Fed. Cir. 1987) (second alteration in original)).

"[I]n an effort to strike the proper balance between protecting patentees while also providing sufficient notice to the public, various rules of law have emerged to constrain when and how the doctrine of equivalents is to be applied." Freedman Seating Co. v. Am. Seating Co., 420 F.3d 1350, 1358 (Fed. Cir. 2005). "One limit on the doctrine of equivalents is the 'all elements' rule." DePuy Spine, 469 F.3d at 1016. The "all elements" rule "require[es] that equivalence be assessed on a limitation-by-limitation basis, rather than from the perspective of the invention as a whole." Id.; see also Warner-Jenkinson, 520 U.S. at 40 ("The determination of equivalence should be applied as an objective inquiry on an element-by-element basis."). The "all elements" rule also provides that "an element of an accused product or process is not, as a matter of law, equivalent to a limitation of the claimed invention if such a finding would entirely vitiate the limitation." Freedman Seating, 420 F.3d at 1358.

"There is no set formula for determining whether a finding of equivalence would vitiate a claim limitation, and thereby violate the ['all elements'] rule." Id. at 1359. "Rather, courts must consider the totality of the circumstances of each case and determine whether the alleged equivalent can be fairly characterized as an insubstantial change from the claimed subject matter without rendering the pertinent limitation meaningless." Id. However, the Federal Circuit has explained that "[t]he vitiation concept has its clearest application 'where the accused device contain[s] the antithesis of the claimed structure.'" Brilliant Instruments, Inc. v. GuideTech, LLC, 707 F.3d 1342, 1347 (Fed. Cir. 2013) (quoting Planet Bingo, LLC v. GameTech Int'l, Inc., 472 F.3d 1338, 1345 (Fed. Cir. 2006)). Put differently, "equivalency cannot embrace a structure that is specifically excluded from the scope of the claims." Athletic Alternatives, 73 F.3d at 1582 (quotation marks omitted). Subject matter is "specifically excluded" if "its inclusion is somehow inconsistent with the language of the claim." Ethicon Endo-Surgery, Inc. v. U.S. Surgical Corp., 149 F.3d 1309, 1317 (Fed. Cir. 1998). For example, in Moore U.S.A. v. Standard Register Co., 229 F.3d 1091, 1106 (Fed. Cir. 2000), the Federal Circuit held that a mailer whose adhesive covered a minority of its edge could not be the equivalent of a claim limitation requiring that the adhesive cover a majority of the mailer's edge. Not only would such reasoning vitiate a claim limitation, but "it would defy logic to conclude that a minority—the very antithesis of a majority—could be insubstantially different from a claim limitation requiring a majority." Id.

The Federal Circuit's application of the "all elements" rule in Planet Bingo is particularly instructive. Planet Bingo was the exclusive licensee of a patent for "alternative methods of playing bingo by coupling numbers with additional 'indicia' or 'markings,' such as colors or shading patterns." 472 F.3d at 1340. "These additional designations overlay a traditional bingo game to produce more winning combinations for more prizes." Id. The relevant claim of the patent at issue included limitations requiring "establishing a predetermined combination as a winning combination for the progressive jackpot pool" and "awarding the progressive jackpot pool to the player when he achieves the predetermined winning combination on the bingo card." Id. at 1342 (quoting U.S. Patent No. 5,482,289). The district court construed "predetermined winning combination" to mean "the precise elements necessary to achieve bingo in a particular game are known before the first bingo ball is drawn." Id. at 1341.

GameTech argued that its game did not infringe because players of its product did not know the winning combination until after the first bingo ball is drawn. Id. at 1342. Planet Bingo countered that GameTech's game infringed under the doctrine of equivalents because GameTech's game "incorporates only an insubstantial variation from the claims because the progressive predetermined winning combination appears right after, rather than right before, the first bingo ball is drawn." Id. at 1344. Planet Bingo argued that "the timing of the predetermined winning combination does not affect the game's prize amounts, odds, or other essential characteristics." Id.

The district court rejected Planet Bingo's argument. Specifically, the district court determined that the claims at issue "require[d] a predetermined winning combination, which was construed to mean before the first bingo ball is drawn. After is the opposite of before, not its equivalent." Planet Bingo, LLC v. GameTech Int'l, Inc., CV-S-01-1295-PMP (PAL), ECF No. 379 at 16. Accordingly, GameTech's game, "in which the winning combination is determined after the first bingo ball is drawn[,] cannot be the equivalent of the predetermined limitation as a matter of law." Id.

The Federal Circuit affirmed. The Federal Circuit held that the "predetermined" limitation "was part of the bargain when the patent issued." Planet Bingo, 472 F.3d at 1344. Citing Moore, the Federal Circuit observed that the Federal Circuit "has refused to apply the doctrine [of equivalents] in other cases where the accused device contained the antithesis of the claimed structure." Id. at 1345. Accordingly, the Federal Circuit refused to "overlook that limitation or expand the doctrine of equivalents beyond its purpose to allow recapture of subject matter excluded by a deliberate and foreseeable claim drafting decision." Id.

Here, the Court finds that Apple's doctrine of equivalents theory would vitiate the user interface limitation requiring "selection of a detected structure." Specifically, as both parties acknowledge, the parties previously agreed that the plain and ordinary meaning of "selection of a detected structure" requires detection of the structure to occur before selection. See ECF No. 1151 at 20; Opp'n at 5-6; Reply at 9-10. In the DA1-A products, detection of the structure occurs after selection. As the Federal Circuit wrote in Planet Bingo, "after is opposite of before, not equivalent." Planet Bingo, 472 F.3d at 1344; cf. Wi-LAN, Inc. v. Apple, Inc., 811 F.3d 455, 462 (Fed. Cir. 2016) (Apple defending jury verdict in its favor by arguing that its products do not infringe a patent for a wireless communication technique under the doctrine of equivalents because its products performed the same processes in a different order than required by the claims, even though the differences in order had no functional effect). Of course, both the doctrine of equivalents and vitiation are fact-intensive, case-specific inquiries. See Cadence Pharm. Inc. v. Exela PharmSci Inc., 780 F.3d 1364, 1371 (Fed. Cir. 2015); Freedman Seating, 420 F.3d at 1359. As a result, the Court does not rely solely on Planet Bingo, although this precedent supports the Court's conclusion. Rather, the Court finds that the '647 patent's specification and Apple's previous positions in this case confirm that the parties' interpretation of the plain and ordinary meaning of the user interface limitation was correct, and that the DA1-A products, in which selection occurs before detection, do not satisfy the user interface limitation by the doctrine of equivalents.

First, throughout the specification of the '647 patent, Apple described detection as occurring before selection and at times emphasized the importance of this order of operations. In other cases, the Federal Circuit has viewed a claim construction's consistency with the specification as evidence that the construction was correct. See, e.g., Wi-LAN, 811 F.3d at 462 (concluding that the "intrinsic record is therefore clear that the asserted claims cover only structure that randomizes data symbols in parallel before combining them" because "[e]very embodiment discussed in the specification randomizes the data symbols before combining them" and "[n]o disclosure in the specification depicts or discusses the possibility of combining before randomizing"). The description of the preferred embodiment consistently refers to detection occurring before selection. Here, for example:

After identifying structures and linking actions, application program interface 230 communicates with application 167 to obtain information on the identified structures so that user interface 240 can successfully present and enable selection
of the actions. In a display-type environment, application program interface 230 retrieves the locations in document 210 of the presentation regions for the detected structures from application 167. Application program interface 230 then transmits this location information to user interface 240, which highlights the detected structures, although other presentation mechanisms can be used. User interface 240 enables selection of an identified structure by making the presentation regions mouse-sensitive . . . .
'647 Patent, col.4:1-13. Similarly, Figures 8 and 9, which "display a flowchart illustrating [the] preferred method for recognizing patterns in documents and performing actions," clearly envision detection of structures occurring before selection of those same structures. See id. at col.5:51-col.6:21; Figs. 8-9. "No disclosure in the specification depicts or discusses the possibility of [selection] before [detection]." Wi-LAN, 811 F.3d at 462. Moreover, not only does the specification only disclose detection occurring before selection, but the specification also stresses the importance of this order in explaining the significance of the invention. Specifically, the specification stressed that the automatic detection and highlighting of structures helped a user visually identify structures in a long document or email, which was a "laborious and cognitively disruptive" process without automatic detection. '647 patent at col.1:16-27. Thus, the specification reveals that detection occurring first was a significant feature of the invention.

This conclusion is reinforced by the expert declaration that Dr. Mowry filed in May 2012. See ECF No. 177 ("Mowry 2012 Decl."). In that declaration, in distinguishing U.S. Patent No. 5,859,636 ("Pandit"), Dr. Mowry emphasized that Pandit required the user to select text before the system recognized that text, whereas the '647 patent required the opposite order. Mowry 2012 Decl. ¶¶ 208, 217. Even assuming that Pandit is not prior art to the '647 patent, and thus cannot serve as the basis for an ensnarement argument, how Dr. Mowry analyzed Pandit in comparison to the '647 patent provides insight into how Apple and its expert initially conceived the scope and importance of various aspects of the '647 patent. Dr. Mowry wrote in 2012:

The plain and ordinary meaning of "a user interface enabling the selection of a detected structure" requires the user interface to enable selection of a structure, by the user, after the structure has already been detected. Pandit, by contrast, reverses these steps. For example, Dr. Cohen cites to one portion of Pandit that discusses the invention "recognize[ing]" text that the user has previously "accented", or selected. Dr. Cohen thus argues that the [user interface] element of
claim 1 is satisfied when (1) the user first selects undetected text and, (2) after such selection, structures are detected in that text. This argument makes no sense in view of the plain language of claim 1, which, as explained, requires that the user be able to pick or choose a detected structure after the system identifies such structures for the user.
Mowry 2012 Decl. ¶ 217 (alterations and emphasis in original). Thus, Dr. Mowry's declaration makes clear that he previously viewed the order in which detection and selection occurred as significant—not as an "insubstantial[] differen[ce]," as he later characterized it. See Mowry 2017 Initial Rep. ¶ 136; Initial Expert Report of Dr. Todd C. Mowry Regarding Infringement of U.S. Patent No. 5,946,647 (August 12, 2013), at ¶¶ 252-54 (opining that Samsung products where detection routines are invoked after the user touches the screen infringe the '647 patent under the doctrine of equivalents).

Accordingly, the Court concludes that the DA1-A products, in which selection occurs before detection, cannot "be fairly characterized as an insubstantial change from the claimed subject matter without rendering the pertinent limitation meaningless." Freedman Seating, 420 F.3d at 1358; see also Planet Bingo, 472 F.3d at 1345 ("[T]he proposed application of the doctrine of equivalents would change 'before' to 'after,' a more marked difference" than a "small variation[] in the degree of achieving a claimed limitation."). As a result, the DA1-A browser products do not infringe the '647 patent under the doctrine of equivalents.

Apple argues that its proposed equivalent would not vitiate the user interface limitation because structures are still detected. See Reply at 11. However, this argument ignores that the construction of the user interface limitation requires that detection occur before detection. Any vitiation analysis must take into account any claim construction that has occurred. See Planet Bingo, 472 F.3d at 1344 (taking claim construction into account in vitiation analysis). Thus, Apple's argument that focuses only on the language of the claim and ignores the construction of that claim is unavailing. --------

b. DA1-A Messenger Products

With respect to the DA1-A messenger products, the parties dispute whether the relevant adjudicated product for purposes of the colorably different analysis is the adjudicated messenger product or the adjudicated Jelly Bean browser. See Mot. at 11 (comparing DA1-A messenger products to the adjudicated Jelly Bean browser); Opp'n at 14 (arguing that DA1-A messenger products should be compared to adjudicated messenger products). The Court need not resolve this dispute, however, because the DA1-A messenger products feature the same design-around as the DA1-A browser products. Thus, even if the Court determined that the DA1-A messenger products were not more than colorably different from the relevant adjudicated product, the Court would conclude that the DA1-A messenger products do not infringe the '647 patent for the same reasons as explained above.

2. DA1-B

Samsung's DA1-B products contain most of the same changes as the DA1-A products, except that detection occurs in the DA1-A products when a user lifts his or her finger up from the screen, whereas detection in DA1-B occurs when a user puts his or her finger down on the screen. Mowry 2017 Initial Rep. ¶¶ 76, 101; Jeffay Rep. ¶ 184. In other words, the DA1-B products contain fewer changes from the adjudicated products than do the DA1-A products because detection occurs in both the DA1-B products and the adjudicated Jelly Bean browser when a user puts his or her finger down on the screen. See Mowry 2017 Initial Rep. ¶ 101. As explained above, the DA1-A browser products are not more than colorably different from the adjudicated browser products because the changes are insignificant from a technical perspective, a functional perspective, and a user's perspective. See supra, Section III.C.1.a.i.2. As a result, the DA1-B products, which contain fewer changes than the DA1-A products, also are not more than colorably different from the adjudicated products for the same reasons. Thus, the Court must determine whether the DA1-B products infringe either literally or under the doctrine of equivalents.

To prove literal infringement, Apple must show that the user interface limitation of claim 1 is satisfied. Specifically, Apple must show that detection of a structure is complete before selection occurs. Apple argues that "the timing of the call to the detection functions in the DA1-B Products is the same as in the Adjudicated Products—in both cases, '[t]he process of detecting a selected structure starts upon finger down.'" Mot. at 15 (quoting Jeffay Dep. 116:7-16) (emphasis and alteration in original). Dr. Mowry states that because "[t]he detection function in DA1-B Products starts immediately upon 'finger down' and is called even before 'finger up' occurs," the "user is therefore selecting a detected structure." Mowry 2017 Initial Rep. ¶ 150. According to Dr. Mowry, "the timing of the detection functions in the DA1-B Products operates the same way in which the Infringing Galaxy S III [does]." Id.

However, at trial, Dr. Mowry explained that detection occurred before selection in the infringing Galaxy S III, which ran the Jelly Bean browser, because detection started and finished while the user was performing a long press. Trial Tr. 867:14-868:11. Under Dr. Mowry's theory, selection was not complete until the user finished the long press. Thus, because detection was complete before the long press was complete, detection occurred before selection. Trial Tr. 867:17-868:11; 868:23-870:8; Mowry 2017 Initial Rep. ¶¶ 42-43.

By contrast, Dr. Mowry does not offer an opinion in his reports on whether detection is complete in the DA1-B products before a short tap is complete. In his deposition, Dr. Mowry admitted that he had performed no empirical tests to determine how long detection took. Mowry Dep. 180:5-16. Dr. Mowry further testified that the detection routines "may or may not be" complete before a short tap is complete. Id. at 178:2-6; 178:17-20. As a result, under either the clear and convincing standard or the preponderance standard, Apple has failed to carry its burden of showing that detection occurs before selection in the DA1-B products. As a result, Apple has not carried its burden to show that the DA1-B products literally infringe.

Apple argues that even if detection is not complete before a short tap is complete, "then there is infringement under the doctrine of equivalents for the same reasons discussed above with respect to the DA1-A Products." Reply at 12. However, for the reasons that the Court explained above, a product in which selection occurs before detection is not equivalent to claim 1, which requires that detection occur before selection. As a result, to the extent that detection occurs after selection in the DA1-B products, the DA1-B products do not infringe under the doctrine of equivalents.

3. DA2

Apple argues that the newly accused DA2 products are not more than colorably different from the adjudicated products because the DA2 messenger products still (1) automatically detect and highlight structures; (2) allow users to select those structures with a short tap; and (3) display a pop-up menu upon that short tap. Mot. at 17. Apple concedes that the pop-up menu that is displayed is not the context menu, but is instead the resolver activity menu. Id. at 18. However, Apple argues that the resolver activity menu is not colorably different from the context menu, because the resolver activity menu "enables the user to choose from one or more actions that are linked to that structure." Id. at 20. Samsung responds that the removal of the context menu "should end the inquiry." Opp'n at 21. In the alternative, Samsung argues that under the Federal Circuit's decision in Ncube, a previously present but unaccused feature—here, the resolver activity menu—is not the proper focus of the colorably different inquiry. Id. at 23-24.

Samsung is correct that under Ncube, a previously present but unaccused feature cannot support a finding that a newly accused product is less than colorably different from an adjudicated product. In Ncube, the patent at issue disclosed technology that "allows a user to purchase videos that are then streamed to a device such as a television." 732 F.3d at 1347. One of the limitations of the relevant claim required "updating a connection service table with [an] upstream physical address." Id. at 1348. SeaChange's infringing product updated the connection table with a ClientID and a SessionID, both of which contain the same "MAC address" information. nCube Corp. v. SeaChange Int'l, Inc., No. 01-011-LPS, 2012 WL 4863049, at *3 (D. Del. Oct. 9, 2012), aff'd by Ncube, 732 F.3d 1346. At trial, the plaintiff's "expert identified the ClientID as the 'upstream physical address'" that satisfied the limitation that required "updating a connection service table with [an] upstream physical address." Id. at *2. "Notably, neither [the plaintiff], nor its expert, ever mentioned the SessionID as potentially being the upstream physical address." Ncube, 732 F.3d at 1348.

After trial, SeaChange modified its product so that the connection table was no longer updated with the ClientID. However, the connection table continued to be updated with the SessionID. 2012 WL 4863049 at *3. The plaintiff argued that the redesigned product was not more than colorably different from the infringing product because the connection table was still updated with the SessionID, which contained the same MAC address information as the accused ClientID. Id. The district court ruled the plaintiff at trial never identified the SessionID as satisfying the "upstream physical address" limitation. 2012 WL 4863049 at *5. "Nor, following trial, did SeaChange ever modify or replace the SessionID as part of its redesign efforts. The [c]ourt [wa]s not convinced that the SessionID performs any different function in the modified ITV system than it did in the infringing ITV system." Accordingly, the court did not find that the SessionID allegations could support a finding of less than colorable differences. Id.

On appeal, the plaintiff argued that the district court erred "by failing to compare the Connection Table of the infringing system, which is updated with both the ClientID and the SessionID, with the Connection Table of the modified system, which is updated only with the SessionID." 732 F.3d at 1350-51. "In short, [the plaintiff] argue[d] that the ClientID and SessionID each contain the same 6-byte MAC address information and, therefore, the Connection Table in both versions of the system is updated with the MAC address." Id. at 1351. Thus, the infringing system's connection table was updated with two MAC addresses, whereas the newly accused system's connection table was updated with only one MAC address. According to the plaintiff, this was not a colorable difference. Id.

The Federal Circuit rejected this argument. Id. Specifically, the Federal Circuit reasoned that "[i]n order for this argument to hold sway, the MAC address must be the portion of the ClientID that meets the upstream physical address limitation of claim 4. The problem is that [the plaintiff] never relied on the MAC address at trial to prove infringement." Id. "Rather, [the plaintiff] relied on the ClientID to prove infringement, but never called out the MAC address as the infringing aspect of that element." Id. As a result, the Federal Circuit affirmed.

Like the SessionID in Ncube, in the instant case the resolver activity menu was present in the adjudicated devices but Apple never "unequivocally alleged prior to the [ongoing royalty] stage" that the resolver activity menu "met th[e] [pop-up menu] claim limitation." TiVo, 646 F.3d at 883-84. Nor did Apple "rel[y] on the [resolver activity menu] at trial to prove infringement." Ncube, 732 F.3d at 1351. Although Apple argues that the resolver activity menu "has taken the place of the context[] pop-up menu," Mot. at 18, this is true only insofar as the resolver activity menu now sometimes appears as the result of the user's selection of a structure instead of the context menu. Apple does not specifically contend that the resolver activity menu provides the same functionality as the context menu. Nor could Apple so contend. Because the resolver activity menu coexisted with the context menu in the adjudicated devices, it is clear that the two menus performed different functions. Like the SessionID in Ncube, the resolver activity menu was not the subject of Samsung's redesign efforts, and there is no evidence that the resolver activity menu performs a different function in the newly accused devices than it did in the adjudicated devices. See nCube, 2012 WL 4863049 at *5. Thus, the Court concludes that the resolver activity menu provides a different function than the context menu. Accordingly, under the Federal Circuit's decision in Ncube, the relevant colorably different comparison is not between the context menu in the adjudicated products and the resolver activity menu in the newly accused products, which was previously present but unaccused.

Apple tries to avoid this result by arguing that "Apple was not required to accuse each successive instance of infringement by Samsung's devices; demonstrating infringement by the first pop-up menu in the Adjudicated Products was sufficient to prove Apple's case." Mot. at 19 n.7. As a result, Apple argues that it may now properly rely on the resolver activity menu to prove that the newly accused products are not more than colorably different from the adjudicated products even though Apple did not previously accuse the resolver activity menu. Id. However, the Federal Circuit rejected this very argument in Ncube. In Ncube, the plaintiff argued that "it need not prove every possible avenue of infringement at trial and, as such, it was under no obligation to prove that the SessionID also met the upstream physical address of the claim." 732 F.3d at 1351. The Federal Circuit held that although a plaintiff "need not prove every avenue of infringement," "the separation of the colorable-differences and infringement components in TiVo . . . preserves values of notice and preservation of trial rights by keeping contempt suitably limited." Id. Thus, the Federal Circuit held that the plaintiff's proof and arguments at trial limited the scope of the colorably different analysis. Id.

Here, because the context menu was completely removed in DA2 products, the relevant inquiry is whether that removal "was significant." See Proveris, 739 F.3d at 1371 (calling argument that removal of accused element renders a product more than colorably different as a matter of law "a misreading of TiVo" and stating that "the court must still determine whether that modification was significant"). In Proveris, the Federal Circuit held that the purported removal of an infringing feature was not significant where "a comparison of the User Manuals [of the infringing and newly accused products] demonstrates that the two products are functionally identical" and "it [wa]s not at all clear from the record whether [the] purported change actually had any effect." 739 F.3d at 1371. Here, by contrast, it is undisputed that the removal of the context menu from DA2 products altered the range of actions that a user is able to perform on a structure. The Court finds that this change—going from a choice of different types of actions in the adjudicated products' context menu to a default action—results in more than a colorable difference between the adjudicated products and the newly accused DA2 products. Cf. TASER Int'l, Inc. v. Stinger Sys., Inc., No. CV07-42-PHX-JAT, 2012 WL 12960852, at *5-6 (D. Ariz. Jan. 18, 2012) (finding design changes that altered device's capabilities, among other changes, rendered newly accused products more than colorably different from infringing products). Because the Court finds that there is more than a colorable difference between the DA2 products and the adjudicated products, the Court does not proceed to an infringement analysis. See TiVo, 646 F.3d at 882. While a jury may find that the resolver activity menu meets the pop-up menu limitation of claim 9, "that should not be decided in a[n] [ongoing royalty] proceeding." Id. at 884.

IV. CONCLUSION

For the foregoing reasons, the Court DENIES Apple's motion for ongoing royalties for the products incorporating the DA1-A, DA1-B, and DA2 design-arounds. The Court GRANTS Apple's motion for ongoing royalties as to the stipulated amount of $6,494,252 plus interest.

The parties shall file proposed judgment forms within seven days from the date of this order.

IT IS SO ORDERED.

Dated: February 15, 2018

/s/_________

LUCY H. KOH

United States District Judge


Summaries of

Apple, Inc. v. Samsung Electronics Co.

UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA SAN JOSE DIVISION
Feb 15, 2018
Case No. 12-CV-00630-LHK (N.D. Cal. Feb. 15, 2018)

applying test to defendant's products that were allegedly "designed around" the infringed patent

Summary of this case from Opticurrent, LLC v. Power Integrations, Inc.

applying the injunction standard on continued infringement by newly accused products to ongoing royalties

Summary of this case from Bridge v. TCL Commc'n Tech. Holdings Ltd.
Case details for

Apple, Inc. v. Samsung Electronics Co.

Case Details

Full title:APPLE INC., Plaintiff, v. SAMSUNG ELECTRONICS CO., LTD., et al.…

Court:UNITED STATES DISTRICT COURT NORTHERN DISTRICT OF CALIFORNIA SAN JOSE DIVISION

Date published: Feb 15, 2018

Citations

Case No. 12-CV-00630-LHK (N.D. Cal. Feb. 15, 2018)

Citing Cases

Opticurrent, LLC v. Power Integrations, Inc.

Some courts have also applied the same test to determine whether a prevailing plaintiff should be awarded…

Bridge v. TCL Commc'n Tech. Holdings Ltd.

Thus, the party seeking to enforce an injunction (or obtain ongoing royalties) must prove both that the newly…