Google LLC v. Sonos, Inc.

Filing 565

ORDER RE #464 #469 MOTIONS TO STRIKE. SIGNED BY JUDGE ALSUP. (whalc2, COURT STAFF) (Filed on 4/12/2023)

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 1 of 19

1
2
3
4
5
UNITED STATES DISTRICT COURT

6
7

NORTHERN DISTRICT OF CALIFORNIA

8
9
10

SONOS, INC.,
Plaintiff,

United States District Court
Northern District of California

11
12
13
14

No. C 20-06754 WHA

v.
GOOGLE LLC,
Defendant.

ORDER RE MOTIONS TO STRIKE
PORTIONS OF EXPERT REPORTS

15
16

INTRODUCTION

17

In this patent infringement action, both sides move to strike portions of each other’s

18

expert reports. For the following reasons, patent owner’s motion is GRANTED IN PART and

19

DENIED IN PART, and alleged infringer’s motion is DENIED.
STATEMENT

20
21

Sonos, Inc. accuses Google LLC of infringing U.S. Patent Nos. 10,779,033; 10,848,885;

22

and 10,469,966. These patents generally concern multi-room “smart” speaker technology.

23

Specifically, the ’033 patent covers technology related to transferring playback between

24

devices, i.e., “casting,” whereas the ’885 and ’966 patents cover technology related to

25

managing groups of smart speakers.

26

In the lead-up to trial, both parties have filed motions to strike portions of each other’s

27

expert reports (Dkt. Nos. 464, 469), as well as new motions for summary judgment (Dkt. Nos.

28

478, 483). Sonos also filed a renewed motion to realign the parties (Dkt. No. 477), which the

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 2 of 19

1

undersigned granted at the hearing after Google withdrew its opposition (Dkt. No. 557). This

2

order considers only the motions to strike.
1.

4

According to Sonos, Google improperly introduced new invalidity theories, non-

5
6

United States District Court
Northern District of California

SONOS’S MOTION TO STRIKE.

3

infringement theories, and non-infringing alternatives in its expert reports.
By way of background, Sonos explains that Google served invalidity contentions

7

pursuant to Patent Local Rule 3-3 on December 6, 2021, which it urged Google to amend to no

8

avail in March 2022 (Sonos Br. 3; Moss Exh. G). Separately, Sonos served interrogatories

9

asking Google to describe its non-infringing alternatives and non-infringement positions on

10

August 7, 2021 (Moss Exh. H at 16, 19–20). In the last week of fact discovery, Google served

11

supplemental responses to these interrogatories, describing its non-infringing alternatives and

12

non-infringement positions on November 21, 2022, and November 29, 2022, respectively

13

(Moss Exhs. I, J). Then, on the last day of fact discovery, November 30, 2022, Google served

14

the opening expert reports of Dr. Samrat Bhattacharjee and Dr. Dan Schonfeld (Moss Exhs. A,

15

D). Google served the experts’ rebuttal and reply reports on January 13, 2023, and January 23,

16

2023, respectively (Moss Exhs. B, C, E, F). Dr. Bhattacharjee’s reports address invalidity and

17

non-infringement of the ’033 patent, whereas Dr. Schonfeld’s reports address invalidity and

18

non-infringement of the ’885 and ’966 patents.

19

Sonos argues that “despite [its] discovery obligations and procedural rules designed to

20

promote disclosure and efficiency, Google waited until serving its expert reports to disclose a

21

broad array of new invalidity and non-infringement theories,” thereby “hid[ing] the ball until

22

fact discovery had closed” (Sonos Br. 2). Google purportedly failed to amend its invalidity

23

contentions, to disclose its new positions in its interrogatory responses, and to supplement its

24

disclosures as required under Federal Rule of Civil Procedure 26(e)(1). For that reason, Sonos

25

asserts that Google’s new theories are untimely and improper, and that they should be struck

26

from Dr. Bhattacharjee and Dr. Schonfeld’s reports.

27
28
2

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 3 of 19

2.

2

Meanwhile, according to Google, it was Sonos that improperly introduced new validity

3
4

United States District Court
Northern District of California

GOOGLE’S MOTION TO STRIKE.

1

and infringement theories in its expert reports.
Sonos served infringement contentions pursuant to Patent Local Rule 3-1 on October 21,

5

2021 (Google Br. 4). Google emphasizes that Sonos was granted leave to amend these

6

contentions several times, most recently on January 12, 2023 (Dkt. No. 447). (In some

7

instances, Google opposed amendment, whereas in others, including the most recent instance,

8

Google did not.) Sonos served the opening expert report of Dr. Douglas Schmidt on

9

infringement of the ’033 patent on November 30, 2022 (Kaplan Exh. 2). It served his rebuttal

10

report on validity and his reply report on infringement on January 13, 2023, and January 23,

11

2023, respectively (Kaplan Exhs. 3, 5).

12

Google argues that Sonos presented new validity and infringement theories in Dr.

13

Schmidt’s rebuttal and reply reports that are based on claim constructions first disclosed in his

14

rebuttal report and that are different than those provided by the Court. Google further argues

15

that Sonos presented new infringement theories in all of Dr. Schmidt’s reports that were not

16

disclosed in Sonos’s infringement contentions, and that Sonos presented new infringement

17

theories in Dr. Schmidt’s reply report that were not disclosed in his opening report (Google

18

Br. 5). Like Sonos, Google asserts that these new theories are therefore untimely and

19

improper, and that they should be struck from Dr. Schmidt’s reports.

20
21
22

The motions will be taken up in turn. This order follows full briefing and oral argument.
ANALYSIS
The patent local rules require parties to disclose infringement and invalidity contentions

23

early. See Patent L.R. 3-1, 3-3. As such, they may not use expert reports “to introduce new

24

infringement theories, new infringing instrumentalities, new invalidity theories, or new prior

25

art references not disclosed in the parties’ infringement contentions or invalidity contentions.”

26

See ASUS Comput. Int’l v. Round Rock Rsch., LLC, 2014 WL 1463609, at *1 (N.D. Cal. Apr.

27

11, 2014) (Judge Nathaniel M. Cousins). But the rules permit amendment of contentions as

28

new and previously unavailable information comes to light in discovery. See Patent L.R. 3-6;
3

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 4 of 19

1

O2 Micro Int’l Ltd. v. Monolithic Pwr. Sys., Inc., 467 F.3d 1355, 1365–66 (Fed. Cir. 2006).

2

And, they do “not require identification of every evidentiary item of proof.” Oracle Am., Inc.

3

v. Google Inc., 2011 WL 4479305, at *3 (N.D. Cal. Sept. 26, 2011) (emphases omitted).

4

Meanwhile, the Federal Rules of Civil Procedure require parties to supplement interrogatory

5

responses on pain of not being able to use new information moving forward unless the failure

6

to supplement is substantially justified or harmless. See Fed. R. Civ. P 26(e)(1), 37(c)(1).

United States District Court
Northern District of California

7

“The threshold question in deciding whether to strike an expert report” or, in this case,

8

select portions of an expert report, is “whether the expert has permissibly specified the

9

application of a disclosed theory or impermissibly substituted a new theory altogether.”

10

Digital Reg of Tex., LLC v. Adobe Sys. Inc., 2014 WL 1653131, at *5 (N.D. Cal. Apr. 24,

11

2014) (Judge Kandis A. Westmore). When the line between permissible application of a

12

disclosed theory and impermissible substitution of a new theory blurs, the district court

13

“revert[s] to a simple question: will striking the report result in not just a trial, but an overall

14

litigation, that is more fair, or less?” Apple Inc. v. Samsung Elecs. Co., 2012 WL 2499929,

15

at *1 (N.D. Cal. June 27, 2012) (Judge Paul S. Grewal).

16

This order begins with Sonos’s motion to strike and associated arguments.

17

1.

18

Starting with its first batch of arguments, Sonos avers that Google improperly introduced

19

previously undisclosed invalidity theories in both Dr. Bhattacharjee and Dr. Schonfeld’s expert

20

reports. Sonos seeks to strike material related to: (A) the Tungsten/NexusQ system’s “Magic

21

Playlist” feature; (B) the YouTube Remote system’s provision of “automatic playback”; and

22

(C) certain Bose products.

23

SONOS’S MOTION: NEW INVALIDITY THEORIES.

A.

TUNGSTEN/NEXUSQ SYSTEM’S “MAGIC PLAYLIST.”

24

First, according to Sonos, Google improperly asserted for the first time in Dr.

25

Bhattacharjee’s opening report that the “Magic Playlist” feature of the Tungsten/NexusQ

26

system anticipated claim 1 of the ’033 patent (Sonos Br. 6–8). Dr. Bhattacharjee’s opening

27

report states that “[i]n addition to allowing users to play user-created playlists, the prior art

28

Tungsten/NexusQ was also able to playback playlists that were generated and stored in the
4

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 5 of 19

1

cloud” (Moss Exh. A ¶ 218). This “allowed users to request that the cloud servers generate

2

and provide a playlist of songs that were related to, for instance, specific tracks or albums,

3

which was called a ‘Magic Playlist’” (Moss Exh. A ¶ 220). As such, “[t]he Magic Playlist

4

[was] a ‘remote playback queue’” per limitation 1.4 of the ’033 patent, which requires a

5

“computing device [] configured for playback of a remote playback queue provided by a cloud-

6

based computing system associated with a cloud-based media service” (Moss Exh. A ¶ 499;

7

’033 patent 17:39–42). Sonos decries that the Magic Playlist feature appeared nowhere in

8

Google’s invalidity contentions, which only disclosed the Tungsten/NexusQ system as prior art

9

(Sonos Br. 6).

United States District Court
Northern District of California

10

But there is no new theory of invalidity here. Admittedly, Google should have been

11

more forthcoming when it stated in its invalidity contentions that “the Tungsten/NexusQ

12

system anticipate[d] the asserted claims of the ’033 patent under at least the interpretation that

13

[Sonos] appears to rely on for its infringement theories” (Google Opp. 5) (internal quotation

14

marks omitted). But Google provided additional information in its invalidity contentions

15

sufficient to disclose its theory of invalidity. Note Google’s invalidity claim chart described

16

how the Tungsten/NexusQ system enabled users to “pull the music directly from the music

17

library in the cloud,” with a screenshot showing one user selecting a playlist to make

18

“[a]vailable offline”:

19
20
21
22
23
24
25
26
27
28

Item [4] from Invalidity Claim Chart for ’033 Patent.

(Moss Exh. K at 9–12).
5

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 6 of 19

United States District Court
Northern District of California

1

Sonos takes issue with the fact that Google did not expressly disclose a “playlist” when it

2

stated that the Tungsten/NexusQ system allowed users to pull music directly from the cloud;

3

after all, “pulling a single media file from a music library in the cloud would not require a

4

‘playlist’ at all, much less one ‘generated’ by a cloud server” (Sonos Reply Br. 2–3) (emphasis

5

in original). But playlists are clearly pictured in the screenshot above. That they could be

6

made “available offline” reflects that they were stored in the cloud. And, that the text directly

7

above the screenshot referred to “[v]arious forms of streaming media . . . be[ing] requested,”

8

such as “live radio program[s]” from services “includ[ing] YouTube and Pandora,” adequately

9

put Sonos on notice that these playlists could be generated by a cloud server (Moss Exh. K

10

at 11–12). Taken together, that was sufficient disclose Google’s theory of invalidity. This

11

order declines to strike material related to the Magic Playlist feature.

12
13

B.

YOUTUBE REMOTE SYSTEM’S “AUTOMATIC PLAYBACK.”

Second, according to Sonos, Google improperly asserted for the first time in Dr.

14

Bhattacharjee’s reply report that the YouTube Remote system anticipated claim 1 of the ’033

15

patent based on its provision of “automatic playback” of lists of videos (Sonos Br. 7–8). Dr.

16

Bhattacharjee’s reply report explains that “[a]lthough the ‘+’ icon c[ould] be used to add a

17

recommended video to the user’s queue, the list of recommended videos c[ould] also be played

18

back without adding them to the user’s queue” (Moss Exh. C ¶ 79). Dr. Bhattacharjee

19

annotates a screenshot from Google’s invalidity claim chart that “shows the YouTube Remote

20

prior art playing back a list of videos provided by the YouTube cloud servers in the ‘Search’

21

workspace — in other words, without using the ‘+’ icon to add the videos to the user’s queue”

22

(ibid.). Dr. Bhattacharjee suggests that the YouTube Remote system thereby allowed for

23

playback of a remote playback queue, not just playback of a local playback queue when a user

24

manually added videos using the “+” icon. The annotated screenshot from Dr. Bhattacharjee’s

25

report is reproduced below:

26
27
28
6

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 7 of 19

1
2
3
4
5
6
7
8
9
10

United States District Court
Northern District of California

11
Annotated Screenshot from Reply Report
of Dr. Bhattacharjee.

12
13
14

(ibid.).

15

This order agrees with Google that its invalidity contentions sufficiently disclosed the

16

underlying theory. Only Dr. Schmidt’s rebuttal report (that Dr. Bhattacharjee’s reply report

17

was directed to) limited Google’s theory to playback of a queue of videos a user expressly

18

added to the “Queue” workspace, elsewhere referred to as the “Queue” tab (see Google Opp. 5

19

n.3). Google’s invalidity claim chart provided screenshots of playback of queues of videos

20

from the “Search” and “Recommended” tabs in its discussion of limitation 1.4, like the one

21

above unannotated (Moss Exh. L at 7, 11). Further, any doubt about the disclosure (and

22

feasibility) of “automatic playback” from such tabs appears to be foreclosed upon review of

23

item [4] from that claim chart, a video of the YouTube Remote system that the parties and the

24

undersigned closely scrutinized in evaluation of other motions (Moss Exh. L).1 It shows

25

playback of a queue of videos from a “Best of YouTube” tab — not the “Queue” tab. Thus,

26

the theory that the queue of videos in the other tabs could have been remote playback queues

27
28

1

Citing https://www.youtube.com/watch?v=EGdsOslqG2s (last visited April 11, 2023).
7

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 8 of 19

1

— and that the YouTube Remote system could have been “operating in a first mode in which

2

the computing device is configured for playback of a remote playback queue provided by a

3

cloud-based computing system associated with a cloud-based media service,” per limitation 1.4

4

— was adequately disclosed, even if Sonos disagrees with it. This order declines to strike

5

material related to the YouTube Remote system’s provision of automatic playback.

6

United States District Court
Northern District of California

7

C.

BOSE PRODUCTS.

Third, Sonos avers that Google improperly relied upon a number of undisclosed Bose

8

products as prior art in Dr. Schonfeld’s reports (Sonos Br. 9–11). According to Sonos,

9

although Google disclosed the primary reference, the Bose Lifestyle 50 System, it did not

10

disclose the “Bose Link communication protocol” and the “Lifestyle SA-2 and SA-3 Stereo

11

Amplifiers” (Sonos Br. 9 (citing Moss Exh. D)). Sonos asserts that the primary reference did

12

not include those products, as demonstrated by the owner’s guide, among other documents

13

(Sonos Br. 9 (citing Moss Exhs. M, N, O, P, Q)). What’s more, according to Sonos, it would

14

be impossible for Dr. Schonfeld to show that these products were part of the primary reference

15

because the primary reference was discontinued before the amplifiers were even released

16

(Sonos Br. 9–10 (citing Moss Exhs. R, S)). As such, “there is little discernable connection

17

between the previously disclosed Bose Lifestyle 50 System, and the newly disclosed

18

communication protocol and amplifiers (other than a common manufacturer)” (Sonos Br. 10).

19

Accordingly, Google’s failure to disclose these products was “inherently prejudicial to Sonos”

20

(Sonos Reply Br. 5).

21

Not so. This order acknowledges that Sonos makes colorable arguments, but Sonos

22

should have made them roughly nine months ago. Google disclosed these Bose products on

23

June 6, 2022, when it served Dr. Schonfeld’s opening report on invalidity for the patent

24

showdown round of summary judgment motions (Kaplan Opp. Exh. 11 ¶¶ 619–722; see, e.g.,

25

¶¶ 653, 677). It served documents relating to these Bose products even earlier, on September

26

30, 2021 (Kaplan Opp. Decl. ¶ 3). As Google observes, Sonos seems to have sat on its hands

27

waiting to strike expert report language, just like it did during the patent showdown. Again,

28

Sonos “strategically chose not to initiate motion practice on this issue.” Google, 2022 WL
8

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 9 of 19

1

3052559, at *5. Google is correct that “[i]f Sonos believed that certain Bose-related theories

2

disclosed in Dr. Schonfeld’s report were missing from Google’s invalidity contentions, Sonos

3

could have raised the issue at any time thereafter,” including during the two depositions of Dr.

4

Schonfeld, one of which was taken after the motion to strike was filed (Google Opp. 9). In any

5

case, as Google observes, Sonos’s technical expert Dr. Kevin Almeroth discussed these Bose

6

products at length in both of his expert reports (Google Br. 10 (citing Kaplan Opp. Exhs. 15

7

¶¶ 215–250, 748–964; 16 ¶¶ 348–408, 1249–1557)).

8

United States District Court
Northern District of California

9

Like the order on Sonos’s prior motion to strike, this order will not reward Sonos’s delay
tactics, particularly when it is clear that Sonos was not actually prejudiced. See BLK

10

Enterprises, LLC v. Unix Packaging, Inc., 2018 WL 5993839, at *2 (C.D. Cal. Oct. 2, 2018)

11

(Judge Karen L. Stevenson). Because any non-disclosure was harmless, this order declines to

12

strike material related to the Bose products.
SONOS’S MOTION: NEW NON-INFRINGEMENT THEORIES.

13

2.

14

Turning to the next batch of arguments, Sonos contends that Google improperly

15

introduced previously undisclosed non-infringement theories in both Dr. Bhattacharjee and Dr.

16

Schonfeld’s rebuttal reports. Specifically, Sonos enumerates five purportedly new non-

17

infringement theories in Dr. Bhattacharjee’s rebuttal report and one purportedly new non-

18

infringement theory in Dr. Schonfeld’s rebuttal report. This order finds that none are, in fact,

19

new theories.

20

First, Sonos argues that Google stated for the first time in Dr. Bhattacharjee’s rebuttal

21

report “that limitation 1.7 of the ’033 Patent is not infringed because the setPlaylist message

22

sent by the Sender to the MDx server is not identical to the setPlaylist message sent by the

23

MDx server to the Receiver” (Sonos Br. 11). Suffice to say, under limitation 1.7, the “Sender”

24

is a control device and the “Receiver” is a playback device. In response to interrogatory 12,

25

Google expressly stated that: (1) “a YouTube application running on the alleged computing

26

device transmits a ‘setPlaylist’ message ‘to one or more MDx servers,’” (2) “[t]he MDx

27

servers then generate a further setPlaylist message,” and (3) “it is the MDx server, not the

28

alleged computing device, that transmits an instruction for the . . . playback device” (Moss
9

United States District Court
Northern District of California

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 10 of 19

1

Exh. J at 47–48). Sonos may disagree that the messages sent and received by the MDx server

2

are, in fact, different such that they fail to satisfy limitation 1.7, but this theory of non-

3

infringement was sufficiently disclosed.

4

Second, Sonos argues that Google stated for the first time in Dr. Bhattacharjee’s rebuttal

5

report “that limitation 1.7 of the ’033 Patent is not infringed because a videoID is not used by

6

Receiver [sic] to retrieve at least one media item in the remote playback queue because

7

(a) Bandaid URLs are used to retrieve rather than videoId and (b) a Receiver obtains only

8

chunks of a media item at a time rather than a whole media item” (Sonos Br. 12). No detailed

9

discussion of technical terms is warranted here. As is arguably betrayed by the sentence

10

structure, Dr. Bhattacharjee’s expert report merely provided additional explanation regarding

11

the theory of non-infringement that Google disclosed in various interrogatory responses.

12

Sonos is correct that “disclosing a technical explanation of how a videoId results in retrieval of

13

a media item is not the same as arguing that a videoId is not ‘use[d]’ by the playback device ‘to

14

retrieve’ a media item” (Sonos Reply Br. 8), but Sonos had what it needed to connect the dots

15

and conduct discovery on this theory of non-infringement — which it did, as Google points out

16

(Google Opp. 15 (citing Kaplan Opp. Exh. 8 at 85:3–94:11, 113:15–115:7)). In other words,

17

this theory of non-infringement was sufficiently disclosed.

18

Third, Sonos argues that Google stated for the first time in Dr. Bhattacharjee’s rebuttal

19

report that limitations 1.4 and 1.7 of the ’033 patent are not infringed because, in essence, a

20

computing device is not configured for playback of the remote playback queue prior to the

21

transfer of playback responsibility to a playback device (Sonos Br. 12). Google’s interrogatory

22

language sufficiently introduced its theory of non-infringement here, expressly stating that

23

“when playing back media on the alleged ‘computing device,’ [an] accused YouTube

24

application plays back a local queue stored on the computing device” such that “Sonos has

25

failed to show that [an] accused YouTube application infringes the ‘remote playback queue’

26

limitations that require playback of the remote playback queue on the computing device”

27

(Moss Exh. J at 60) (emphases added). Sonos’s argument appears to be motivated not by

28

prejudice or surprise but rather by its disagreement with Google that the accused products can
10

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 11 of 19

1

have both a local playback queue and a remote playback queue. To say that Sonos’s

2

contentions did not allow for this possibility, however, is a step too far. This theory of non-

3

infringement was sufficiently disclosed.2
Fourth, Sonos argues that Google stated for the first time in Dr. Bhattacharjee’s rebuttal

United States District Court
Northern District of California

4
5

report “that a Google Hub device is not a ‘computing device’ as claimed by claim 1 of the ’033

6

Patent” (Sonos Br. 12). But this was directly responsive to an infringement argument made in

7

Dr. Schmidt’s opening report about initiating playback on a Google Hub device and then

8

transferring playback to a second Google Hub device (see Google Br. 9 –11 (citing Kaplan

9

Exh. 2 ¶¶ 178, 247)). According to Google, it could not have disclosed this position earlier

10

because Sonos did not disclose this “new theory of infringement” earlier (Google Opp. 18).

11

For that reason, Google moves to strike Sonos’s infringement contentions related to the Google

12

Hub device, which Sonos opposes (Google Br. 9–11; Sonos Opp. 17–21). As will be discussed

13

below, this order disagrees that Sonos disclosed a new theory of infringement involving the

14

Google Hub device. In any event, because both sides’ experts spoke to the Google Hub

15

device’s alleged infringement, this order declines to strike related responsive material.
Fifth, Sonos argues that Google stated for the first time in Dr. Bhattacharjee’s rebuttal

16
17

report “that a [Google] Hub device running YT Main app does not satisfy limitations 1.5–1.6

18

of the ’033 Patent because it pauses playback when the Cast icon is selected” (Sonos Br. 12).

19

This order declines to strike such material involving the Google Hub device for the reasons

20

stated directly above.
Lastly, Sonos argues that Google stated for the first time in Dr. Schonfeld’s rebuttal

21
22

report “that the Accused Google Controllers do not ‘cause storage’ of zone scenes and

23

therefore do not infringe asserted independent claims 1 and 9 of the ’966 patent” (Sonos Br.

24

13) (emphasis omitted). According to Sonos, Google’s interrogatory responses only argued

25
26
27
28

Sonos’s related argument that Dr. Bhattacharjee newly and thereby improperly interpreted
remote playback queue in limitation 1.4 to refer to the same remote playback queue in limitation
1.7 is flatly rejected. As Google observes, this is not expert opinion but rather a well-settled
aspect of interpreting claim language (Google Opp. 16–17 (citing X One, Inc. v. Uber Tech. Inc.,
440 F. Supp. 3d 1019, 1034–35 (2020) (Judge Lucy H. Koh)).
11
2

United States District Court
Northern District of California

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 12 of 19

1

this “lack of storage” with respect to dependent claim 3, from which asserted claim 4 depends

2

(ibid.). But as explained by Google, just because it supplemented its interrogatory response to

3

provide additional disclosure regarding non-infringement of claim 3 does not mean that it

4

disclaimed its original response that all of the asserted independent claims did not meet the

5

“causing storage” limitations (Google Opp. 19 (citing Kaplan Opp. Exh. 18 at 39–42)).

6

Google incorporated its previous response by reference into its supplemental responses. As

7

such, Google’s theory of infringement was sufficiently disclosed.
SONOS’S MOTION: NEW NON-INFRINGING ALTERNATIVES.

8

3.

9

Turning to the final batch of arguments, Sonos asserts that Google improperly introduced

10

three previously undisclosed non-infringing alternatives in both Dr. Bhattacharjee and Dr.

11

Schonfeld’s rebuttal reports.

12

First, Sonos contends that Google stated for the first time in Dr. Bhattacharjee’s rebuttal

13

report that a non-infringing alternative existed at the time of infringement “consisting of the

14

‘receiver device not [] send[ing] the accused getWatchNext and/or getPlayer requests,’ with

15

the ‘receiver device’ instead ‘send[ing] a request to a Onesie agent’” (Sonos Br. 15 (quoting

16

Moss Exh. B ¶ 288)). But this was disclosed in, inter alia, Google’s response to interrogatory

17

18. In that response, Google explained that the playback device could “send a request to a

18

Onesie agent,” which would be tasked with obtaining media data by making “GetWatchNext”

19

and “GetPlayer” requests, and would then “stream the media data from the Onesie agent to the

20

receiver device” (Kaplan Opp. Exh. 24 at 13)). That Google later incorporated into its

21

response showdown opinions regarding the Onesie agent and the ’615 patent did not serve to

22

change or erase this disclosure regarding the Onesie agent and the ’033 patent (see Google

23

Opp. 21–22).

24

Next, Sonos contends that Google improperly expanded on what was disclosed as “NIA

25

#4” in Google’s interrogatory response and “Alternative #3” in Dr. Bhattacharjee’s rebuttal

26

report: continuing playback at a control device after playback has been transferred to a

27

playback device (Sonos Br. 15). According to Sonos, Alternative #3 contained three additional

28

alternatives to infringement (Sonos Br. 16). This order disagrees. That this alternative “would
12

United States District Court
Northern District of California

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 13 of 19

1

apply to ‘all of the YouTube applications,’ including YouTube Music” was never disclaimed;

2

rather, Google expressly disclosed that the alternative applied to “Google’s accused products,”

3

which would cover all of the YouTube applications (Sonos Br. 16 (emphasis in original);

4

Kaplan Opp. Exh. 24 at 14). Meanwhile, that “the alternative could be implemented so that all

5

of the YouTube applications default to pausing playback of the media . . . upon transfer” was

6

merely explanation of Google’s existing theory in response to language in Dr. Schmidt’s

7

opening expert report, which stated that an accused product is configured for playback of the

8

remote playback queue even when it is paused and not playing back media (Sonos Br. 16

9

(emphasis in original); Moss. Exh. B ¶ 280).

10

But this order agrees with Sonos that Google never disclosed this alternative “could also

11

be implemented so that the video and audio would continue to playback on the mobile device

12

after transferring playback” (Sonos Br. 16). Google suggests that its discussion of playback

13

continuing with “only the video alone (and the audio muted)” was “but one example of the

14

disclosed NIA” (Google Opp. 23). Yet there is no language in its interrogatory response —

15

e.g., “e.g.” — that suggests this was but one example (Kaplan Opp. Exh. 24 at 14). As such,

16

Sonos’s motion to strike as to the relevant language in paragraph 280 of Dr. Bhattacharjee’s

17

rebuttal report is GRANTED.

18

Finally, Sonos argues that Google improperly advanced a new non-infringing alternative

19

in Dr. Schonfeld’s rebuttal report, called “No Identification of Groups as Zone Scenes” (Sonos

20

Br. 17 (citing Moss Exh. E ¶¶ 178–81)). Note Dr. Schonfeld himself acknowledged that he did

21

not discuss this non-infringing alternative in his opening report (Moss Exh. E ¶ 178).

22

According to Google, this argument “elevates form over substance” because this non-

23

infringing alternative “follows directly from the Court’s discussion of naming a group” in the

24

prior order on the ’885 patent (Google Opp. 23–24). Even so, it is unclear why Google did not

25

incorporate this non-infringing alternative when it supplemented its interrogatory responses

26

after the patent showdown, and why it did not incorporate this non-infringing alternative in Dr.

27

Schonfeld’s opening report. This was insufficiently disclosed.

28
13

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 14 of 19

United States District Court
Northern District of California

1

Google points out that Dr. Almeroth discussed this non-infringing alternative in his reply

2

report (Google Opp. 24 (citing Kaplan Opp. Exh. 23 ¶¶ 268–77)). But this discussion was

3

sparse, and understandably so. Dr. Almeroth expressly “note[d] that Dr. Schonfeld ha[d] not

4

provided sufficient details regarding this new alleged non-infringing alternative, and as a

5

result, [Dr. Almeroth did] not have enough information to fully evaluate whether this alleged

6

non-infringing alternative would be non-infringing, technologically viable, available to

7

Google, or commercially acceptable” (Kaplan Opp. Exh. 23 ¶ 268). It cannot be said that this

8

nondisclosure was substantially justified or harmless. See Fed. R. Civ. P 37(c)(1). Sonos’s

9

motion to strike as to paragraphs 178–81 from Dr. Schonfeld’s rebuttal report is GRANTED.
GOOGLE’S MOTION: NEW CLAIM CONSTRUCTION.

10

4.

11

This order now turns to Google’s motion to strike and associated arguments.

12

First, Google argues that Sonos introduced a new claim construction in Dr. Schmidt’s

13

rebuttal report (Google Br. 5–7). During the patent showdown, the Court construed the term

14

“playback queue” in claim 13 of the ’615 patent as “a list of multimedia content selected for

15

playback.” Google, 2022 WL 3046752, at *5. According to Google, Dr. Schmidt’s rebuttal

16

report “proposed a different claim construction” because it suggested that a playback queue is

17

only a playback queue if it possesses “additional purported requirements” (Google Br. 5–6).

18
19
20
21
22
23
24
25
26

Dr. Schmidt’s rebuttal report states, in pertinent part:
Thus, according to the Court’s Order, I understand that the claim term “playback
queue” refers to a “list of multimedia content selected for playback” with the
following characteristics:
•
•
•
•

The playback queue is the list of media items that is used for playback;
The playback queue contains the entire list of media items selected for
playback;
The playback queue is not being used merely to process the list of media
items for playback;
The playback queue is the queue that “runs the show.”

(Kaplan Exh. 3 ¶ 107).
This order does not read Dr. Schmidt’s language as “propos[ing] a different claim

27

construction” but rather as applying it to rebut contentions in Dr. Bhattacharjee’s reports.

28

Sonos’s expert can fairly provide further details about an earlier opinion to rebut the responsive
14

United States District Court
Northern District of California

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 15 of 19

1

opinions of another expert. See MasterObjects, Inc. v. Meta Platforms, Inc., 2022 WL

2

4856269, at *4 (N.D. Cal. Oct. 3, 2022). Further, this order credits Dr. Schmidt’s deposition

3

testimony that he did not expressly articulate these characteristics in the same way in previous

4

reports because here he was responding to what he perceived as flawed new theories raised by

5

Dr. Bhattacharjee (see Richter Exh. 4 at 143:18-144:14). According to Dr. Schmidt’s rebuttal

6

report, Dr. Bhattacharjee had previously represented that a local playback queue and a remote

7

playback queue were mutually exclusive, and his latest reports took a different position

8

(Richter Exh. 2 ¶¶ 301–05). The bulleted characteristics were therefore included to support Dr.

9

Schmidt’s argument that the relevant prior art can only have one playback queue and —

10

because the prior order indicated that the local playback queue was this playback queue —

11

there was no remote playback queue in the prior art. The merits of this argument will be taken

12

up elsewhere. Google may disagree with Sonos’s application of the Court’s claim

13

construction, but this does not warrant striking it from Dr. Schmidt’s rebuttal report.
GOOGLE’S MOTION: NEW INFRINGEMENT THEORIES.

14

5.

15

Google makes two primary arguments regarding infringement theories that were

16
17

allegedly first introduced in Sonos’s expert reports.
First, Google argues that Sonos’s infringement theories based on the “PlaylistService”

18

and “BigTable” should be struck (Google Br. 7–9). Specifically, Google asserts that Sonos’s

19

infringement contentions did not disclose that the accused remote playback queue is stored on

20

a cloud server. It points to language in Sonos’s infringement contentions that identifies the

21

remote playback queue as “a ‘Watch Next’ queue[] provided by one or more cloud servers”

22

(Google Br. 7 (quoting Kaplan Exh. 1 at 4)) (emphasis added). In other words, according to

23

Google, the Watch Next queue is a “list of media identifiers contained in the Watch Next

24

message that is sent to a YouTube application, not a queue that is stored in the cloud” (Google

25

Reply Br. 6). As such, disclosure of where and how the Watch Next queue is stored in Dr.

26

Schmidt’s opening and reply reports — “by the ‘PlaylistService’ in a distributed storage

27

system referred to as ‘BigTable’” — was purportedly improper because Google was not put on

28

notice of the underlying theories of infringement (Google Br. 8). Sonos responds that Google
15

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 16 of 19

1

“has long understood that Sonos accuses a playback queue in the cloud” rather than a message

2

providing the media item identifiers from the cloud, and that Google had reasonable notice

3

about the relevance of PlaylistService (Sonos Opp. 9). This order agrees.

United States District Court
Northern District of California

4

Sonos points to ample statements and documentation demonstrating Google knew that

5

Sonos was accusing a queue stored in the cloud (Sonos Opp. 9–11). This includes Dr.

6

Bhattacharjee’s opening report, filed the same day as Dr. Schmidt’s opening report (Richter

7

Exh. 7). In his opening report, Dr. Bhattacharjee explained that Sonos “contends that the term

8

‘remote playback queue’ covers a ‘playback queue’ that is ‘in the cloud’ (e.g., a cloud queue),”

9

and that Sonos’s infringement contentions “identify[] the ‘remote playback queue’ as ‘a

10

“Watch Next” queue’ with ‘recommended videos’ that is provided by ‘one or more cloud

11

servers’” (Richter Exh. 7 ¶ 87)).

12

Moreover, Sonos points to ample statements and documentation demonstrating

13

PlaylistService was sufficiently disclosed (Sonos Opp. 11–13). This includes a reproduced

14

excerpt from a Google document that explains how “[t]he PlaylistDocumentService provides a

15

representation of the queue” and “[i]t calls into the Playlist Service for video IDs in the

16

playlist” (Sonos Opp. 11 (quoting Richter Exhs. 5, 15)). Sonos even provided a source code

17

trace with citations for the code responsible for calls to the PlaylistDocumentService and

18

PlaylistService (Sonos Opp. 11 (citing Richter Exh. 5 at 24)). This is more than adequate.

19

In reply, Google notes that Sonos has provided “shifting (and inconsistent) infringement

20

theories regarding the term ‘remote playback queue’ throughout this litigation” and “these

21

subsequent inconsistent positions do not change Sonos’s contentions” (Google Reply Br. 7–8).

22

But Sonos amended its infringement contentions with leave of the Court and should not be

23

faulted for doing so. True, Sonos did not provide a source code trace for PlaylistService

24

specifically, but “[t]hat a particular document or source code file was not cited in a party’s

25

infringement disclosures does not automatically preclude the party from using that document

26

or file to support a theory that was timely disclosed.” See Oracle, 2011 WL 4479305, *3

27

(emphasis in original). Sonos was not required to identify every evidentiary item of proof —

28
16

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 17 of 19

1

or specify that the overarching distributed storage system happened to be called BigTable. The

2

theory that Sonos accused a playback queue stored in the cloud was timely disclosed.

United States District Court
Northern District of California

3

Second, as alluded to above, Google argues that Sonos’s infringement theory for Google

4

Hub devices should be struck (Google Br. 9–11). According to Google, Sonos’s Hub-based

5

infringement contentions were limited to a “single playback path” involving a user initiating

6

active playback at the Google Hub device via input at the Google Hub device’s touchscreen

7

display (Google Reply Br. 12; see also Google Br. 10). Meanwhile, Dr. Schmidt’s report

8

introduces two new playback paths, the first involving a user initiating active playback at the

9

Google Hub device via casting from a smartphone to the Google Hub device, and the second

10

involving a user initiating active playback at the Google Hub device using voice (Google Br.

11

9–10). Sonos responds that Google has interpreted the limitation 1.4 of the ’033 patent to

12

require the computing device to be actively playing back to be satisfied, which is why it spoke

13

to these playback paths at all (Sonos Opp. 17). But in any event, according to Sonos, its

14

contentions did not provide for only one playback path. Again, this order agrees.

15

Nothing in Sonos’s infringement contentions limited Sonos’s theory of infringement to a

16

single “playback path” (Google’s term). And surely it was not a secret to Google that playback

17

on its devices could be initiated via casting and voice commands. Sonos’s contentions cite to

18

one of Google’s own support pages, which explained that “[w]ith YouTube built-in to your

19

Google Nest display,” “you can use voice commands like ‘Play [name of video]’” (Sonos Opp.

20

18 (citing Richter Exh. 5 at 8)). Here, no new theory of infringement was disclosed. This was

21

simply additional explanation invited by Google’s interpretation of active playback.

22

Accordingly, this order declines to strike material related to “PlaylistService,” “BigTable,” and

23

Google Hub devices.
GOOGLE’S MOTION: NEW DOCTRINE OF EQUIVALENTS THEORIES.

24

6.

25

Finally, Google argues that Dr. Schmidt introduced two new doctrine of equivalents

26

theories in his reply report related to select limitations of the ’033 patent (Google Br. 11–12).

27

According to Google, Dr. Schmidt improperly argues that these limitations are met under the

28

doctrine of equivalents (1) “irrespective of whether the contents of the list [i.e., the specific
17

United States District Court
Northern District of California

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 18 of 19

1

songs or videos in the list] provided by the YouTube cloud infrastructure is the same or

2

different before and after Casting”; and (2) “irrespective of whether the list of one or more

3

media items selected for playback is stored in the same or a different location in the YouTube

4

cloud infrastructure before and after Casting” (Google Br. 11 (quoting Kaplan Exh. 5 ¶¶ 179–

5

88)). Google asserts that neither of these arguments was included in Sonos’s infringement

6

contentions and there is no justification for Sonos to have waited until Dr. Schmidt’s reply

7

report to present them (Google Br. 11–12). Sonos responds that Dr. Schmidt’s doctrine of

8

equivalents opinions are directly responsive to non-infringement positions improperly

9

advanced for the first time in Dr. Bhattacharjee’s non-infringement report — namely, that

10

Sonos has failed to show that devices are not configured for playback of remote playback

11

queues (Sonos Opp. 21–22). According to Sonos, these are the same arguments that Sonos

12

challenged in its motion to strike. As such, “[i]f the Court strikes these new non-infringement

13

positions from Dr. Bhattacharjee’s report, then Dr. Schmidt need not offer his doctrine of

14

equivalents positions” (Sonos Opp. 22). Because this order did not strike those non-

15

infringement positions, however, it declines to strike this responsive material.
CONCLUSION

16
17

For the foregoing reasons, Sonos’s motion to strike is GRANTED IN PART and DENIED IN

18

PART. This order strikes part of paragraph 280 of Dr. Bhattacharjee’s rebuttal expert report

19

and paragraphs 178–81 of Dr. Schonfeld’s rebuttal expert report. Google’s motion to strike is

20

DENIED in its entirety.

21

On these motions alone, counsel submitted 135 pages of briefing, 1,130 pages of

22

unsealed exhibits, and 3,756 pages of conditionally sealed exhibits and accoutrements — all to

23

strike, in the end, fewer than five paragraphs from two expert reports. (Separately, on the

24

summary judgment motions, counsel submitted 164 pages of briefing, 270 pages of unsealed

25

exhibits, and 1,286 pages of conditionally sealed exhibits and accoutrements, plus 322 slides

26

for the hearing.)

27
28

Alas, much ink was spilled for little purpose. Most “new theories” proved to be
permissible applications of disclosed theories or elaborations of evidence relevant to proof of
18

Case 3:20-cv-06754-WHA Document 565 Filed 04/12/23 Page 19 of 19

1

disclosed theories. It appears counsel moved to strike not out of prejudice but to secure an

2

advantage for summary judgment and trial, emblematic of the worst of patent litigation.

3

IT IS SO ORDERED.

4
5

Dated: April 12, 2023.

6
WILLIAM ALSUP
UNITED STATES DISTRICT JUDGE

7
8
9
10

United States District Court
Northern District of California

11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
19