Waymo LLC v. Uber Technologies, Inc. et al

Filing 2516

(PUBLIC REFILING OF #2497 pursuant to #2498 ) ORDER ON #2202 DAUBERT MOTION RE LAMBERTUS HESSELINK. (whalc2S, COURT STAFF) (Filed on 1/23/2018)

1
2
3
4
5

IN THE UNITED STATES DISTRICT COURT

6
FOR THE NORTHERN DISTRICT OF CALIFORNIA
7
8
9

WAYMO LLC,

11
For the Northern District of California

United States District Court

10

12
13
14

No. C 17-00939 WHA

Plaintiff,
v.

Defendants.
/

15
16

ORDER ON DAUBERT
MOTION RE LAMBERTUS
HESSELINK

UBER TECHNOLOGIES, INC.;
OTTOMOTTO LLC; and OTTO
TRUCKING LLC,

After two continuances of the trial date and in advance of the third final pretrial

17

conference in this action for trade secret misappropriation, this order GRANTS IN PART and

18

DENIES IN PART defendants’ Daubert motion to exclude plaintiff’s technical expert.

19

Defendants Uber Technologies, Inc., and Ottomotto LLC (collectively, “Uber”) move to

20

exclude plaintiff Waymo LLC’s technical expert Lambertus Hesselink from opining about

21

(1) Waymo’s saved development time estimates for alleged trade secret numbers 25, 90, and

22

111, and (2) Waymo’s “bottleneck” theory for all alleged trade secrets (Dkt. No. 2202).

23

With respect to saved development time estimates for alleged trade secret numbers 25,

24

90, and 111, the relevant paragraphs of Hesselink’s report are reproduced here in full (Dkt. No.

25

2201-4 (citations omitted)):

26
27
28

455. With regards to Trade Secret No. 25, I understand that
Waymo’s specific testing parameters and scenarios are derived
from test data accumulated from millions of miles driven by test
vehicles over at least a two year period. This evidence is
consistent with my expectation from reviewing the documentation

1
2
3
4
5
6
7
8
9

11
For the Northern District of California

United States District Court

10

12
13
14
15
16
17
18
19
20
21
22

relevant to Trade Secret No. 25, which involves a sophisticated set
of sensing parameters relating to a carefully selected set of
important “use cases” that are most likely to drive the overall
LiDAR design. It is my opinion that a similarly comprehensive set
of self-driving parameters could only be independently derived if
Defendants had undertaken the same years of testing and millions
of test miles driven. Thus, it is my opinion that Defendants’
misappropriation of Trade Secret No. 25 saved them at least 2
years of development time in their self-driving vehicle program[.]
456. With regards to Trade Secret No. 111, I understand that
Waymo and its predecessor 510 Systems devoted at least 1 year in
ruling out
LiDAR system for
medium-/long-range sensing. This is consistent with my
expectation in the field where, as discussed above, there is no
settled approach to the design of medium-/long-range LiDAR
sensor design, and I would expect a company independently
developing technology to have to eliminate certain design options
through the trial-and-error process. Thus, it is my opinion that
Defendants’ misappropriation of Trade Secret No. 111, and their
immediate pivot to exactly the “right” architecture for the Fuji
design, saved Defendants at least 1 year of development time in
their self-driving vehicle program.
457. With regards to Trade Secret No. 90, as discussed above,
Defendants’ [sic] accelerated their knowledge of fiber-laser
technology by acquiring Tyto LiDAR in the Spring of 2016. For
several years prior to this, Tyto LiDAR — at the direction of
Anthony Levandowski — exploited Waymo’s trade secret
information regarding fiber laser technology in order to create a
lower cost design for their “Owl” device. For example, at least by
November 2013, Tyto LiDAR had “defined a plan to reduce the
cost of the fiber laser and bring the BOM cost down to $9,500 by
January 2014.” Tyto’s continued work on lowering the cost of its
own, custom-built fiber laser from late 2013 until its acquisition by
Ottomotto in May 2016 further enhanced Tyto’s value to
Defendants. It is therefore my opinion that the 2 year, 5 month
period of time spent by Tyto exploiting Waymo’s Trade Secret No.
90 is a reasonable approximation of the time saved by Defendants
in not having to independently develop their own fiber laser
technology from the ground up.
As Uber points out, the foregoing portions of Hesselink’s report essentially amounted to

23

a black box. Hesselink simply recited fact evidence of how much time Waymo supposedly

24

spent developing alleged trade secret numbers 25 and 111, and how much time Tyto supposedly

25

spent exploiting alleged trade secret number 90, then concluded with no discernible analysis or

26

reasoning whatsoever that Uber saved at least that much time in its own development efforts.

27

To give just one non-exhaustive example, Hesselink made no attempt to explain why Waymo,

28

Uber, and Tyto — three very different companies with different resources, personnel, and
2

1

development plans — would necessarily take the same amount of time to develop the same

2

technology. His conclusions amounted to attorney argument, not expertise.*
Hesselink came closest to bridging the gap between his data and opinions in Paragraphs

3
4

455 and 456, where he injected conclusory assertions that the amount of time supposedly

5

invested by Waymo in alleged trade secret numbers 25 and 111 comported with his

6

expectations. Such bald assertions, however, constitute mere ipse dixit and fall well short of the

7

reliability required by Federal Rule of Evidence 702. See Daubert v. Merrell Dow Pharms.,

8

Inc., 509 U.S. 579, 589–95 (1993); Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997).

9

Because Hesselink failed to provide a reliable foundation for his opinions on saved
development time estimates for alleged trade secret numbers 25, 90, and 111, and because there

11
For the Northern District of California

United States District Court

10

was simply too great an analytical gap between the data he cited and the opinions he proffered,

12

those opinions will be excluded.
With respect to the “bottleneck” theory, the relevant paragraphs of Hesselink’s reply

13
14

report are reproduced here in full (Dkt. No. 2201-5 (citations omitted)):
282. Dr. Lebby opines that the proposed re-designs would not
result in a delay to Uber’s overall self-driving car effort. I
disagree. It is my opinion that the re-design times proposed by
Defendants will at least result in an equivalent schedule impact to
Defendants’ overall self-driving car program. In other words,
every day of re-design time proposed by Defendants will result in a
minimum of a day of delay in Defendants [sic] overall self-driving
car technology, and most likely many more as outlined above in
paragraph 281.

15
16
17
18
19

283. This opinion is based on the importance of the mediumrange LiDAR solution to the Defendants’ overall self-driving car
solution, the fact that any changes in the medium-range LiDAR
system will impact other aspects of Defendants [sic] self-driving
car technology, and the fact that certain aspects of self-driving car
technology depend on a final LiDAR design, including the
software needed to interpret the LiDAR sensor data. For example,
it would be my expectation that Uber’s proposed re-designs . . .
will, in turn, require reconfigurations to the perception software
that cannot be undertaken until the re-design work is completed.
Additionally, re-designing the mechanical design . . . will require
that the other hardware interfaces be similarly re-designed in order
to properly interact with the proposed re-designed components.

20
21
22
23
24
25
26
27
*

28

Incredibly, Hesselink’s curriculum vitae, in a section titled “legal expertise,” boasts, “Over the years
I have gained a reputation as a worldwide recognized expert on optics and holography. I have been an expert
witness in over 15 major case [sic] and I have never lost a case” (Dkt. No. 2239-2 at 15 (emphasis added)).

3

1

First, Uber complains that this opinion was not adequately disclosed in Hesselink’s

2

opening report. As the quoted portions show, however, Hesselink proffered the opinion in

3

response to a contrary opinion by Uber’s expert. Hesselink will therefore be allowed to present

4

his opinion at trial, but only on rebuttal after Lebby opines to the contrary. Second, Uber

5

contends Hesselink’s analysis was “conclusory, methodologically unsound, and unreliable.”

6

This order disagrees. Hesselink adequately explained his reasoning for why, due to the

7

interlocking nature of self-driving car development, re-design of LiDAR components would

8

have a ripple effect that would delay the overall development timeline. Uber’s arguments that

9

Hesselink’s reasoning was flawed and his conclusion was incorrect may have merit but go to
the weight of his testimony, not to its admissibility. Third, Uber contends “an optical engineer

11
For the Northern District of California

United States District Court

10

who has never worked on self-driving cars . . . is not qualified to offer such an opinion.” But

12

Hesselink’s “bottleneck” opinion did not exceed the scope of his optical engineering

13

qualifications and was not so specialized that only an expert who personally worked on self-

14

driving cars could render it. Again, Uber’s criticism goes to the weight of Hesselink’s

15

testimony, not to its admissibility.

16

As stated herein, Uber’s motion is GRANTED IN PART and DENIED IN PART. Hesselink

17

will be permitted to testify on rebuttal about his “bottleneck” theory but will not be permitted to

18

opine on saved development time estimates for alleged trade secret numbers 25, 90, and 111.

19
20

IT IS SO ORDERED.

21
22

Dated: January 18, 2018.

WILLIAM ALSUP
UNITED STATES DISTRICT JUDGE

23
24
25
26
27
28
4