One Night in November, Part 5: GPS Isn’t Always Reliable

Driver’s Vehicle was equipped with a device that allowed its position to be tracked using Global Positioning System information.  Pedestrian wanted to use those data as evidence of Vehicle’s speed in the moments before the collision.  I argued the GPS data were unreliable.  Impossible you say as you rely on Google to navigate you around town?

  • Vehicle carried a GPS tracking device, but the speed data were unreliable.

Pedestrian subpoenaed the company that provided the GPS service and obtained data from the morning the accident occurred.  Pedestrian argued these data indicated Vehicle was traveling at a rate in excess of the speed limit in the moments before the accident.[1]   As Pedestrian was the proponent seeking to admit these data into evidence, he bore the burden of demonstrating admissibility.[2]  To lay that foundation, Pedestrian deposed a NRCP 30(b)(6) designee of the company that provided the device and tracking service.

The designee was the founder and CEO.  He gave a general description of how the system operates and the data that are collected, such as “latitude, longitude, speed, heading, device ID, and time.”  When Vehicle was on and moving more than 7 kilometers per hour, data are gathered and transmitted every ten seconds.  He was provided the spreadsheet that the parties received and was asked what they data included.

13        Q.   So the information here, can you tell us
14   what it includes?
15        A.   It includes the device name, the time,
16   the miles per hour of the device, the heading.
17   Yep, that’s one thing I left out also of the data
18   stream is the number of satellites that the device
19   is reading from.  It also includes switches,
20   ignition status, latitude, longitude and address.
21   I also left out the ignition status and the data
22   stream.

Pedestrian also asked the designee if he had any questions about the speed calculation’s accuracy.

10        Q.   This is something that you do as part of
11   your business on a regular basis?
12        A.   It’s — yes.
13        Q.   And do you have any reason to doubt the
14   accuracy of the information in this report?
15        A.   No, none whatsoever.
16        Q.   The speeds indicated on this report, any
17   reason to believe they are over or understated?
18        A.   No.
19        Q.   Should be spot on, fair to say?
20        A.   They should be fairly accurate.

15        Q.   The server that the information is stored
16   on, is it secure?
17        A.   Yes.
18        Q.   Any reason to think or any reason to
19   suspect that this data could have been fabricated
20   or modified in some way by an outside party?
21        A.   Impossible.[3]

The designee may have been confident in the speed data, he could not testify to the methodology or algorithm utilized to create it. I spent some time going back and forth with him about how the system worked and then got to this segment:

14        Q.   So again, going back to my question,
15   then, the triangulation tells me where I am but
16   how — maybe this is because I’m not a computer
17   scientist — how does determining where you are
18   determine your velocity at that given moment?
19        A.   So as far as the exact explanation of how
20   that works, I am not an engineer.  I don’t think I
21   could give you the correct response to that.  But I
22   do know that the device makes the calculation based
23   on the data it receives, based on satellite
24   position and timing, and then it transmits that
25   velocity to our servers which we then are able to
1   see the data.
2        Q.   I appreciate that clarification.
3             Who would we need to speak to, then, if
4   we wanted to try to get an answer to that specific
5   type of question?
6        A.   It would probably have to be an engineer
7   at the device manufacturer’s headquarters.
8        Q.   Do you happen to know who the device
9   manufacturer was for this particular unit?
10        A.   Yes, I do.  It’s — the name of the
11   device manufacturer is a company called [name].

Stated another way, the foundational testimony needed to establish that these speed calculations are reliable would come from someone else.  Pedestrian never disclosed a witness or documentation from that other company to establish the foundation for how the speed data were calculated and whether the speed data gathered were accurate.

  • The foundation is needed because of known errors in the data.

Some smarty pants may think this was just wishful arguing and GPS data are reliable otherwise most of us would never find our way anywhere.  But the designee confessed there are known errors in the data.

19        Q.   What type of errors have you encountered
20   that affect the velocity calculations?
21        A.   I’ve seen, for example, what’s known as a
22   GPS bounce occur.  And that happens when, say, a
23   device is parked in a covering where the signal is
24   blocked and you have a very low satellite count,
25   maybe two or three satellites.  Well, the device
1   thinks that it’s not at the position that it is at
2   and it is at another location and it will return
3   erroneous data.
4        Q.   Have you seen other errors that report
5   erroneous velocity?
6        A.   I have seen it, yes.
7        Q.   Can you give us some examples of things
8   that you’ve seen?
9        A.   A parked vehicle, for example, in a
10   garage reporting a speed of 100 miles an hour.  And
11   that would be a classic example of the signal
12   bouncing.

These types of errors were consistent with errors that my clients had previously encountered and described in deposition.

  • What happened?

A motion was filed; but never decided.

[1] The speed limit in the area was 50.
[2] Pfingston v. Ronan Eng’g Co., 284 F.3d 999, 1004 (9th Cir. 2002) (“Because Pfingston seeks admission of Rizvi’s statements, he bears the burden of proof of admissibility.”)
[3] Everything can be hacked, it just takes time, patience, and inevitable human error.

One thought on “One Night in November, Part 5: GPS Isn’t Always Reliable

Comments are closed.