|
Hugh Roddis
Pre-Trial Evidentiary Hearings
February 17, 2004
Direct Examination by Rick Distaso
DISTASO:
Mr. Roddis, where are you employed?
RODDIS:
Company called Orion Electronics Limited.
DISTASO:
What is your title there?
RODDIS:
I'm the president of the company.
DISTASO:
And just very briefly, what type of company is Orion Electronics?
RODDIS:
We design and manufacture electronic equipment.
DISTASO:
And does that include covert global positioning system devices?
RODDIS:
Yes.
DISTASO:
Could you briefly state for the record your educational background?
RODDIS:
I have a degree in electrical engineering from the University of London back in
1965. I was certified as a Professional Engineer in Nova Scotia in 1974.
DISTASO:
Let me stop you there. Have you been a Professional Engineer, Licensed
Professional Engineer in Canada since 1974?
RODDIS:
Yes.
DISTASO:
Okay.
JUDGE:
Excuse me, Mr. Distaso. That's got a lot of bass gain. Can everybody understand
what the witness is testifying to? Because it sounds like,
DISTASO:
It reverberates quite a bit back there.
DISTASO:
And Orion Electronics, is that a company that you founded?
RODDIS:
Yes.
DISTASO:
When did you found the company?
RODDIS:
January of 1965.
DISTASO:
Do you sit on any scientific advisory councils in Canada?
RODDIS:
Yes. I'm a member of the Premier's Advisory Counsel on Innovation. I was past
member of the Council of Applied Science and Technology, which is a government
advisory board. And I'm presently a member of NSERC, which is the National
Science Engineering Research Council. Advisory Board on University Grant
Applications for Electromagnetics in Canada as a whole.
DISTASO:
And has, you told us Orion builds GPS tracking systems. How long has Orion been
building those type of systems?
RODDIS:
Roughly since 1992, roughly.
DISTASO:
Could you state briefly for the Court your employment history?
RODDIS:
Yes. I moved to Canada in 1965, after I graduated from university. I was
employed there as a design engineer for military equipment. Designed radio
location systems for aircraft, and sonobuoys. That's underwater acoustic
location systems. Worked for Litton Systems, for Martin Marrietta, people like
that, until I started Orion.
DISTASO:
And you have obviously been at Orion ever since 1975?
RODDIS:
Yes. Judge Delucchi 65. You said 65.
DISTASO:
I'm sorry, right.
RODDIS:
Pretty much. It was a period when the business wasn't doing very well and I
worked as an outside consultant. Interesting period. When people answered the
phone and said can I speak to the president, and the president doesn't work here
any more. But that ended in 1991. For a period of for years, I did outside
consulting for military equipment.
DISTASO:
All right. And have you previously testified as an expert in courts in the
United States and Canada regarding covert GPS technology?
RODDIS:
I have.
DISTASO:
How many times have you testified in courts in the United States?
RODDIS:
Twice previously.
DISTASO:
And what, just tell me the locations of those, of that testimony?
RODDIS:
Once was in Las Vegas, one was in Augusta, Georgia.
DISTASO:
I believe the Las Vegas was a state court; is that right?
RODDIS:
I believe so, yes.
DISTASO:
And the one in Augusta, Georgia, was a Federal Circuit Court?
RODDIS:
Yes.
DISTASO:
And have you also testified in Canada regarding this technology?
RODDIS:
Yes, I have.
DISTASO:
How many times?
RODDIS:
Just once.
DISTASO:
Was that in a federal court in Canada, or a state court? I don't know what the
system is there.
RODDIS:
It would be, would have been the equivalent to a Federal Court.
DISTASO:
Your Honor, I'm going to offer Mr. Roddis as an expert in covert GPS technology.
GERAGOS:
Can I take him on voir dire?
JUDGE:
Sure.
GERAGOS:
Thank you.
Voir Dire Examination by Mark Geragos
GERAGOS:
Good morning, Mr. Roddis.
RODDIS:
Good morning.
GERAGOS:
The first, what was the first location that you testified regarding GPS?
RODDIS:
That would have been the one in Canada.
GERAGOS:
Okay. And what year was that?
RODDIS:
Forgive me. I'm not sure. It would have, it would have been, I think, 2001.
Three years ago.
GERAGOS:
What was the issue that you were called to testify on?
RODDIS:
Yes, some other equipment had been used, and there was some question about the
use of the equipment, much as in this particular case.
GERAGOS:
Was part of reason that you were called was there had been some perceived
malfunction of the equipment, and you were there to explain what the malfunction
of the equipment was?
RODDIS:
In this particular case, the equipment had been used very oddly. The person had,
the investigator had, in fact, put the unit in his vest pocket. And they have,
there was some, I wouldn't say malfunction, but the record was erratic due to
the fact that his pocket was obscuring the satellite signals.
GERAGOS:
And was the object for you to try to explain why the data was erratic?
RODDIS:
Yes.
GERAGOS:
That's why you were called to testify, not to confirm what the data was, so to
speak, but to explain why the data didn't make sense?
RODDIS:
Exactly.
GERAGOS:
And in that case, was that a, by the way, was that a criminal case, if you were
aware?
RODDIS:
It was a murder case.
GERAGOS:
Okay.
RODDIS:
It was an appeal to a murder case, actually.
GERAGOS:
Okay. And this was on appeal, somebody had a post-conviction proceeding, and you
were called to testify?
RODDIS:
Exactly.
GERAGOS:
Now, in this post-conviction proceeding in which you were called to testify,
were you actually tracking, or was the device, by the way, was this an Orion
device?
RODDIS:
Yes.
GERAGOS:
Was the Orion device monitored in any way by your company when it was in use?
JUDGE:
Mr. Geragos, can I interrupt? Doesn't go to expertise or anything. Going into
the subject of his testimony. Maybe save that information for cross examination.
GERAGOS:
I wanted to lay, at least get a feel for the thought. If the Court would want,
as well, what his previous acts of testimony, Judge Deluchhi: I do. You are
going into detail about what his testimony was. I'm concerned he is being
offered as an expert in cross examination, goes to expertise in this particular
field. If it's necessary to go into this, that's fine. But,
GERAGOS:
I think to some degree is to establish that previous occasions were different
issues.
JUDGE:
Within reason, okay.
GERAGOS:
Was that information, is that testimony, did that testimony in Canada involve
tracking of someone?
RODDIS:
Yes.
GERAGOS:
And was it tracking of the person who actually had the receiver in their pocket?
RODDIS:
Exactly.
GERAGOS:
Okay. And was the receiver that was put in the pocket covertly placed there?
RODDIS:
I suppose, well, the user obviously knew it was there.
GERAGOS:
So it was not covert then, was it? User knew he had a tracking device in his
pocket?
RODDIS:
In that definition of covert, yes.
GERAGOS:
Now, the Las Vegas case, can you tell me what kind of a case that was?
RODDIS:
Robbery with violence. Mark Geragos" Okay. And what year did you testify?
RODDIS:
Two years ago, I think.
GERAGOS:
Okay. And in that case, what was the subject of your testimony?
RODDIS:
I suppose to explain to the Court how the unit was working, and I guess just
explain how the technology worked.
GERAGOS:
Wasn't there also an issue in that case about some data, and the data being
erratic?
RODDIS:
No.
GERAGOS:
Was there any data that you testified about in that case?
RODDIS:
I'm not sure how you describe that. I didn't testify to the data per se. I
testified to the fact that the tracks, the way the equipment worked, and how the
tracks were developed.
GERAGOS:
And was, were you testifying as to Orion data, or the Orion machine receiver?
RODDIS:
Yes. Orion machine.
GERAGOS:
Now, the third place that you say you testified was Augusta, Georgia. What year
was that?
RODDIS:
Last year.
GERAGOS:
And who called you as a witness in that case?
RODDIS:
In that case, it was the prosecution.
GERAGOS:
And in that case, was, what was the subject matter?
RODDIS:
It would have been the same as the previous one. Yes, same as the previous one,
inasmuch as the Court went to the description of how the equipment worked, and
so on.
GERAGOS:
Did you bring with you today any court orders that resulted from your testimony
at a hearing, from either Las Vegas or Augusta Georgia?
RODDIS:
To be truthful, after I testified I don't really know what happens after that,
without actually contacting somebody. I don't have any idea.
GERAGOS:
All you know, you went there and you testified about what you thought the
technology does, and left, and you don't have any other information after that?
RODDIS:
I heard verbally that in, I think in all of the cases that the accused was found
to be guilty. But I know nothing more than that. That's just a verbal thing.
GERAGOS:
Now, the company you formed was formed in 1975; is that correct?
JUDGE:
65.
GERAGOS:
I think he's given a CV that I have got that says formed Orion in 75. Is that
correct?
JUDGE:
I think he testified in 65, didn't you? 65 or 75?
RODDIS:
I'm sorry, no, I formed it in 75. 65 was when I graduated from university. I'm
sorry if that was confusing.
JUDGE:
So you are correct, Mr. Geragos. 1975.
GERAGOS:
You did not become licensed as a Professional Engineer until 1974; is that
correct?
RODDIS:
True.
GERAGOS:
The design engineer, you worked as a design engineer for Litton Industries,
Martin Marrietta?
RODDIS:
Right.
GERAGOS:
And that was having to do with submarine systems and low level air defense
radar?
RODDIS:
Yes. I worked, another company I worked at during that time called, at that time
was called E.M.I.-Cossor. E-M-I hyphen C-o-s-s-o-r. It's now called Hermes. It's
changed. And they worked in the electronics world.
GERAGOS:
Now, the, your present occupation, what do you do right now?
RODDIS:
I'm president of the company. Most of it is, I also acted, up until very, very
recently, as Chief Engineer as well. But now I just oversee what's going on.
GERAGOS:
And as president of the company, you have recently sold Orion?
RODDIS:
I did, yes.
GERAGOS:
And you have stayed on as an, in your present capacity as president?
RODDIS:
Yes, I have. I have two years to do. And I have gone through just over one of
those years.
GERAGOS:
And is one of the terms of the sale of Orion that you maintain some kind of
stock or stock option?
RODDIS:
No. It's just a cash sale. It's done.
GERAGOS:
And for the two years that you are on the payroll, so to speak, are you in
charge of everything in terms of creating a profit, and profit motive, for the
company?
RODDIS:
Up until, in fact, up until now, yes. Although they have got somebody else to
take over; and, in fact, while I'm away, they are moving my office furniture to
an another office, they tell me. I'm losing that delight.
GERAGOS:
You still are president. That's still your title?
RODDIS:
Yes. For another two months.
GERAGOS:
Now, the, I think Mr. Distaso asked you since when you have been building GPS
systems, and you answered 1991?
RODDIS:
Yes.
GERAGOS:
And the systems that you are building since 1991, are those high value, or low
cost, or both?
RODDIS:
Generally speaking high cost. We built oil spill tracking devices using GPS for
some time. They were perhaps lower cost. Most of what we do is high value.
GERAGOS:
And as an, are you familiar with the Informer ST600?
RODDIS:
Yes.
GERAGOS:
Is that something that's produced or sold by Orion?
RODDIS:
Yes. It's a series of 601, 602, 603, 604.
GERAGOS:
Were you involved in the manufacture of, or design of, that particular unit?
RODDIS:
Yes.
GERAGOS:
Is that yes?
RODDIS:
Yes.
GERAGOS:
What year was that?
RODDIS:
What year? I'm sorry, I'm not prepared for that one. It would have been , the
first 600 units would have been probably five years ago.
GERAGOS:
Okay. Thank you. I have no further questions on voir dire. As to the expertise,
there is an objection. I don't believe that there is a satisfactory showing.
JUDGE:
All right. The objection is overruled. The Court will accept Mr. Roddis as an
expert in GPS technology. Go ahead.
Direct Examination by Rick Distaso, resumed
DISTASO:
Mr. Roddis, were you contacted by the Modesto Police Department to review some
specific tracks regarding the use of one of your devices?
RODDIS:
I was.
DISTASO:
Let me show you this binder in front of you. It's marked People's 1. If you
could just flip through it, and just take a moment and flip through it, and then
let me know if you have, if those are the tracks that you have reviewed in this
particular case.
RODDIS:
Excuse me. I'm sorry. Taking a little longer.
DISTASO:
Go ahead.
JUDGE:
Take your time, Mr. Roddis.
RODDIS:
Yes.
DISTASO:
Those particular tracks, do you have an opinion as to whether the, actually, let
me start over.
JUDGE:
If I can interrupt. When you say these particular, are you referring to what's
marked 1-A through Q, and so forth? Are you referring to those?
DISTASO:
I am, your Honor. And I'll reference to them specifically. I think it would be
easiest.
JUDGE:
So I can keep track. Mr. Geragos would be much obliged if you could do that.
DISTASO:
Mr. Roddis, regarding People's 1-A, it's a track on 1-8 and 1-9 of 2003. Do you
have an opinion as to whether or not, as to the, whether the two specific sets
of data in that track are accurate?
RODDIS:
It certainly looks so to me, yes.
DISTASO:
And the instances we're talking,
JUDGE:
Excuse me. That's 1-A?
DISTASO:
1-A.
JUDGE:
1-A.
DISTASO:
And the data that there was a problem with, where there was a one degree of
offset on 1-8, and let me show you that. One degree of offset on 1-8, it's 1-QQ.
Do you see that portion of the track?
RODDIS:
I do.
DISTASO:
Do you have, do you have an explanation for the Court as to why that one degree
of offset occurred?
RODDIS:
Yes, I do.
DISTASO:
Go ahead.
RODDIS:
Yes. I'm just trying to, I'm just trying to get it straight in my head to
explain it.
DISTASO:
No problem.
RODDIS:
Within our units, we have two sort of systems, well, three, actually, but two
systems right here. We have the GPS receiver, which I believe the Court knows is
manufactured by Trimble Navigation. And we have our part of the unit, which
takes the information out of the Trimble unit, stores it, processes it, and
handles the other operations that go on. In this particular instance, the
Trimble unit had a very, had a firmware problem in it which only occurred when
the unit crossed over from one line of latitude or longitude on the zero,
exactly 61 degrees. Say, for example, when it crossed over from 60.9999 to 61,
just momentarily, the output would be wrong. Our unit interpreted that as one
degree offset error. And so the data, in itself, was, if you like, correct
except for the fact that the,
GERAGOS:
I'm sorry, the data was?
JUDGE:
Correct except for the fact,
RODDIS:
Except for the fact that that one degree would be, the degrees section would be
one degree larger than it should be.
DISTASO:
Let me stop.
JUDGE:
I can I interrupt for just a second? You said a firmware problem? Is that what
you said?
RODDIS:
Firmware.
JUDGE:
What is that, just so I know?
RODDIS:
What I mean there, is the computer program that the Trimble receiver uses to
operate within itself. Firmware, I mean the computer code, computer program
that's imbedded inside the Trimble unit, and which we have no control over.
JUDGE:
You have no control over?
RODDIS:
Right. It does its thing, and we take the output from it.
DISTASO:
And, Mr. Roddis, this problem that occurred, it occurred for six minutes; is
that correct, on this particular track?
RODDIS:
Yes, I believe so. I'd have to check that. Yes.
DISTASO:
Go ahead.
RODDIS:
Yes, exactly.
DISTASO:
And then, and at the end of the six-minute time, the unit corrected itself, and
this one degree of offset was corrected?
RODDIS:
Yes. I have to go into a bit more explanation about that one.
DISTASO:
Go ahead.
RODDIS:
Why it took six minutes to correct. Because, in essence, I don't want to get too
complex here. But there are two or three, in fact, ways of taking the data or
information out of our unit. You can do it directly. In other words, once the
unit is stored, or this information, all this information about the vehicle's
movement, you can take the unit away from the vehicle, connect it into a
computer, and have the data drawn on a map. You can also connect the unit over a
wireless radio link, like a cell phone, if you like, and download the data
remotely, so you don't actually have to go to the car and collect the unit. And
the third way is that if you like, you can actually connect to the unit, again
over the wireless link, and just watch the movement of the vehicle in what we
call real time. And, in that case, the data for that is also stored on the base
station computer.
DISTASO:
Okay.
RODDIS:
Now, in this case I understand that interpreting in this was done in
over-the-air download. And an over-the-air download, in order to speed the data
up, because the wireless network is quite slow, in terms of transmitting data,
in order to speed that process so we're not sitting down on the phone for hours
downloading, we compress the data. And we compress the data in a very simple
way. The parts of the data that don't change very often, for instance, the date,
we might download a thousand records on the same day sometimes. Simply to send
the date with every record, it doesn't change. And the same for some of the
other things. One of the things is the number of degrees, and the other the
hour. So if, as in a case like this, we get the one-degree offset, that will
maintain with the record until one of those major things rolls over, at which
point we send all of the data again to make sure the record is correct. If you
look at this one, the last of the bad records, or bad, I would say, one-degree
offset records, occurred at 5:59:55. So the next record would have been at six.
It would have rolled over an hour. And that caused the system to reset, to go
back to its correct operation, with the correct,
DISTASO:
When it hit the 6:00 o'clock hour mark, the unit corrected itself and it started
interpreting data correctly?
RODDIS:
Yes.
DISTASO:
Regarding this instance, if the unit had been this only, this compression issue
only occurs on over-the-air downloads; is that right?
RODDIS:
In this unit, yes. 600, that's true.
DISTASO:
Okay. And let me go to another record and have you just talk about it briefly.
In People's 1, let's see 1-JJJ, there is another one-degree offset in the middle
of that page. Do you see that?
RODDIS:
Yes. Goes back to here, I think.
DISTASO:
Let me show you 1-NNN. Again there is another one-degree offset. Was the problem
in this particular track, this one degree offset similar or the same problem
that you just described to us?
RODDIS:
Yes. The same problem. It's the one-degree issue, yes.
DISTASO:
Okay. And then let me also go to 1-A. Do you see the track in 1-A? This track in
front of you?
RODDIS:
Yes.
DISTASO:
There is a small spike kind of portion in Fresno. Have you reviewed the
information that corresponds to that little spike?
RODDIS:
Yes, I have.
DISTASO:
Let me show you 1-EE.
RODDIS:
Yes.
DISTASO:
And this particular spike, Doctor Loomis previously had testified that this
behavior was most likely caused by interference with the GPS receiver. And do
you agree with that, with his assessment there?
RODDIS:
In general, yes, I do. By the fact that when I say, in general, is that I have
no other explanation for it. Looks like something that could be caused by
interference.
DISTASO:
And that particular behavior took about five minutes to correct, and, do you
agree with that? That roughly, or so, in five minutes that behavior was
corrected?
RODDIS:
Yes.
DISTASO:
Other than these three problems which we've identified, the remainder of the
information in People's 1, the tracks of the Chevy S-10, the Rover and the Dodge
Dakota, did you find the rest of the information contained in those tracks to be
accurate and valid?
RODDIS:
I would. Yes, I would.
DISTASO:
Okay. Did you feel, let me just ask you a couple of questions about the Orion,
Orion software. Let me show you People's 1, I think that's DD. Just make sure.
1-DD. The data that's outlined on that record of 1-DD, does that, do you
recognize that format of the data?
RODDIS:
Yes.
DISTASO:
And is that how Orion reports the data?
RODDIS:
Yes. That's a typical printout from Orion software.
JUDGE:
Did you say 1-DD?
DISTASO:
1-DD. If you look at 9:51, does that represent an actual GPS fix?
RODDIS:
That's where we would record the identifier.
DISTASO:
And then there is the date?
RODDIS:
The date, yes.
DISTASO:
Then moving across the 1-DD from the left to the right, you have the time?
RODDIS:
Yeah.
DISTASO:
And then what is the minus 8.0 GPS fix? What does mean?
RODDIS:
Actually linked. There is an "L" there. The "L" stands for local time, because
the system itself measures in GMT, Greenwich Mean Time. The minus 8 is the
offset that was used to calculate the local time.
DISTASO:
And then after that is some numbers. Can you describe for the Court what those
numbers are?
RODDIS:
Yes. This is the latitude and longitude.
DISTASO:
Okay. Oh. So, all right. So then, actually, the, looking at this record 1-DD
looks like a bad print, because, would you agree with that? Because after that
you have miles per hour. Looks like it's merging into,
RODDIS:
It is, yes. Printout is not quite right.
DISTASO:
Printout on 1-DD is not correct.
RODDIS:
But if it wasn't, it would be the speed now, and then the motion status.
DISTASO:
Let me show you 1-G. And we can finish talking about that. This is a similar
data set, correct?
RODDIS:
Yeah.
DISTASO:
And on 1-G there is a miles per hour; is that right?
RODDIS:
Yeah.
DISTASO:
And how does the unit calculate miles per hour?
RODDIS:
This is not an absolute calculation. It's an indicator for the user. So what
it's done, this is just calculated from the change in latitude, latitude and
longitude. It's not measured on the vehicle. It doesn't come from the Trimble
receiver. This is just a calculation based on the change of position. So it goes
from A to B. We calculated he must have done 20 miles an hour to go from A to,
DISTASO:
The speed in this device is not calculated to the speedometer?
RODDIS:
No.
DISTASO:
And if, for example, like the one degree of offset, if the unit says that I am
here in Modesto, but I've one degree of offset which shows me sixty miles away,
it's going to report a very high speed to try to calculate, make up for that?
RODDIS:
Just it will be a ballistic missile, or something, yes.
DISTASO:
And so this particular speed, I think we have gone over that. And then let's
talk about now the motion-no-motion action, like in this one on 1-G is a good
example. It says 56 miles per hour, but the unit is showing no motion. How can
that be?
RODDIS:
Again, you must understand this equipment is not designed for issuing speeding
tickets, or anything. This is a tool for use by investigators. And that
motion-no-motion action is not derived from the GPS at all.
DISTASO:
Let me stop you there.
RODDIS:
Totally unrelated to that.
DISTASO:
The motion-no-motion sensor of the unit is not linked to the GPS portion of the
unit?
RODDIS:
No.
DISTASO:
Completely separate?
RODDIS:
It's a vibration sensor in the unit itself.
DISTASO:
Okay. So if the unit is parked, and someone is, you know, bouncing up and down
on the bumper, or something, the unit might actually register motion?
RODDIS:
Exactly.
DISTASO:
In that case would the GPS unit change the position of the vehicle?
RODDIS:
No.
DISTASO:
And would the speed show anything different?
RODDIS:
No.
DISTASO:
So let me give you another example then. So if someone is driving down a very
smooth highway, and the unit is covertly placed in a position where it is not
getting very many or no vibration, does the motion sensor, is it possible to
have a situation where the vehicle is moving, registering, speed but registering
no motion?
GERAGOS:
Objection. Speculation as phrased.
JUDGE:
I think he can give an opinion because he's an expert. Overruled. Can you give
an opinion, Mr., I keep elevating you to doctor.
RODDIS:
I don't mind.
JUDGE:
Mr. Roddis, I'm sorry, Mr. Loomis is the doctor. Mr. Roddis.
RODDIS:
Yes, it is. It's what we call the Cadillac-on-a-freeway effect. If you have a
very smooth ride, and perhaps the engine is not pushing too hard, the vibration
sensor responds a lot better to an engine driving hard, or somebody changing
lanes a lot, or whatever. It's actually a useful tool. Sometimes it's, you can
tell how the vehicle has been driven.
DISTASO:
If this, so just, finally, the motion sensor portion of the unit is not related
to the actual GPS fix locations of the unit?
RODDIS:
No, it's not. And in actual fact, you may even see a place where the vehicle is
stationary. And it shows no motion, but the speed has some speed, small speed,
like the one-mile-an-hour, two-mile-an-hour. That's simply because the GPS fix
is changing slightly. Just the statistical error in the measurement. And the
unit calculates very small velocity, because the unit, I don't know if, for
example, if we look at one of these things here, sorry, excuse me.
DISTASO:
Take your time.
RODDIS:
There is a place in here where, according to what I interpreted anyway, the unit
was stationary quite long periods to the street address. And yet if you look at
it, it isn't one dot on the map. Here is 1-J, is kind of a blurry thing. That's
just statistical variation of the fix as it sits there for a long, long time.
Just builds up a blob. We used to call them star pattern in the old days when it
was quite obviously it was bad. Nowadays it's lot more accurate and the thing is
a relatively small blob. When the unit sees that the variation of the dots in
there may cause the calculation to show some small motion, like one mile an
hour, two miles an hour, even though, in fact, the unit isn't moving.
DISTASO:
Okay. So like in 1-J, if a unit is parked, you have a series of blots or dots,
that's just the unit continually reacquiring its position?
RODDIS:
That's right. Yes, it does. It does it every second, in essence. It's kind of
been there a long time.
DISTASO:
Thank you, Mr. Roddis. I have no more questions.
JUDGE:
I just have one question, Mr. Roddis. Based upon your examination of People's
Number 1, do you have an opinion as to whether or not the correct scientific
procedures were used in this case?
GERAGOS:
Objection. There is no foundation.
JUDGE:
Overruled.
DISTASO:
Mr. Roddis, just so you know, People's 1 is the entire binder with all the
tracks. Entire binder with all the tracks.
JUDGE:
The question was, is it correct, do you have an opinion as to whether or not
correct scientific procedures were used in this case?
RODDIS:
I would believe so, yes.
JUDGE:
All right.
DISTASO:
I have nothing further.
Cross Examination by Mark Geragos
GERAGOS:
What do you base that belief on?
RODDIS:
What? I'm sorry.
GERAGOS:
What do you base that belief on?
RODDIS:
Oh, that's an interesting question. The track, as a whole,
GERAGOS:
Let me ask you one question first.
DISTASO:
Objection, your Honor. The witness wasn't finished with his answer.
JUDGE:
Mr. Geragos, let him finish.
GERAGOS:
Were you finished?
JUDGE:
Were you done with your answer, Mr. Roddis?
RODDIS:
The track as a whole, because you see a lot of these tracks. When I have been
developing this equipment, for days on end I would go out and drive around, do
things and then come back and look at the tracks. There is certain small things
that, sort of small things that happened. You get used to seeing what the track
looks like. So, for example, sometimes it might be slightly off the road, or
might jump reading, or so on. This track has all the characteristics of what I
would call a decent track. It's not something that's generated out of thin air.
It's something that has characteristics of a GPS track as I know it.
GERAGOS:
Well, is that the only information that you were given regarding this unit and
the use of the unit?
RODDIS:
How do you mean? We were asked to look at these tracks and analyze it, but
that's all.
GERAGOS:
Did they tell you, for instance, that the, one of the Orion brand GPS data
logging systems was installed on a vehicle?
RODDIS:
Yeah.
GERAGOS:
And the tracking device was installed, and that the vehicle was released, and
they were unable to maintain a track, or retrieve data information, and they
said it was due to equipment problems, and that that lasted for a period of time
from January 3rd to January 21st. Were you aware of that?
RODDIS:
When I analyzed the data initially, no. But when I went back and looked at the
serial numbers of the units, it turns out that that was the case in our records,
yes.
GERAGOS:
Okay. Where are your records that reflect that? Are they here?
RODDIS:
I don't have them here.
GERAGOS:
Where are they?
RODDIS:
In Technical Support records at the home base.
GERAGOS:
Okay. Now, you are looking at something as we're talking; is that correct?
RODDIS:
Yeah.
GERAGOS:
Now, is the item that you are looking at a January 13th, 2004, letter?
RODDIS:
Yes.
GERAGOS:
I think that's previously been marked, your Honor, as pretrial Exhibit C.
JUDGE:
Let me look. Yes.
GERAGOS:
Can I ask you,
JUDGE:
Well, it's, C is the, refers to telephone conversation with Orion. B is the
letter from Peach and supervisor.
DISTASO:
C actually refers to the telephone conversation, but is also a letter as well.
JUDGE:
That's the letter.
RODDIS:
It says July 9th. Refers to the prior one.
GERAGOS:
Correct. If I could, rather than having you testify, put those in front of you.
Can I hand you the two exhibits that have already been marked?
RODDIS:
Yes.
GERAGOS:
That's C, and the other one that you have in front of you is a July 9th letter
which was marked as Pretrial Hearing Exhibit Number, or letter B.
RODDIS:
Yeah.
GERAGOS:
Now, you don't have the data sets here that show that from January 3rd to
January 24th, the January 3rd to the 21st, that the device, one of these devices
was inoperable; isn't that correct?
RODDIS:
I'm sorry. You are asking if I have that information here in front me?
GERAGOS:
Yes. Do you have that information here?
RODDIS:
I have that information in my head, but not in the paper, no.
GERAGOS:
So in your head is January 3rd to the 21st, the tracking device that was
installed on a vehicle that was released to Mr. Peterson was inoperable; is that
correct?
RODDIS:
Yes. One of the two tracking devices that I have information on was inoperable,
that's correct.
GERAGOS:
Did you determine why the unit was inoperable?
RODDIS:
Yeah.
GERAGOS:
What was the reason the unit was inoperable?
RODDIS:
Well, there two issues as I understand it from our records. The first one was,
GERAGOS:
Could I ask you a question? Are you reading from a document as you speak? Just
looking down?
RODDIS:
I'll put it away. I'm not reading from the document, yes.
GERAGOS:
You say, as you understand it, are you testifying based upon your knowledge, or
on something somebody else told you?
RODDIS:
I based, my knowledge is based on looking at our records at Orion.
GERAGOS:
Okay. And looking at your records at Orion, did you,
DISTASO:
Objection, your Honor. The witness, can the witness please be allowed to finish
his answer?
JUDGE:
I think this is an answer to the question. Let it,
GERAGOS:
Looking at the records that were, that you have at Orion, what was your, what's
your explanation as to why the unit did not either track or provide accurate
data sets for a period of three weeks?
RODDIS:
As I started to say, there were two reasons. The first one was, initially, that
there was a problem with the cell phone connection. One thing that happens from
time to time, cell provider had inadvertently set up the connection wrongly, or
something. After that was resolved, there was still a problem with the unit. And
so the unit was retrieved and sent back to Orion. We looked at it. We looked to
the data in the unit, and we discovered that the one of the components, a
microprocessor, had a bad connection between one of its pins on this computer
chip and the rest of the circuitry that had caused the malfunction of the unit.
GERAGOS:
Well, the unit did, in fact, show the vehicle to be in the City of Modesto, and
you were given information that the vehicle was not, in fact, in the City of
Modesto; is that correct?
RODDIS:
I don't understand that question. I don't think I was given that information.
GERAGOS:
There was some limited tracks that were produced or generated by this unit that
you say malfunctioned, correct?
RODDIS:
Yeah.
GERAGOS:
And the limited tracks that were produced showed the vehicle to be in the City
of Modesto; is that correct?
RODDIS:
I don't know. I saw some tracks that were briefly shown to me that were
downloaded from the unit. I'm not sure where they were. I didn't look at the
city or place. I just looked at the form and function of the track and the data
to see if I felt the data was valid or not.
GERAGOS:
Well, didn't it concern you, when the judge asked you whether correct scientific
procedures were used in this case, wouldn't it concern you if, take a look at
the data and see where the data is indicating the unit is, or the receiver is,
and then determine whether or not there was surveillance or other things, other
factors that would have shown whether the receiver was actually working
properly.
RODDIS:
Forgive me. I thought I was testifying to the information in this book number
one, and it doesn't have any information from, that was downloaded from the bad
unit.
GERAGOS:
Right. And what I'm asking you about is, there were two Orion units that were
used; isn't that correct?
RODDIS:
Yes.
GERAGOS:
One of them malfunctioned. Wasn't a total failure, correct?
RODDIS:
No.
GERAGOS:
Well, would you use any of the information on that one?
RODDIS:
Yes. One of the, some of the information in this book came from that unit, but
it was before it was malfunctioned.
GERAGOS:
Can you show me data on there that shows that it was actually operating before
it malfunctioned?
RODDIS:
It comes from the file Rover 03 Jan 26 1418 Skultety, this one in here.
GERAGOS:
Find that. You are pointing to 1 double X? Are you aware that 1 double D starts
on January 24th, by that time the receiver had already been installed and
withdrawn, and it had been determined that it was malfunctioning, so you are
not, do you want to change the previous statement you have made, that in terms
of the units, data being there,
JUDGE:
That's two questions.
GERAGOS:
I'll just ask you one. Again, January 3rd through January 21st, January 3rd
through January 21st were the operative dates when the unit was installed. You
have just shown me 1 double X, which shows January 24th. That's not the unit, is
it?
RODDIS:
How do you mean it's not the unit?
GERAGOS:
Well, the unit you said was returned to your, to Orion, wasn't it, after it was
removed from the car?
RODDIS:
I'm sorry. This is serial number 7476, which is the unit I'm told was the one
that was wrong.
GERAGOS:
Who told you that?
RODDIS:
I got it from the information that Orion, and from this letter. We were told
that this is a track, and that's the serial number. So I looked up that serial
number to see the records.
GERAGOS:
What I'm telling you is, I'll show you the reports if you want to see them , is
that the unit that was installed on the automobile, the tracking device was
installed on January 3rd and was removed the week of January 21st. Okay? So can
we accept that for one second, that information?
RODDIS:
Sure.
GERAGOS:
Now, that unit was then sent back to you, correct? Because you are able to
determine that there was what you called a pin error, correct?
RODDIS:
I'm not sure of the date. Actually, yes, that unit was sent back.
GERAGOS:
Let me just read you another sentence from the police report that says, during
the week of January 21st, 03, he returned the Land Rover to the Modesto Police
Department to replace of the headliner that was partially removed during the
search for evidence. Do you understand, do you see that?
RODDIS:
I do see that, yes.
GERAGOS:
Do you see where, on January 3rd, they arranged for installation of a tracking
device into the vehicle?
RODDIS:
Yeah. Yes.
GERAGOS:
Do you see that was released to Mr. Peterson?
RODDIS:
Yes.
GERAGOS:
That was released, they were unable to maintain a track or retrieve data track
information?
RODDIS:
Yes.
GERAGOS:
That was due to equipment problems with the tracking unit?
RODDIS:
Yup.
GERAGOS:
You got all that? And it says the tracking data shows the vehicle within the
City of Modesto, and is very sporadic. Do you see all that?
RODDIS:
I see all of that.
GERAGOS:
Can you show me one item in that exhibit, that notebook, that reflects that
unit's sporadic information from January 3rd to January 21st of the year 2003?
RODDIS:
This record that I have here says it's from the 24th.
GERAGOS:
Exactly. And by the 24th, that unit was somewhere in Canada, and somebody was
determining what the problem was with the unit, so this could not be the same
information from that unit that was disabled, could it?
DISTASO:
Objection, your Honor.
RODDIS:
I'm not sure.
JUDGE:
Hold on just one second. Everybody is talking. First of all, what's the basis
for your objection?
DISTASO:
That's misstating the testimony. Mr. Roddis testified he doesn't know when the
unit was sent back. And I'll make an offer to the Court, there is going to be a
witness going to testify the unit wasn't sent back the second time until April.
JUDGE:
I think he's entitled to ask that question, see if Mr. Roddis, can you read the
question back?
GERAGOS:
I'll reask it. Tell me, when did you get this information as to the unit having a pin
problem as you described it?
RODDIS:
When did I personally get the information?
JUDGE:
I think he's asking,
GERAGOS:
Exactly.
RODDIS:
I'm not sure. I talked to the Technical Support people before I left to come
here.
GERAGOS:
Is one of the Technical Support people a Mr. Peach?
RODDIS:
Yes.
GERAGOS:
Now, Mr. Peach was involved in the technical support analyzation of what the
problems were in this case; isn't that correct?
RODDIS:
Yeah.
GERAGOS:
Okay. Now, can you tell me, where in that Exhibit 1, can you show me anything in
there that reflects what data sets were being generated by that unit between
January 3rd and January 21st?
RODDIS:
I'd have to go back and look; but I don't think this particular track looked at,
the one that's on the Rover, I assume that's the one you are talking about. The
other ones were generated by different units. This one starts on the 24th, which
is after the period that you are talking about. I'm not sure of the question.
GERAGOS:
Well, let me ask you something. What unit is is it that you are looking at right
now on 1 double X?
RODDIS:
What unit? 7476. Serial number 7476.
GERAGOS:
Do you have any way of determining from 7476 when it was installed?
RODDIS:
No.
GERAGOS:
Okay. Do you have any idea of, any way to determine whether or not it's been, by
the way, can you calibrate this unit in any way?
RODDIS:
No.
GERAGOS:
Do you have any, do you have any instructions on how to install the unit and
activate the unit?
RODDIS:
We do training courses, yes.
GERAGOS:
Okay. What training courses were given to Modesto Police Department in regards
to the Informer ST600 series?
RODDIS:
I'm not so sure I understand. One of the technicians, I wasn't aware that I
would be asked that question. I understand one of the technicians attended one
of our training courses. But I'm not sure about that for sure. That's just my
understanding.
GERAGOS:
Okay. Wouldn't you want to know that before you answered the judge's questions
as to whether or not the approach here was scientific in this particular case?
Wouldn't you want to know that information?
RODDIS:
No. Because I understood and interpreted the question, I may have been wrong. I
understood I was to look at this data to see whether or not this data, in my
opinion, reflected a proper use of the equipment. And, in my opinion, it did.
GERAGOS:
Well,
JUDGE:
Can I, just for just a second. Kelly-Frye we're doing here doesn't try to
ascribe to professionalism by which data was used. That's a different issue. The
question is, was the proper scientific methodology used. When it was used,
professional or not, would be a question for the jury to decide.
RODDIS:
Yes.
JUDGE:
As I understand your question, the proper methodology was used. Whether or not
it was professionally used is another issue.
RODDIS:
I do understand that.
JUDGE:
Do you understand?
RODDIS:
I do. This is from, when I said professionally used, I mean used in a manner
which generated data which I felt was good data.
GERAGOS:
Now, have you seen data that's not good data that was generated by this same
unit?
RODDIS:
Yes, I have.
GERAGOS:
And what was the, what fix took place between January 21st and January 24th, in
that three-day span, what did your company do to fix the data that the machine,
DISTASO:
Objection, your Honor.
RODDIS:
I haven't that information.
DISTASO:
Objection. The question as stated between January 21st and January 24th.
GERAGOS:
January 21st and the 24th.
JUDGE:
Okay. Is that your objection?
DISTASO:
My objection is that's not the, I'll let it go. Never mind. Withdraw it. My
understanding is that that period of time this particular unit was not
functioning properly, the testimony was it was turned back in to the witness's
place of business and it was repaired, returned back.
JUDGE:
Is that what happened?
DISTASO:
No, that's not correct. I mean, just so the Court knows,
GERAGOS:
Rather than make an offer of proof, I'd like to get his best testimony here
before Mr. Distaso,
JUDGE:
I'll overrule the objection. Go ahead.
GERAGOS:
Now, are you aware of anybody at the Modesto PD who would have been able to fix
this machine in that three-day time period, between the 21st and the 24th?
RODDIS:
I'm not aware of what the Modesto, facilities Modesto Police Department has.
GERAGOS:
Are you aware of where this particular unit was placed in the automobile?
RODDIS:
No.
GERAGOS:
So when you are testifying as to whether the proper methodology is used, isn't
one of the things you want to know is where the unit was placed?
RODDIS:
No.
GERAGOS:
You don't. To you it doesn't matter if it's on top of the car with an antenna
that's at the skyline, or whether it's underneath the car with a signal that is
not able to track the open sky?
RODDIS:
Certainly that does matter to me. By looking at the track, I can see that the
antenna was most likely placed in a reasonable position because, just because of
the consistency of the fixes.
GERAGOS:
You would use the term that this was a good investigative tool; is that correct?
RODDIS:
Yes.
GERAGOS:
You certainly would not consider this to be an, at least this low-cost unit as
applied here to be something that you would want to rely on absent other
information; isn't that correct?
RODDIS:
There are a number things you said there I'm not sure about, because, first of
all, what's the low-cost unit you are referring to as low cost?
GERAGOS:
Well, that term has been used by Doctor Loomis that this is a low-cost unit. Is
that something you disagree with Doctor Loomis about?
RODDIS:
The unit that he, I imagine he's talking about his GPS receiver, which is
intended for use in what he would call a low-cost unit, which has, I assume, I'm
not wishing to put words in his mouth, but is equivalent in comparison to units
used for, say, surveying or other uses, which would require very high accuracy
indeed. So to that extent, yes. But I wouldn't say, call this unit, as you
implied, a low-cost unit.
GERAGOS:
He characterized it as a, gave it a distinction. He gave a distinction between
what he called high-value low-cost. Do you recognize that distinction?
RODDIS:
Yes.
GERAGOS:
And would you characterize this a high-value or low-cost unit?
RODDIS:
I would say, I think that that decision is placed on what you wanted to use the
equipment for.
GERAGOS:
Well, in the case we want to track somebody, somebody's movements, or object's
movements, what would you use the object for? Wouldn't that be a low-cost unit?
RODDIS:
No, I don't think so. I think that this unit, in fact, is one of the better ones
on the market, and it is, in fact, a high-value unit in terms of its use.
GERAGOS:
So you would disagree with Doctor Loomis who says that this is a low-cost unit
RODDIS:
I doubt he said this is a low-cost unit. I
<<<a few lines of testimony were lost>>>
GERAGOS:
a tunnel and cannot get the, underground parking area, where it's difficult or
impossible to get a GPS satellite signal, have a motion detector to see whether
the car stops motion or starts motion, may be more valuable.
<recess>
GERAGOS:
Do you have the Exhibit B in front of you, the letter from Mr. Peach?
RODDIS:
I have C here. Oh, yes, B. Yep.
GERAGOS:
Now, have you read Exhibit B?
RODDIS:
I have.
GERAGOS:
Would you agree with the analysis by Mr. Peach as contained in Exhibit B?
RODDIS:
I, partially. I agree with it as modified by the circumstances of C.
GERAGOS:
I didn't ask you if you agreed with it as modified. I asked you if you agreed
with it as B?
DISTASO:
Objection, your Honor. The witness answered the question properly.
JUDGE:
I don't think so. Overruled.
RODDIS:
Okay. I'm not, Exhibit,
GERAGOS:
Exhibit B makes a variety of statements.
RODDIS:
It does, yes.
GERAGOS:
Okay. And most of those statements are wrong; isn't that correct?
RODDIS:
No.
GERAGOS:
Well, which ones are wrong?
RODDIS:
I don't actually see any segment that's totally wrong, however, there are some
that may be taken out of context, I believe. In fact, since this letter was
written on July 9th we have done some more analysis and had more information as
to a, particularly this one degree issue that we talked about earlier, which was
not reflected, was not known at the time of this letter was written. And so
there are some statements in here which I would modify based on knowledge gained
since we wrote it.
GERAGOS:
Okay. Do you see do you see the portion where he states in there that because
the antenna has been placed or presumably it's referred to in a covert position,
or hidden or secret location? Do you see where he talks about that?
RODDIS:
I think that's mentioned twice. Could you give me the paragraph?
GERAGOS:
How about I come up and actually show you.
RODDIS:
There's one mentioned at the back.
GERAGOS:
The "We feel that some of these occurrences could have been related to GPS
antenna placement and/or interference"?
RODDIS:
Yes.
GERAGOS:
Okay. Do you agree with that?
RODDIS:
And/or interference, yes.
GERAGOS:
Do you agree with the antenna placement?
RODDIS:
I don't think so, no. And that's one of the issues; I think that it says antenna
placement and/or interference, and I believe the one issue that we looked at
here, the spikey thing, was largely caused by interference, not antenna
placement.
GERAGOS:
Okay. Now, the, do you see the last paragraph here where he says: I suspect the
resulting calculation got pulled over by the GPS receiver's averaging program
designed to reject transient spikes. Once a better choice of satellites was
available, it acquired another one and started to correct itself. Due to the
covertness of this device, it often cannot be put in an optimal position to see
the satellites. Do you agree with that?
RODDIS:
Not, not wholly, no. I, I, you must have, understand that we had no knowledge of
where the antenna was placed, and this is a general statement based on many,
many cases that we've looked at. So it, I'm sorry, can I read this one again?
GERAGOS:
Sure.
RODDIS:
The business about the, the, some inadvertent jamming or interference can mimic
this problem. And that's the key sentence. In this case, I suspect that it held
on too long to a satellite, perhaps due to the antenna placement. So the antenna
placement there relates to how long it would hold on to the satellite.
GERAGOS:
Are you,
RODDIS:
When it went over the horizon, the resulting calculations got pulled over by the
GPS averaging program. That is one potential answer to that particular question,
but as I mentioned earlier, that particular question was the one in, bitmap four
with track name ten oh three January oh nine. And, and we now feel that's
probably more likely caused by the interference, which was mentioned in this
letter, rather than the antenna placement.
GERAGOS:
Well, you understand that Mr., Dr. Loomis has testified, and I have it here,
that he disagrees with that statement in its totality because our receiver
cannot track satellites that go over the horizon. We have a protection against
that. Were you aware that the device that's contained in the ST600 does not
track satellites that go over the horizon?
RODDIS:
Well, that is two points to that one. The first one, yes, I was aware of what
that one does, we, and I am aware of it. However, the letter actually doesn't
say that it tracks satellites going over the over horizon. It says satellites
which have just gone over the horizon. In other words, it's no longer tracking
them. It's lost tracking them. And at that point when it's lost tracking,
there's a moment of confusion whereby the thing doesn't know quite what to do
and the averaging circumstances can momentarily produce momentary results until
it acquires another satellite. So we don't actually say in this letter that it
did track over the horizon. The second issue,
GERAGOS:
You don't say,
DISTASO:
Objection, your Honor. He,
JUDGE:
Let him finish his answer, Mr. Geragos.
RODDIS:
The second issue was an issue that Trimble had an issue that did not, that it
had a protection for tracking. There is a setting in the unit that allows you to
set the angle at which it will detect satellites, so if the satellites get too
long you can actually put a setting in the unit to say I want it to be no more
than 15 degrees or 10 degrees or 5 degrees, and I'm sure Dr. Loomis understood
that. When we use this equipment, it's specifically designed for covert use, we
sometimes have to set that angle lower than you would normally set it simply
because it's one, some position is better than no position at all.
GERAGOS:
What would,
RODDIS:
Therefore we do, do that.
GERAGOS:
Well, what did you do in this case?
RODDIS:
I don't know. The equipment's automatic. I would suspect that, as I say, it
doesn't track it going over the horizon. It doesn't. It can't. But it would try
and track it as close to the horizon as it could get.
GERAGOS:
You don't think that when Dr. Loomis read this and he said he disagrees with the
statement by Mr. Peach, do you think he was reading it incorrectly?
RODDIS:
I suspect we were, perhaps wrote the letter poorly and didn't make that clear.
But if I read it myself, if I read it myself, I would perhaps see that it could
be misinterpreted, and I think that's probably what he did. In my sense I say it
reads this way: I suspect that it held on too long to a satellite, perhaps due
to antenna placement, which just went over the horizon. So he could interpret
that to saying it tracked it down over the horizon, but what it meant to say was
that it tracked it too close to the horizon, and therefore it had some problem
before this satellite disappeared. And,
GERAGOS:
Were you aware,
RODDIS:
, probably a poorly worded sentence.
GERAGOS:
Well, you reviewed that letter, didn't you, before it going out?
RODDIS:
Yes, and I read it the way I thought I read it. Obviously other people
interpreted it differently.
GERAGOS:
Well, you had had a discussion with a panel that reviewed all of those tracks;
isn't that correct?
RODDIS:
We, we got together to review it, yes.
GERAGOS:
And that included you, Mr. Peach, technical support and engineering; isn't that
correct?
RODDIS:
Yep.
GERAGOS:
And there were 13 tracks that were sent to you and you reviewed all of those
personally, correct?
RODDIS:
We reviewed them together, yes.
GERAGOS:
Okay. Now the, specifically the, Mr. Peach and you reviewed this; said that in
this particular case with this bitmap, what happened was the device held on to
the satellite too long, and he ascribed that to antenna placement; isn't that
correct?
RODDIS:
No.
GERAGOS:
He never said that?
RODDIS:
No. It says here: I suspect that it held on to the satellite perhaps due to
antenna placement. It doesn't categorically state that.
GERAGOS:
Are you aware that Mr. Peach has made the statement he suspected it could have
been in each situation where the antenna was located within the vehicle that was
not in an optimal position to view the satellite? Were you aware that that's a
statement made by Mr. Peach?
RODDIS:
I'm not sure, but I wouldn't, yeah, it might be one of them that I don't....
GERAGOS:
Well, would you say, would you agree or disagree with the statement that the
antenna placement affects whether it held on to the satellite too long?
RODDIS:
Antenna placement can, certainly, because if it's poorly placed, then the unit
would hold on to the satellite too long. As I said, I think that post-analysis
after this letter says that probably did not happen.
GERAGOS:
Well, are you aware that Orion specifically has made the statement, through Mr.
Peach, that the covert placement can affect whether it holds on too long with
the satellite?
RODDIS:
That's absolutely true. That, as a general statement that's correct.
GERAGOS:
Okay. And could I show you where what I asked that specific question of Dr.
Loomis, page 496, why it's held on too long with the satellite or made a poor
satellite choice; isn't that correct? He makes that statement, yes. You disagree
with that? I disagree. Our receiver cannot track satellites that go over the
horizon. We have a protection against that. Are you aware of that?
RODDIS:
I understand that's what he said. But I also say that I think he's probably
talking about two different topics.
GERAGOS:
You think Dr. Loomis is talking about something different than what you're
talking about?
DISTASO:
Objection, your Honor. It's been asked and answered.
JUDGE:
Yeah, sustained.
GERAGOS:
Well, are you aware that there are some Trimble receivers that can track a
satellite below the horizon?
DISTASO:
Objection, your Honor. Relevance.
JUDGE:
No. Let him answer it. Go ahead.
RODDIS:
No, I did not know of that. I didn't think they could.
GERAGOS:
Okay. Now, the, the, specifically this device as reflected, or as installed,
have you, you reviewed some items at the break; isn't that correct?
RODDIS:
Oh, with regards to the data on the unit sent back, yes.
GERAGOS:
Okay. You're now aware that the unit that was installed on the Land Rover on
January 3rd did malfunction between the 3rd and the 21st, correct?
RODDIS:
Yes. That's right, yes. That was due to the wireless antenna.
GERAGOS:
That it, what was, what was explained to you as to what happened there?
RODDIS:
What I understand happened, and you must understand that I wasn't there, not
involved in it, but that they installed the unit and they had some trouble
connecting with it and which was subsequently traced down to problems with the
wireless antenna. Which was replaced or repaired. And the unit then subsequently
operated.
GERAGOS:
Now, you had indicated, too, that there was, in response to Mr. Distaso's
question that you will show various readings in there, in the data sets that
have miles per hour when the car actually may not be moving at all?
RODDIS:
Yes.
GERAGOS:
And that's because there, the machine is constantly correcting itself; isn't
that correct?
RODDIS:
Right.
GERAGOS:
Over,
RODDIS:
Yeah.
GERAGOS:
And the reason it's correcting itself is because there are inherent problems
with the way that the machine calculates its position; isn't that correct?
RODDIS:
No.
GERAGOS:
Well, the, it has a, specifically in this case a, has to get a setting from at
least three satellites; isn't that correct?
RODDIS:
Three or four, yeah.
GERAGOS:
What's the reason for the fourth?
RODDIS:
It needs to have an initial time. If it doesn't have the time it actually needs
four, otherwise it can do it, time; otherwise it can do it with three.
GERAGOS:
Are you familiar with the term or acronym WAOS?
RODDIS:
Yes, uh-huh.
GERAGOS:
What is that?
RODDIS:
Good point. Wide Area Orientation System, I believe.
GERAGOS:
Okay. Does this device, the informer, ST600 --
RODDIS:
No.
GERAGOS:
, have that?
RODDIS:
No.
GERAGOS:
Are you familiar with the acronym NDGPS?
RODDIS:
No.
GERAGOS:
You don't,
RODDIS:
ND, I'm not sure what ND stands for.
GERAGOS:
How about a differential,
RODDIS:
Or differential, sure, yes.
GERAGOS:
Nation, Nationwide Differential GPS?
RODDIS:
Yeah. Doesn't have that.
GERAGOS:
Doesn't have that at all. So if it doesn't have those, are you aware that
without those, that the, that there's a body of literature that says that
without one of those two methods incorporated into the device, that that
produces great -- great inaccuracy.
RODDIS:
Oh, you want me to give you some GPS history again.
GERAGOS:
I don't want, no. I don't want you --
RODDIS:
The,
GERAGOS:
I want you to tell me why don't you have a differential GPS installed in this
device?
RODDIS:
Because nowadays you don't need it.
GERAGOS:
And your opinion as to why you don't need it is why?
RODDIS:
Because as I said to history, that previously there was a thing called selective
availability, and selective availability, which was one of the good things
President Clinton did, was arrange to have it discontinued, but selective
availability meant that for commercial uses it was a jitter or a noise imposed
on the system which created some substantial error. And it was done
intentionally so that people couldn't use this for missiles or whatever,
whatever, whatever. Since President Clinton's last year, whatever that was, the
last year. 2000? Anyway, he had that switched off. Differential GPS was
primarily a way of correcting for that error. And so when that error was
switched off and differential GPS was no longer required to give a reasonable
accuracy for the system, the Wide Area Orientation System gives greater accuracy
and is used, is intended to be used for aviation in part. So, but in this case
we don't believe that that accuracy is required for this particular use of this
unit. To us it doesn't really matter terribly much if it is ten feet down the
street or ten feet in the opposite direction. Most of the time having the
measurement within a few feet is accurate enough.
GERAGOS:
Well, the, have you ever heard of the term Positional Dilution of Precision?
RODDIS:
Yes.
GERAGOS:
Okay. And isn't that, can't you get bad, and it goes by the acronym PDOP?
RODDIS:
Yeah.
GERAGOS:
You can get bad PDOP, which in simple terms mean unable to compute the precise
location to the GPS receiver due to bad geometry; isn't that right?
RODDIS:
Uh-huh.
GERAGOS:
Yes?
RODDIS:
Yes.
GERAGOS:
One of the primary reasons for a bad PDOP is a covert positioning of the
antenna; isn't that correct?
RODDIS:
I wouldn't say private, I'm sorry. It is a reason, yes.
GERAGOS:
Okay. And in fact, isn't there a body of thought that, that that condition, a
bad PDOP, will undoubtedly occur whenever you have a covert placement of an
antenna?
RODDIS:
No.
GERAGOS:
Isn't there also, you can't, do you ascertain a bad PDOP with this particular
unit?
RODDIS:
No.
GERAGOS:
And how do you do that?
RODDIS:
Well, actually it's not recorded.
GERAGOS:
Right. That's not,
RODDIS:
But the unit does do it. It does it on the fly, as it were.
GERAGOS:
Do you have that recorded anywhere?
RODDIS:
No.
GERAGOS:
And when you say it does it on the fly, unless you have the data
recorded, we don't know that it's done it, do we?
RODDIS:
No, probably not. It's part of the internal workings of the unit. I will say
that covert antenna is a very poor term to use because it could be covertly on
top of the vehicle, in which case it would be fine. You can put a very wide
range of instruments and a bad PDOP can be caused by parking in between two
large buildings, as well as a covert antenna.
GERAGOS:
Except you don't know where the antenna was placed in this case.
RODDIS:
No. You asked me a general question and I was responding with a general answer.
GERAGOS:
Well, let me ask you a specific question. The Land Rover, the Chevy S10, the
Lincoln town car, the Dodge Dakota, or even a detective unit from the Modesto
PD, where was the antenna placed on any of those cars?
RODDIS:
I don't know.
GERAGOS:
Yet you are willing to say there's no problem with bad PDOP even though we have
no output of it whatsoever and you don't know where the antenna was placed?
RODDIS:
I have looked at these tracks. If you have a bad PDOP, a bad PDOP is a transient
thing in terms of where the satellites are, and satellites go around all the
time. So one moment you might have a bad PDOP, ten minutes or an hour later you
might have a good one. So what happens basically if that occurs, in a bad
antenna placement it blocks the view of the sky so you can't see the satellite,
and that occurs at cyclic intervals as the satellites go around. So what would
happen in the track is that you would get pieces missing out of the track when
the PDOP was bad, either one or two spots or three spots, but the track overall
still remains true to itself. And when I look at this track, I don't see many of
those, if any, the occasional case where there's been a dropped, dropped fix. So
I would assume that the antenna positions were fairly good.
GERAGOS:
Well, the, when you say you looked at this information, you've only been given a
partial amount of the information for that,
RODDIS:
Right. And that's what I'm basing my opinion on. I haven't seen the rest.
GERAGOS:
Well, as, as an engineer, as a scientist, or as an expert, wouldn't you want to
see all the data sets that were involved in the tracking devices for this
particular operation?
RODDIS:
I've been asked to look at these and I assume that they're representative or the
court wishes me to see these ones. I haven't looked at the other ones. I would
if somebody asked me to, but I haven't.
GERAGOS:
Well, I'm asking you to,
RODDIS:
No.
GERAGOS:
What I'm asking you, if I'm telling you they are not representative, and you
know now from your review at the break that one of these devices was completely
inoperational for three weeks, wouldn't that be, wouldn't that tend to negate
your claim that these are representative when you know that one device failed
for three weeks in a row?
RODDIS:
Oh, I'm sorry if, I didn't intend to say these are representative. I've just
based my opinion on these tracks that I have seen. And, and as far as I can see
within these tracks, that the units were working correctly at that point. With
two, well, when I say correctly, there were two issues with the one degree
offset, which we have explained, and understand now.
GERAGOS:
Do you know what a multi-path of the GPS signal is?
RODDIS:
Yeah.
GERAGOS:
Now, what is, doesn't the multi-path, isn't that a critical problem in a GPS
unit?
RODDIS:
Multi-path is another name for signal reflections. So you get more than one
signal arriving at the receiver at the same time. That can be an issue. However,
the receivers are very good at eliminating it and taking it out. And, again, if
you
do
have a lot of multi-path, it's reflected in core tracks and great variability in
the positions.
GERAGOS:
Well, the, the spike that's on the particular exhibit that's reflected in
Fresno; you know which one I'm talking about?
RODDIS:
Yes, I do.
GERAGOS:
Have you ever seen that particular spike before? In --
RODDIS:
You mean that type of spike in other cases?
GERAGOS:
Yes.
RODDIS:
That particular one is a big one, and it's rare. I've seen one once before like
that. And, again, it was near an airport. But I must say in my experience that
that type of spike is rare.
GERAGOS:
Does that mean that that's the only time you've seen it? Or one other time?
RODDIS:
Yes, one other time.
GERAGOS:
And so we have four days, basically four days' information here; is that
correct?
RODDIS:
Possibly. I haven't done, I didn't do the math, but possibly.
GERAGOS:
Okay. Out of those four days of information, I've now told you that one of these
other receivers produced by your company malfunctioned for three weeks, you've
just testified that you've seen a spike that you've only seen one other time in
your, what, almost 30 years of doing this, that you also are seeing something
that's called a firmware error when it crosses a longitudinal or latitudinal
line?
RODDIS:
Yeah.
GERAGOS:
And you're comfortable just taking these four little data sets out of a larger
group of maybe 3,000 pages' worth of information and saying that in this case
everything's hunky-dory?
RODDIS:
I haven't, I'm sorry, but I have been asked to testify on these tracks; and
these tracks, to my viewpoint, are typical of decent tracks of the type I have
seen over many years. That spike, if you like, is like a flash of interference
on a TV set. You see a good program and a burst of interference, but the program
before and afterwards is fine. And that's the way I see this one.
GERAGOS:
Well, were you aware that the Modesto PD had installed an Orion tracking unit
onto an unmarked police vehicle to do their own test?
RODDIS:
Perhaps. I don't know.
GERAGOS:
Did anybody tell you they had done their own test?
RODDIS:
No.
GERAGOS:
Did they tell you that they instructed the person to go to three separate points
with the tracking device in the car?
RODDIS:
No. They may have done it, sure.
GERAGOS:
Well, did they tell you that they, that the devices, when installed, did not
track accurately?
DISTASO:
Objection, your Honor. That's not, that's, he has no good faith basis for that.
GERAGOS:
Status was it was not,
JUDGE:
Hold on a second.
GERAGOS:
, old data.
JUDGE:
Do you have something here to support your question?
GERAGOS:
Absolutely. Page 24014 of Bates stamped --
JUDGE:
Do you have that, Mr. Distaso?
GERAGOS:
Now,
JUDGE:
Wait. On your representation, I'l let you ask the question. Go ahead.
GERAGOS:
Thank you. Did they tell you that the system when they, when this did this was
producing sporadic communications and that there was some type of interference
in the area precisely in the area where you're testifying to about Exhibit
number 1? And, did anybody tell you that, give you that piece of information?
RODDIS:
I'm, I'm not, what, no. And it's a very wide area. I'm not sure
which part you're talking about, where the interference would have been.
GERAGOS:
The Berkeley Marina.
RODDIS:
Berkeley Marina. Okay.
JUDGE:
I think you've got to give him that exhibit number.
GERAGOS:
I can do that.
RODDIS:
I'm not sure what relevance it has, but I didn't know any of that.
GERAGOS:
Well, you're unaware of any of that information; isn't that correct?
RODDIS:
Right.
GERAGOS:
Okay. Were you aware that, remember you had pointed me to a location where there
was a receiver that started producing data sets for January 24th; do you
remember that?
RODDIS:
Yes.
GERAGOS:
Now, were you aware that somebody in Modesto PD had apparently decided to fix
the problem with the equipment between the 21st and the 24th?
RODDIS:
I was told that on the break, yes.
GERAGOS:
Were you told that by Mr. Distaso? Isn't that correct?
RODDIS:
Yeah, uh-huh.
GERAGOS:
Is that a yes?
RODDIS:
Yes.
GERAGOS:
Now, the, do you recommend, by the way, is there a warranty of some kind on
these units?
RODDIS:
Yes.
GERAGOS:
Is the warranty good if somebody other than an authorized Orion technical
support person does the repairs on the unit?
RODDIS:
Our, our policy is that we receive units back and repair them with, without
arguing about what happened. So I don't, there's is no policy in place on that
one.
GERAGOS:
Do you recommend that people try to repair the units themselves?
RODDIS:
I think that it depends on what they're doing by "repair." If you're talking
about changing an antenna or moving an antenna or doing something like that,
that's hardly a repair. It's just unplugging an antenna and plugging a new one
in. Opening the box and fiddling around inside would be a different issue. I
doubt very much they did that, though.
GERAGOS:
Did you not tell me, though, that when the unit was sent back there was a pin
error? Isn't that what you said?
RODDIS:
There was one of the microprocessors was not plugged in properly.
GERAGOS:
Okay. Did that require taking the device apart?
RODDIS:
For us, yes.
GERAGOS:
And to find out that there was a microprocessor that was malfunctioning?
RODDIS:
Yes.
GERAGOS:
Okay. Is that something that you would consider to be just a random unplugging
or plugging in that could be accomplished by somebody who's not trained on this
particular device?
RODDIS:
I doubt very much whether it would be undetected without any, I mean people
could do what they want to do. I don't, I can't stop people doing things, but it
could, people could remove that, I suppose, but it would seem that we, we saw no
traces of that, I don't think, when we got the unit back again.
GERAGOS:
Now, is there, is it your understanding when these devices are used that you
need a base unit of some kind?
RODDIS:
A computer, yes. And a modem.
GERAGOS:
Okay. What was used in this case?
RODDIS:
I'm not sure.
GERAGOS:
Well, doesn't the base unit or, give you greater accuracy if you have one?
RODDIS:
No, no. It's just like a modem on a computer, a way of getting data.
GERAGOS: And how was the data saved in this case?
RODDIS:
It was on a computer. Well, downloaded, the data is contained in the unit and
then downloaded to a computer and then the personnel would then take the data
from them.
GERAGOS:
And where is that software? Who provides the software for the data set
retrieval?
RODDIS:
We do.
GERAGOS:
And have you looked at that? Have you looked at all of the data set retrievals
that were done in this case?
RODDIS:
No. Only these ones that we've looked at here.
GERAGOS:
Have you looked, do you recommend that the, when possible that you clear the
memory on the unit?
RODDIS:
It depends on the length of the case. The unit has a limited memory in it and so
many thousand records, and so you don't want it to run out of memory, so if you,
if you are taking a lot of records and set to take a lot of records over a long
period, then you would be recommended to clear the memory to make room for new
records, yes.
GERAGOS:
Okay. Now, the July 9th letter by Mr. Peach that you had reviewed prior to it
going out, that's a correct statement, isn't it?
RODDIS:
Yes.
GERAGOS:
Okay. Now, that, he at first was analyzing the bitmap one track for the Dakota
and the bitmap three track for the S-10, correct?
RODDIS:
Right.
GERAGOS:
And he and you believed that when crossing the line of latitude or longitude,
the GPS receiver occasionally does not update all decimal places of the
longitude; wasn't that your explanation at the time?
RODDIS:
This was to do with what I mentioned earlier about the compression issue.
GERAGOS:
That was, that analysis was wrong as of July 9th, 2000 and 3, correct?
RODDIS:
Yes. Well, that's correct. As of, when I say it's wrong, it's partially correct.
We later discovered through more experiments and more, the problem with the
Trimble receiver, which is actually triggering this issue. It is a compression
problem with the download, but it would have been triggered off by the Trimble
issue that we didn't discover until after this letter was written.
GERAGOS:
When you wrote or approved the other letter sent on July 9th, 2003: If the
timing is such that this happens when a fix is taken, one fix can be exactly one
degree off. For the most part our units are designed to ignore this data, but
sometimes the bad data gets recorded; That statement is inaccurate, isn't it?
Phrased better, you no longer believe that statement?
RODDIS:
Let me just read that one just to make sure it's in context here. Yes. It's
probably too generous a statement, but that refers to the fact that when the
Trimble unit put out the A, if you look, I don't know if Dr. Loomis explained
all that, but instead of a zero, there are various malfunctions that, that, of
data that can occur, and a unit is programmed to generally disregard or, or get
rid of all the bad stuff. That particular one had not been detected and had not
been rectified at that point. And so it did get through the system and did get
recorded, yes.
GERAGOS:
And you had a, and you and Mr. Peach had an alternate explanation which you were
comfortable with when you sent it to the Modesto PD which later, as of January,
proved to be wrong; isn't that correct?
RODDIS:
That explanation was our best explanation due to our knowledge at that point,
yes.
GERAGOS:
And your knowledge at that point on July 9th of 2000 and 3 was some time after
these data sets were recorded back in January of 2000 and 3, correct?
RODDIS:
Correct, yes.
GERAGOS: Okay.
RODDIS:
We're not involved in that,
JUDGE:
Wait. What's your answer?
RODDIS:
I'm sorry. Yes, except that we, we have no knowledge of what was going on here
until much later in January, yes.
JUDGE:
Go ahead.
GERAGOS:
Okay. Now, the next sentence is that: Normally the software automatically
corrects the problem. However, in order to speed up the download over-the-air,
the whole parts of degrees are only sent when a major change occurs. Therefore
some records which follow will also be wrong, until the compression engine
resets upon a change in latitude, longitude, for example, it crosses back, or
when the time change, changes increments such that the minute value changes. Or
the minute value changes. You, you no longer believe that either, do you?
RODDIS:
No, I, that's exactly what I said earlier this morning.
GERAGOS:
You don't believe that anymore; isn't that correct?
RODDIS:
No. That is the same thing I had said was true earlier this morning. It's the
compressing issue.
GERAGOS:
And, as you say: The direct download is not compressed, this type of error would
not be seen. Do you agree with that?
RODDIS:
I do except for the fact this was not a direct download.
GERAGOS:
This type of thing can also happen due to poor GPS antenna placement, although
this may not be exactly one degree. Do you agree with that?
RODDIS:
Yes. That's true. And, in fact, I would say that it would be normally much less
than one degree.
GERAGOS:
Due to the covertness of this device the GPS antenna often cannot be put in an
optimal position for the GPS satellites to see 100 percent of the time. Do you
agree with that sentence?
RODDIS:
That, that's true. As I said, I'm not sure it's true in this case, but generally
it is true.
GERAGOS:
And: The records, which have been thrown off, are one degree of latitude or
longitude from where it should be. That is still accurate, is it not?
RODDIS:
Well, it's certainly true in this case. If you, if you look at, we're saying the
records which we see in this track, which were wrong, were in fact exactly one
degree off, which is what that sentence says, yes.
GERAGOS:
Correct. So the conclusion or, from the data set was accurate. It's just your
reasoning behind what the problem was was incorrect at that time; isn't that
right?
RODDIS:
I'm sorry, I didn't understand that sentence.
GERAGOS:
Well, you had, had a conclusion that there was a one degree shift, right?
RODDIS:
Yes.
GERAGOS:
Of either longitude or latitude?
RODDIS:
Right.
GERAGOS:
And your reasoning for it was that occasionally the GPS receiver does not update
all decimal places of the longitude; isn't that correct?
RODDIS:
What, what, our reasoning at that time was our best estimate, because, remember,
this was something that we hadn't seen before, was that it was a, an error in
the decompression, which is what we talked about there. Because that's the only
thing that we could see which would cause an exactly one degree error. As it
turns out, we talked about it later, that it turned out to be a problem with the
Trimble receiver, which also caused the one degree error.
GERAGOS:
Right. So the explanation that you posited for the one degree error had nothing
to do with all the explanations in that paragraph, it had to do with something
that was endemic, if you will, to the Trimble receiver?
RODDIS:
Right.
GERAGOS: And that didn't stop you as a scientist or as an engineer from trying
to give a possible explanation as to what was going on at the time in July,
correct?
RODDIS:
Correct.
GERAGOS:
Okay. Now, you also had, you see where in the July 9th letter it talks about
bitmap number two track name Rover for January 3rd?
RODDIS:
Right.
GERAGOS:
Okay. Now, here there's an explanation about that this may be related to a
problem with data which the census bureau receives from the County of Modesto;
isn't that correct?
RODDIS:
That's what it says, yeah.
GERAGOS:
Now, Orion on its web site actually touts the fact that it has done work with
Modesto in mapping the city; isn't that correct?
RODDIS:
No.
GERAGOS:
Are you familiar with your web site?
RODDIS:
I am, but I doubt very much it has that in it.
GERAGOS:
There's nothing in there, no reference to Modesto and the use of Orion GPS
satellite technology from the City of Modesto?
RODDIS:
Absolutely not.
GERAGOS:
Okay.
RODDIS:
Not that I know of, anyway.
GERAGOS:
Now, do you agree that the bitmap two track named Rover problem, that that was
related to the U.S. census bureau information, that paragraph there in Exhibit,
RODDIS:
It was a mapping problem with the data that they were using for the maps that
they used at that time, which were, which were derived from the TIGER data.
That's what I understand, because when we changed to more recent data, a map
with better data, then the errors went away. You must understand that with the
GPS system, the map accuracy is just as important as the GPS accuracy.
GERAGOS:
Now, the, do you, so you're still, you still stand by the explanation that's
reflected on Exhibit B as to bitmap two having to do with the census information
being inaccurate?
RODDIS:
I, I understand that this, yes, I, to the, to the extent that when we replaced
this data with a more modern map set, more accurate map set, the problem went
away, yes.
GERAGOS:
Now, the portions where it says: The error in the data increases the further
away you get from your chosen datum; towards the western edge of the county, the
error is much less. --
RODDIS:
Yes.
GERAGOS:
, if you look carefully around the edges of the county; you agree with all that?
RODDIS:
In general, yes, that's true.
GERAGOS:
Okay. And bitmap four track named S-10 for January 3rd, do you see that?
RODDIS:
Yes, I see that.
GERAGOS:
On Exhibit B? Do you also, do you disagree or agree this is an example of a GPS
receiver making a poor satellite choice?
RODDIS:
This I believe refers to the spike that we talked about earlier this morning.
And again, our opinion is, we talked about it later and since, is that it's, it
does say possible receiver making a poor satellite choice or some form of radar
interference. We've decided since that we think it's more likely to be radar
interference; but, yes, so I do agree with that sentence.
GERAGOS:
Okay. Now, the, on the January 13th letter, were you also involved in that?
Which is Exhibit C.
RODDIS:
Yes. Yes, indeed.
GERAGOS:
Okay. And in that letter you say your previous explanation regarding compression
was feasible but has since been shown to be incorrect. In fact, our
compression system was working properly. Is that an accurate statement?
RODDIS:
Yes. It is involved in the fact, though, that the, the, the spike lasted for
several minutes rather than going away. The compression system was working
correctly, but it still, the fact that it exists meant the thing lasted longer
than it would have done otherwise.
GERAGOS:
And then, lastly, the bitmap four, all of your experts agree that the portion of
that track is not valid and we do rarely see unexpected GPS positions. Is that
accurate?
RODDIS:
Yes. Right.
GERAGOS:
And you're not sure of the real cause, but the presence of the nearby airport
leads you to suspect that the equipment there may have contributed to the
problem?
RODDIS:
Exactly.
GERAGOS:
Okay. Now, have you done any studies or have there been any studies done that
you're aware of that check GPS accuracy in the presence of airports?
RODDIS:
Well, there have been a lot of studies done in that regard. I haven't seen any
in this particular instance. The reason that we say that is some of the airport
equipment uses frequencies which might, which relate to the frequencies which
are used for GPS, radio frequencies, that is, and therefore could contribute to
interference, but I haven't seen any further than that.
GERAGOS:
And how far away is the Berkeley Marina from an airport; do you know?
RODDIS:
I have no idea.
GERAGOS:
How far away is the location where that spike was from an airport?
RODDIS:
There were two airports on the map, one of which was fairly close and in a
matter of perhaps a mile, I believe. The other one was further away. I'm just
guessing. I'd have to look at it on here. I'll show you. Excuse me. Yes. It's,
according to this map, probably five miles away.
GERAGOS:
So one of your, one of your possible explanations for this rare occurrence is
that a, some kind of a radio transmitter or radio equipment up to five miles
away could have affected the,
RODDIS:
Exactly. It was, as I said, one airport, which was in fact much closer than
that, less than one mile, I think, on the map when I saw it.
GERAGOS:
So based upon what you know about GPS, being within five miles of an airport
could cause some occurrence that here would spike off approximately how many
miles?
RODDIS:
Seven.
GERAGOS:
How many?
RODDIS:
Seven.
GERAGOS:
Seven?
RODDIS:
I, I, yes. The, the actual length of the spike is almost irrelevant, but once
the interference occurs, it's difficult to say what would happen. I, I must say,
again, that this is a rare thing, very rarely seen. It's our best guess of what
happened in this particular instance. There may have, for all I know, been some
other interference close by. We're just guessing about the airport.
GERAGOS:
What other kind of interference would cause a seven mile glitch?
RODDIS:
Well, the most likely one, which is very unlikely, but is a GPS jamming system,
which is used by the military fairly often. If somebody was testing a GPS
jamming system, that's the most likely issue, most likely thing. And sometimes
that is, has been used recently with, because of the security issues since 9/11.
That's the most likely one. Other than that, as I said, a very high power radar
is potential; but, again, it's not, it's not, not seen very often.
GERAGOS:
But it happened in this case. It happened in this case?
RODDIS:
I'm not happy with it, but,
GERAGOS:
No, I said,
RODDIS:
, nevertheless,
GERAGOS:
No, I said it happened in this case?
RODDIS: I would suspect that something like that happened in that case, but I
don't know what.
GERAGOS:
Okay. But it, it's a fair statement that we, one of the other problems that was
detected here, also, you had a reasonable explanation for it in July, and you
revised that explanation six months later, correct?
RODDIS:
I think those things are rather different, though. In one case we have an
exactly one degree error, which can only be triggered by certain things, in
which we did know it was most likely associated with our unit. And so we were
looking for reasons, which were pretty obvious reasons at the time, and then
were corrected. But it's still the same reason, but a different triggering
thing. This other one here, as I said, it's like a burst of interference on a TV
screen. You don't know where that interference comes from, you just know that
it's there. The track doesn't have the right characteristics over that area and
something's wrong, but I don't know what it came from. I think there were two
different, two different things.
GERAGOS:
Which, were either of these receivers the same receiver, the receiver with the
data sets that you have as Exhibit 1 in front of you, are either of those that
are in there, contained in there, the same receiver that malfunctioned from
January 3rd to January 21st?
RODDIS:
I believe so. I believe that the bitmap two one, yeah, that I have on this
letter.
GERAGOS:
And the,
RODDIS:
The receiver itself, I don't think it malfunctioned, if we want to talk about
the receiver. I think it was the, the other parts of the circuitry. However,
maybe that's
semantics.
GERAGOS:
I'm sorry, it was what?
RODDIS:
However, I said maybe that's semantics. But I don't think the receiver itself
was bad. I think it was the memory recording parts of the circuitry.
GERAGOS:
Well, when you say the, you're not able to tell what malfunctions unless you get
a data set; isn't that correct?
RODDIS:
I had, no, well, it depends which malfunction we're talking about and what we're
doing.
GERAGOS:
Start with the one that you answered me, the bitmap two.
RODDIS:
Yes.
GERAGOS:
Okay. That's the receiver that malfunctioned from January 1st to January 20,
January 3rd to the 21st, correct?
RODDIS:
Yes. That's right.
GERAGOS:
Somehow somebody in Modesto PD got it up and running again by January 24th,
correct?
RODDIS:
I believe that was an antenna change. Just putting a better or new antenna on.
GERAGOS:
Except when you got it back you said there was a microprocessor error?
RODDIS:
Yes, but that was, that was in June, or much later in the year, as I now
realize. It was a different issue.
GERAGOS:
Okay. So we had three separate issues that we know about with this one receiver
in the span of six months, correct?
RODDIS:
I think, well, one of those issues for sure, the antenna, was nothing to do with
the unit per se. It was due to the fact that a bad antenna was connected to it.
I don't think the unit itself was the problem.
GERAGOS:
That's, you're basing that on what the DA whispered in your ear at the break.
DISTASO:
Objection, your Honor.
RODDIS:
Well --
GERAGOS:
The DA,
JUDGE:
The, no,
GERAGOS:
, what the DA told you at the break.
JUDGE: No, objection overruled. Is that what the district attorney told
you?
RODDIS:
Well, what I was told was that the unit wasn't working and then they replaced
the antenna and it suddenly started working, so I assume that the problem was a
bad antenna.
GERAGOS:
And the,
RODDIS:
But it was based on what they told me; yes, you're quite right.
GERAGOS:
Right. You didn't have any information separate and apart; you didn't do any
investigation on that, did you?
RODDIS:
No.
GERAGOS:
Then that same unit apparently had a, what we've identified here as the bitmap
two track name Rover problem as well, correct? It's the subject of both letters,
right?
RODDIS:
Was that the same unit?
GERAGOS:
Isn't that what you just said? Bitmap two is the same unit?
RODDIS:
The, the unit that had the, the spike on it that we just talked about
with the radar, and whatever, I think was a different unit.
GERAGOS:
I think you're correct. I think you had identified the unit that was,
inoperational, I mean was not operating from January 3rd to the 21st as being
the same one as bitmap two; is that, is that your testimony?
RODDIS:
Yes. Serial number 7476, yes.
GERAGOS:
Okay. That's also the same unit that you say when it was returned to you some
time in May had a microprocessor or pin error; isn't that what you said?
RODDIS:
Yes.
GERAGOS:
Okay. So is it a fair statement that there were three problems with that unit in
the span of five months?
RODDIS:
I'm not sure what the third problem was.
GERAGOS:
Well, we have a bitmap two issue here of some kind that required explanation, in
both letters?
RODDIS:
It has nothing to do with the unit. That was just the maps. The maps the unit
displayed on. Had nothing to do with the unit.
GERAGOS:
How about the data sets that were produced?
RODDIS:
I don't think,
GERAGOS:
Were there problems with the data sets?
RODDIS:
They were fine, I think.
GERAGOS:
Wasn't the whole, wasn't the whole problem that you discussed that the data sets
had a problem because of the, the maps that were used, that they did not plot it
correctly?
RODDIS:
Well, I think you would, the, the, the map that the data is displayed on has
nothing to do with the unit. So, I mean, you could take the data out of the unit
and display it on a blank sheet of paper if you wanted to. I, I think that it's
wrong to say that that unit had a problem with its maps. The map is something
totally separate from the unit.
GERAGOS:
Well, the, what good is the unit if it's not able to display it onto a correct
map? It's useless, isn't it? I mean you, you can't plot, the whole idea of the
unit is to plot a location. So if it's not plotting correctly on a map, what
good is the unit?
RODDIS:
I'm sorry, I don't understand that. Because if the unit can be producing
perfectly good data, and yet if you use the wrong map with it, then it's not
going to display properly. But it's nothing wrong with the unit, it's the map
that's the problem.
GERAGOS:
And your explanation for that, or the Orion's explanation, was that it had
something to do with the maps, correct?
RODDIS:
Yes.
GERAGOS:
Are you aware that Dr. Loomis disagreed with that explanation?
DISTASO:
Objection, your Honor. Misstates his testimony.
RODDIS:
I doubt,
JUDGE:
Hold on.
RODDIS: he did, but if he did,
JUDGE:
Hole on, Doctor.
GERAGOS:
I'll give him the page and line number.
JUDGE: All right.
GERAGOS:
The, the units, at the very least this unit, which you identified by the serial
number, had failed or had a problem twice within a five month period, correct?
If you don't include the mapping problem.
RODDIS:
It had a problem with an antenna and had a subsequent problem with it, something
internally, yes.
GERAGOS:
Well, somebody told you it had a problem with an antenna?
RODDIS:
Yes; that's right, absolutely.
GERAGOS:
And it had a microprocessing unit or pin error, correct?
RODDIS:
That we know about.
GERAGOS:
And the third error, which required two separate letters to try to analyze,
you've laid off on the maps; isn't that correct?
RODDIS:
Yes.
GERAGOS:
Okay. Now, the GPS, the system itself, is based upon in large part things that
are called pseudo ranges; is that correct? Do you know what a pseudo range is?
RODDIS:
I do.
GERAGOS:
What is a pseudo range?
RODDIS:
I'm not sure why in is important to this case, though. It's an internal
calculation that's done to, to, in regards to the satellite locations in
distances and times from them.
GERAGOS:
And isn't that part of the methodology that's used for a GPS device?
RODDIS:
Oh, absolutely, but it's,
GERAGOS:
That's why it's important to this case.
JUDGE:
Let him finish his answer, Mr. Geragos.
GERAGOS:
Okay.
JUDGE:
He's answering your question here. You said "absolutely," then he stepped on
your answer.
RODDIS:
Yes. Those, those are partial, those are, how shall I say it, intermediate steps
in the calculation of latitude and longitude, which is carried out within the
Trimble receiver. Our receiver doesn't see any of those. Our, our equipment
doesn't actually see any of those calculations.
GERAGOS:
But isn't that part of the methodology of GPS positioning?
RODDIS:
Absolutely.
GERAGOS:
Okay. Now, does multi-path or multi-path results, can that result in a, produce
multi, the doctrine, if you will, or the, the incidents of multi-path, can that
have an impact on a true pseudo range?
RODDIS:
Yes, it could, but the equipment is designed to reject those things.
GERAGOS:
And specifically the equipment is designed to, you said, correct that?
RODDIS:
Reject.
JUDGE:
Reject.
GERAGOS:
Reject that. And that's when the condition is specifically where the antenna is
on the top of the car roof; isn't that correct?
RODDIS:
No. I think it would certainly apply when the antenna is on the car roof, but
certainly the equipment does the rejections and the calculations wherever the
antenna is.
GERAGOS:
Doesn't the, doesn't it read in multi-path result under longer than true pseudo
ranges when the antenna is not in an optimal place?
Hugh Roddis” I think that multi-path is more likely if the antenna is not in
place,
GERAGOS:
If you,
RODDIS:
, not in the best place, yes.
GERAGOS:
Right. And if your multi-path results, then you get a longer than true pseudo
range, correct?
RODDIS: Potentially, yes.
GERAGOS:
And if you have a longer than true pseudo range being measured between the
antenna and the satellite, what happens?
RODDIS:
You'd get an incorrect location.
GERAGOS:
Okay. And that condition would occur more often than not when the antenna is not
in plain view on the top of the car roof; isn't that correct?
RODDIS:
I would say it's more likely to happen, but I wouldn't say it happened more
often than not, no.
GERAGOS:
Okay. It's more likely to happen when the antenna is covertly placed, correct?
RODDIS:
Yes.
GERAGOS:
Now, in this case the, you have said that the antenna was bad from, that's what
you were told, between January 3rd and January 21st, correct?
RODDIS:
Not the GPS antenna.
GERAGOS:
Which antenna?
RODDIS:
The wireless antenna.
GERAGOS:
Okay. And do you know where the wireless antenna was placed?
RODDIS:
No, but sometime in that period. I don't know, I'm not sure I remember.
GERAGOS:
Do you remember, do you know where the GPS antenna was placed?
RODDIS:
No. I'm sorry. Did you ask if the wireless antenna was placed when?
GERAGOS:
Yes. Do you know when it was?
RODDIS:
No, I don't, and I don't know where the GPS antenna was either.
GERAGOS:
Okay. And wouldn't that be critical to determining whether or not those
multi-path errors or multi-path occurring in, for you to make a scientific
determination as to whether or not you were getting longer than true pseudo
ranges?
RODDIS:
What happens when you get bad pseudo ranges, a lot of multi-path is you get
erratic data. The tracks that I have seen that I have here don't show evidence
of that erratic data, except in some very minor instances, which I find actually
more interesting than not.
GERAGOS:
Well, when you say that there is an erratic data, we have situations, here I'll
just pick the first one on 1 GG. We have fixes that are occurring here where the
speed of the car is ranging from a hundred and five miles per hour up to 2300
and 23 miles per hour, and going back down to 97; is that correct?
RODDIS:
Yeah. That's what it says. Now, let me just have a look where this is and what
it looks like on the map.
GERAGOS:
Okay.
RODDIS:
That's this one. That is the one where the radar stopped going, which we've
explained as what we think happened there.
GERAGOS:
The, the data that's here, you say self-corrected for how long?
RODDIS:
No, well, I'm not sure if, well, if you look at it on the chart here, it came
back in here, wherever that was.
GERAGOS:
You're pointing to 1 double E?
RODDIS:
Right. And you can see that's the spike where the radar started to go wrong up
here, where it started to hit, say, several hundred miles an hour. You remember
I said that if the speed is just calculated by the space between the dips. When
the spikes started to go off in the wrong direction, the speeds were calculated
to be very, very high because the positions were wrong during that one spike
that we discussed several times this morning.
GERAGOS:
Now, the, you, that's the only explanation you can find for that? Or you have
actually no explanation for that other than there may have been radio or radar
interference, correct?
RODDIS:
Yes, that's right. But the rest of the tracks doesn't show a lot of incidents of
multi-path or...
GERAGOS:
Let me show you 1 double D, which was shown to you this morning.
What
happened in this instance? Which is January 9th also, but in the morning.
RODDIS:
I'm sorry, what do you mean?
GERAGOS:
Well, where is the, where is the miles per hour, where is the longitude, where
is the latitude?
RODDIS:
Well, the longitude and the latitude there, I don't know. This particular page
has been printed out oddly. It's been printed out with a column on top of
another column, but the, the speed is there, but it's the way it's been printed
out. I don't know why it was printed out that way.
GERAGOS:
Well, do you have an explanation as you sit here as to why 1 double D appears to
have two columns collapsed?
RODDIS:
No. No. No, something to do with the printer and the computer.
GERAGOS:
Well, does 1 double D refer to 1 BB?
RODDIS:
I, I don't know. I think so, probably.
GERAGOS:
Could you take a look at that and tell me if it does?
RODDIS:
Yeah. Yep. It seems to.
GERAGOS:
Okay.
RODDIS:
This one section of that does.
GERAGOS:
Fair to say that 1 double B indicates that the object that the antenna and/or
the receiver had been placed on was moving?
RODDIS:
Well, parts of it. That looks stationary there, I would think.
GERAGOS:
Aren't there parts of it here that are also plotted all the way along?
RODDIS:
Which are moving, possibly.
GERAGOS:
Okay. Doesn't this show at all times on 1 double D that there was never any
motion?
RODDIS:
Yes, that's right.
GERAGOS:
Okay. Wouldn't that be, wouldn't that be a possibility of multi-path?
RODDIS:
No. Remember I said that the motion on that one is simply a vibration detector.
GPS.
GERAGOS:
Well, the vibration, the vibration detector was apparently never activated the
entire time this was going on where the car was driving?
RODDIS:
Possibly. I don't know how it was installed. I don't know anything about where
it's installed in this case. I do know one instance where they installed it
inside somebody's seat and that took all the vibration away totally, but I don't
know, I don't know where this was installed or what it was done, but that's,
that's the case.
GERAGOS:
Okay. The next one we have is the one in Fresno. So we're going through these in
order. The one I just showed you, you don't have any explanation for why there's
no motion. The very next one is the one where it's extremely rare, and you said
that you've only seen this one other time, correct? This reflects the Fresno
occurrence which you said you still don't know?
RODDIS:
Yes.
GERAGOS:
You just posited something?
RODDIS:
Yes.
GERAGOS:
Okay. Let's go to the next. Now, this one also,
JUDGE:
Give me a number, Mr. Geragos.
GERAGOS:
Excuse me?
JUDGE:
When you say "this" one,
GERAGOS:
Yeah, I was going to.
JUDGE:
Define,
GERAGOS:
1 double J. Now, this also shows the car going from a hundred and five miles an
hour up to 2300 and back down again?
RODDIS:
Well, it's the same spike we've talked about.
GERAGOS:
As this? The same spike as the one I just showed you?
RODDIS:
I believe so. Which one are you showing me?
GERAGOS:
You have this, you have this. Are they 1 JJ and 1 GG, all the same?
RODDIS:
Let me just check.
GERAGOS:
Okay.
RODDIS:
Yeah, it's the same information.
GERAGOS:
Well, then why do you have it plotted here on 1 HH different than you have it
plotted here on 1 double E?
RODDIS:
I think the segments of the same data, not, it's the same data, but
these are segments of it.
GERAGOS:
Well, if they were segments of it, why would you have the same data sets behind
it?
RODDIS:
I don't know. I didn't create the folder, so I don't know.
GERAGOS:
All right. Now, you have one double, 1 WW. Is this also the same data set?
RODDIS:
Yeah.
GERAGOS:
Okay. And this also shows the same spiking on it?
RODDIS:
It does.
GERAGOS:
Okay. And you have 1, let's see, this is 1 double P; is that correct?
RODDIS:
Yeah.
GERAGOS:
There it shows that his car was going 38 thousand 800 and 84 miles per hour with
no motion; is that correct?
RODDIS:
No.
GERAGOS:
Then that same unit apparently had a, what we've identified here as the bitmap
two track name Rover problem as well, correct? It's the subject of both letters,
right?
RODDIS:
Was that the same unit?
GERAGOS:
Isn't that what you just said? Bitmap two is the same unit?
RODDIS:
The, the unit that had the, the spike on it that we just talked about with the
radar, and whatever, I think was a different unit.
GERAGOS:
I think you're correct. I think you had identified the unit that was
inoperational, I mean was not operating from January 3rd to the 21st as being
the same one as bitmap two; is that, is that your testimony?
RODDIS:
Yes. Serial number 7476, yes.
GERAGOS:
Okay. That's also the same unit that you say when it was returned to you some
time in May had a microprocessor or pin error; isn't that what you said?
RODDIS:
Yes.
GERAGOS:
Okay. So is it a fair statement that there were three problems with that unit in
the span of five months?
RODDIS:
I'm not sure what the third problem was.
GERAGOS:
Well, we have a bitmap two issue here of some kind that required explanation, in
both letters?
RODDIS:
It has nothing to do with the unit. That was just the maps. The maps the unit
displayed on. Had nothing to do with the unit.
GERAGOS:
How about the data sets that were produced?
RODDIS:
I don't think,
GERAGOS:
Were there problems with the data sets?
RODDIS:
They were fine, I think.
GERAGOS:
Wasn't the whole, wasn't the whole problem that you discussed that the data sets
had a problem because of the, the maps that were used, that they did not plot it
correctly?
RODDIS:
Well, I think you would, the, the, the map that the data is displayed on has
nothing to do with the unit. So, I mean, you could take the data out of the unit
and display it on a blank sheet of paper if you wanted to. I, I think that it's
wrong to say that that unit had a problem with its maps. The map is something
totally separate from the unit.
GERAGOS:
Well, the, what good is the unit if it's not able to display it onto a correct
map? It's useless, isn't it? I mean you, you can't plot, the whole idea of the
unit is to plot a location. So if it's not plotting correctly on a map, what
good is the unit?
RODDIS:
I'm sorry, I don't understand that. Because if the unit can be producing
perfectly good data, and yet if you use the wrong map with it, then it's not
going to display properly. But it's nothing wrong with the unit, it's the map
that's the problem.
GERAGOS:
And your explanation for that, or the Orion's explanation, was that it had
something to do with the maps, correct?
RODDIS:
Yes.
GERAGOS:
Are you aware that Dr. Loomis disagreed with that explanation?
DISTASO: Objection, your Honor. Misstates his testimony.
RODDIS:
I doubt,
JUDGE:
Hold on.
RODDIS:
he did, but if he did,
JUDGE:
Hole on, Doctor.
GERAGOS:
I'll give him the page and line number.
JUDGE:
All right.
GERAGOS:
The, the units, at the very least this unit, which you identified by the serial
number, had failed or had a problem twice within a five month period, correct?
If you don't include the mapping problem.
RODDIS:
It had a problem with an antenna and had a subsequent problem with it, something
internally, yes.
GERAGOS:
Well, somebody told you it had a problem with an antenna?
RODDIS:
Yes; that's right, absolutely.
GERAGOS:
And it had a microprocessing unit or pin error, correct?
RODDIS:
That we know about.
GERAGOS:
And the third error, which required two separate letters to try to analyze,
you've laid off on the maps; isn't that correct?
RODDIS:
Yes.
GERAGOS:
Okay. Now, the GPS, the system itself, is based upon in large part things that
are called pseudo ranges; is that correct? Do you know what a pseudo range is?
RODDIS:
I do.
GERAGOS:
What is a pseudo range?
RODDIS:
I'm not sure why in is important to this case, though. It's an internal
calculation that's done to, to, in regards to the satellite locations in
distances and times from them.
GERAGOS:
And isn't that part of the methodology that's used for a GPS device?
RODDIS:
Oh, absolutely, but it's,
GERAGOS:
That's why it's important to this case.
JUDGE:
Let him finish his answer, Mr. Geragos.
GERAGOS:
Okay.
JUDGE:
He's answering your question here. You said "absolutely," then he stepped on
your answer.
RODDIS:
Yes. Those, those are partial, those are, how shall I say it, intermediate steps
in the calculation of latitude and longitude, which is carried out within the
Trimble receiver. Our receiver doesn't see any of those. Our, our equipment
doesn't actually see any of those calculations.
GERAGOS:
But isn't that part of the methodology of GPS positioning?
RODDIS:
Absolutely.
GERAGOS:
Okay. Now, does multi-path or multi-path results, can that result in a, produce
multi, the doctrine, if you will, or the, the incidents of multi-path, can that
have an impact on a true pseudo range?
RODDIS:
Yes, it could, but the equipment is designed to reject those things.
GERAGOS:
And specifically the equipment is designed to, you said, correct that?
RODDIS:
Reject.
JUDGE:
Reject.
GERAGOS:
Reject that. And that's when the condition is specifically where the antenna is
on the top of the car roof; isn't that correct?
RODDIS:
No. I think it would certainly apply when the antenna is on the car roof, but
certainly the equipment does the rejections and the calculations wherever the
antenna is.
GERAGOS:
Doesn't the, doesn't it read in multi-path result under longer than true pseudo
ranges when the antenna is not in an optimal place?
RODDIS:
I think that multi-path is more likely if the antenna is not in place,
GERAGOS:
If you,
RODDIS:
not in the best place, yes.
GERAGOS:
Right. And if your multi-path results, then you get a longer than true pseudo
range, correct?
RODDIS:
Potentially, yes.
GERAGOS:
And if you have a longer than true pseudo range being measured between the
antenna and the satellite, what happens?
RODDIS:
You'd get an incorrect location.
GERAGOS:
Okay. And that condition would occur more often than not when the antenna is not
in plain view on the top of the car roof; isn't that correct?
RODDIS:
I would say it's more likely to happen, but I wouldn't say it happened more
often than not, no.
GERAGOS:
Okay. It's more likely to happen when the antenna is covertly placed, correct?
RODDIS:
Yes.
GERAGOS:
Now, in this case the, you have said that the antenna was bad from, that's what
you were told, between January 3rd and January 21st, correct?
RODDIS:
Not the GPS antenna.
GERAGOS:
Which antenna?
RODDIS:
The wireless antenna.
GERAGOS:
Okay. And do you know where the wireless antenna was placed?
RODDIS:
No, but sometime in that period. I don't know, I'm not sure I remember.
GERAGOS:
Do you remember, do you know where the GPS antenna was placed?
RODDIS:
No. I'm sorry. Did you ask if the wireless antenna was placed when?
GERAGOS:
Yes. Do you know when it was?
RODDIS:
No, I don't, and I don't know where the GPS antenna was either.
GERAGOS:
Okay. And wouldn't that be critical to determining whether or not those
multi-path errors or multi-path occurring in, for you to make a scientific
determination as to whether or not you were getting longer than true pseudo
ranges?
RODDIS:
What happens when you get bad pseudo ranges, a lot of multi-path is you get
erratic data. The tracks that I have seen that I have here don't show evidence
of that erratic data, except in some very minor instances, which I find actually
more interesting than not.
GERAGOS:
Well, when you say that there is an erratic data, we have situations, here I'll
just pick the first one on 1 GG. We have fixes that are occurring here where the
speed of the car is ranging from a hundred and five miles per hour up to 2300
and 23 miles per hour, and going back down to 97; is that correct?
RODDIS:
Yeah. That's what it says. Now, let me just have a look where this is and what
it looks like on the map.
GERAGOS:
Okay.
RODDIS:
23. That's this one. That is the one where the radar stopped going, which we've
explained as what we think happened there.
GERAGOS:
The, the data that's here, you say self-corrected for how long?
RODDIS:
No, well, I'm not sure if, well, if you look at it on the chart here, it came
back in here, wherever that was.
GERAGOS:
You're pointing to 1 double E?
RODDIS:
Right. And you can see that's the spike where the radar started to go wrong up
here, where it started to hit, say, several hundred miles an hour. You remember
I said that if the speed is just calculated by the space between the dips. When
the spikes started to go off in the wrong direction, the speeds were calculated
to be very, very high because the positions were wrong during that one spike
that we discussed several times this morning.
GERAGOS:
Now, the, you, that's the only explanation you can find for that? Or you have
actually no explanation for that other than there may have been radio or radar
interference, correct?
RODDIS:
Yes, that's right. But the rest of the tracks doesn't show a lot of incidents of
multi-path or...
GERAGOS:
Let me show you 1 double D, which was shown to you this morning. What happened
in this instance? Which is January 9th also, but in the morning.
RODDIS:
I'm sorry, what do you mean?
GERAGOS:
Well, where is the, where is the miles per hour, where is the longitude, where
is the latitude?
RODDIS:
Well, the longitude and the latitude there, I don't know. This particular page
has been printed out oddly. It's been printed out with a column on top of
another column, but the, the speed is there, but it's the way it's been printed
out. I don't know why it was printed out that way.
GERAGOS:
Well, do you have an explanation as you sit here as to why 1 double D appears to
have two columns collapsed?
RODDIS:
No. No. No, something to do with the printer and the computer.
GERAGOS:
Well, does 1 double D refer to 1 BB?
RODDIS:
I, I don't know. I think so, probably.
GERAGOS:
Could you take a look at that and tell me if it does?
RODDIS:
Yeah. Yep. It seems to.
GERAGOS:
Okay.
RODDIS:
This one section of that does.
GERAGOS:
Fair to say that 1 double B indicates that the object that the antenna and/or
the receiver had been placed on was moving?
RODDIS:
Well, parts of it. That looks stationary there, I would think.
GERAGOS:
Aren't there parts of it here that are also plotted all the way along?
RODDIS:
Which are moving, possibly.
GERAGOS:
Okay. Doesn't this show at all times on 1 double D that there was never any
motion?
RODDIS:
Yes, that's right.
GERAGOS:
Okay. Wouldn't that be, wouldn't that be a possibility of multi-path?
RODDIS:
No. Remember I said that the motion on that one is simply a vibration detector.
GPS.
GERAGOS:
Well, the vibration, the vibration detector was apparently never activated the
entire time this was going on where the car was driving?
RODDIS:
Possibly. I don't know how it was installed. I don't know anything about where
it's installed in this case. I do know one instance where they installed it
inside somebody's seat and that took all the vibration away totally, but I don't
know, I don't know where this was installed or what it was done, but that's,
that's the case.
GERAGOS:
Okay. The next one we have is the one in Fresno. So we're going through these in
order. The one I just showed you, you don't have any explanation for why there's
no motion. The very next one is the one where it's extremely rare, and you said
that you've only seen this one other time, correct? This reflects the Fresno
occurrence which you said you still don't know?
RODDIS:
Yes.
GERAGOS:
You just posited something?
RODDIS:
Yes.
GERAGOS:
Okay. Let's go to the next. Now, this one also,
JUDGE:
Give me a number, Mr. Geragos.
GERAGOS:
Excuse me?
JUDGE:
When you say "this" one,
GERAGOS:
Yeah, I was going to.
JUDGE:
Define,
GERAGOS:
1 double J. Now, this also shows the car going from a hundred and five miles an
hour up to 2300 and back down again?
RODDIS:
Well, it's the same spike we've talked about.
GERAGOS:
As this? The same spike as the one I just showed you?
RODDIS:
I believe so. Which one are you showing me?
GERAGOS:
You have this, you have this. Are they 1 JJ and 1 GG, all the same?
RODDIS:
Let me just check.
GERAGOS:
Okay.
RODDIS:
Yeah, it's the same information.
GERAGOS:
Well, then why do you have it plotted here on 1 HH different than you have it
plotted here on 1 double E?
RODDIS:
I think the segments of the same data, not, it's the same data, but these are
segments of it.
GERAGOS:
Well, if they were segments of it, why would you have the same data sets behind
it?
RODDIS:
I don't know. I didn't create the folder, so I don't know.
GERAGOS:
All right. Now, you have one double, 1 WW. Is this also the same data set?
RODDIS:
Yeah.
GERAGOS:
Okay. And this also shows the same spiking on it?
RODDIS:
It does.
GERAGOS:
Okay. And you have 1, let's see, this is 1 double P; is that correct?
RODDIS:
Yeah.
GERAGOS:
There it shows that his car was going 38 thousand 800 and 84 miles per hour with
no motion; is that correct?
RODDIS:
It's very clever.
GERAGOS:
Yeah.
RODDIS:
No, that was that one degree spike. One degree spike that we talked about and
changed. And so when the thing jumped one degree, then it looked as though it
was doing thousands of miles an hour. But it jumped, in that case it was just
that one thing. Yeah.
GERAGOS:
Now, this data set shows that for a period of time there was no movement
whatsoever; isn't that correct? Before and after?
RODDIS:
Yes. It was parked, I believe. Or stationary.
GERAGOS:
If he was parked, let me ask you a question. On 1 double P, you're saying that
this spike took place because he had finally, or the latest explanation is,
according to Exhibit C, that they crossed an area of a longitudinal line?
RODDIS:
Yeah.
GERAGOS:
Or latitudinal line?
RODDIS:
Yes, that's right.
GERAGOS:
Well, here,
RODDIS:
If he was,
GERAGOS:
that would seem to indicate from your data set that isn't what occurred; isn't
it?
RODDIS:
No, that is what occurred.
GERAGOS:
Well, then why isn't there movement before and after? Didn't you expect to see
the car moving into that location as opposed to being parked as you just
testified to?
RODDIS:
No, this, this is, it's neat, in fact, from a scientific point of view, if not
from a legal one. He was parked exactly on, on a line of longitude, so the
statistical variation of the fix, as you said, you don't get a dot, you get a
little blurred thing. So statistically at one reading there he jumped over that
line, and that's what caused the problem.
GERAGOS:
Well, if it was a firm, you called it before I think a firmware error from
Trimble? The latest explanation for this was that it was Trimble's error?
RODDIS:
Yes.
GERAGOS:
Why didn't it then occur at the first time when he hit the longitudinal line, is
what you posited is the explanation for this?
RODDIS:
No, that's not true.
GERAGOS:
Well, isn't there,
DISTASO: Objection, your Honor.
JUDGE:
Let him answer the question, Mr. Geragos.
GERAGOS:
Sure.
JUDGE:
Go ahead.
RODDIS:
When it's parked, I already said, and I don't know what Dr. Loomis said, but the
error from the Trimble receiver just comes out once, once as it passes over the
line. We don't take all of the calculated readings from the Trimble receivers.
We only take them when we need them. So in this case this, it looks as though
there are every five seconds apart, so we would have had four or five readings
that we would have ignored. So when he first came in possibly we ignored it. But
when he was sitting there, on a parked spot, in a statistical variation, at some
point we hit one that we didn't, that we picked up because it happened to be
right, and then that made us jump. But we don't,
GERAGOS:
Wait.
RODDIS:
take a reading,
GERAGOS:
You're taking, there's how many readings prior to that jump? There's at least 15
readings?
RODDIS:
Sure.
GERAGOS:
Well, then what are you saying? That it just happens on the 16th reading? It
just so happens he parks on the exact longitudinal line and all of a sudden the
16th time you take a reading all of a sudden it jumps 38 thousand miles per hour
and then comes back on the next reading? You know that doesn't make any sense.
RODDIS:
No, it does make sense,
JUDGE:
That's argumentative. You can answer the question.
DISTASO:
No, Objection, your Honor.
GERAGOS:
Let me ask,
JUDGE:
No, he can answer. Do you have an answer to that, Mr. Roddis?
RODDIS:
I'm just trying to remember where I was. But, yes, I think, I do have an answer,
and it is statistics. I say it's statistical variations, and he's sitting there
parked, we're getting random readings around the thing. I don't know which
readings we hit. We're picking every fifth reading out. I'm not at all surprised
there was only one of those that caused this error. It could have been, if he
had been parked two feet in the other direction, it might have been zero.
GERAGOS:
Isn't this, when you say it's taking random locations,
RODDIS:
I didn't say random.
GERAGOS:
Isn't that what you said? You just testified that it was taking random locations
around,
RODDIS:
No, I said there was statistical variations. If you look at it, 18, 18, 18, 17,
15, 15, 15, then jumps over here, goes to five zero, and this, even these are
the, jumps on the other side of the line and jumps back down here. So we've got
nine point nine nine nine here, and point nine nine nine five here,
GERAGOS:
You still haven't explained to me why there are one, two, three, four, five,
six, seven, eight, nine, ten, 11, 12, 13, 14, 15, 16, 17, 18, 19, readings prior
to, and when it's supposedly parked because it shows no motion as well, and zero
miles per hour, there's 19 readings, which are just a subset of probably, what,
a hundred and 50 readings that could have taken place during that time period?
RODDIS:
Sure.
GERAGOS:
And out of that it just so happened out of the hundred and fifty-first reading
it takes off at a hundred and 38 thousand miles per hour, or you have this one
degree,
RODDIS:
It depends, he's parked near the edge of this line. It depends where that
statistical number goes over the line. I have no idea. It could happen once or
ten times, but you can't tell me either.
GERAGOS:
Well, right, I can't tell you, but you can't tell me. But we're sitting here in
a court of law saying we want to admit this as forensic evidence. DISTASO:
Objection, your Honor. Argumentative.
RODDIS:
I have,
JUDGE:
Again argumentative.
GERAGOS:
Would you go,
JUDGE:
Question and answer.
GERAGOS:
You have two explanations, which have been, one that you conceded now was
incorrect, which was reflected in Exhibit B. You then come up with another
explanation in January of this year, which is reflected in Exhibit C. Did
anybody go out and take this receiver and go park it on a longitudinal line and
see if they could recreate this occurrence?
RODDIS:
Yes, they did.
GERAGOS:
And did it happen?
RODDIS:
Yes, it did.
GERAGOS:
When?
RODDIS:
When?
GERAGOS:
When?
RODDIS:
I'm not too sure. We did it in Nova Scotia because our plant happens to be 45
degrees latitude, almost exactly. We just went down the road and tried it. We
waited there long enough, and it didn't happen immediately, but we waited there
long enough and it did happen eventually, yes.
GERAGOS:
Did you use a new system to see if you could get rid of that?
RODDIS:
We corrected the firmware after, subsequently, to that, adding in a way of
ignoring the A, if you like; and yes, it did get removed.
GERAGOS:
Now, here's 1 double P where we showed you before, at 5:53 there was this jump
that shows by 38 thousand miles per hour, right?
RODDIS:
Yes.
GERAGOS:
And that GPS fix at that point was at one two one point nine nine nine nine
eight, correct?
RODDIS:
Right.
GERAGOS:
Now, if we go to 1 double T, we show a GPS fix at 6:00 o'clock?
RODDIS:
Correct.
GERAGOS:
What do we show there?
RODDIS:
Another jump.
GERAGOS:
38 thousand 800 and 80 miles per hour?
RODDIS:
Exactly.
GERAGOS:
Okay. And where are we then on the longitudinal line?
RODDIS:
Let me just have a look over here a second. There. This is the one you're
looking at.
GERAGOS:
No.
RODDIS:
I'm sorry,
GERAGOS:
5:53?
RODDIS:
Let me just look at this for a second. This is the one where it jumps for four
minutes, or whatever it is, completely. So that's the first speed jump that you
see is the jump of the one minute error, and the second high speed one is where
it jumps back again. And if you look at it, it occurs at exactly on the hour,
which is where the compression system rolls over, six point zero zero zero zero
zero p.m.
GERAGOS:
That's from 5:53?
RODDIS:
Right.
GERAGOS:
When you have one jump?
RODDIS:
Yeah.
GERAGOS:
Now, in the interim, after 5:53, we have a location that shows that this car is
moving, do we not?
RODDIS:
It is moving.
GERAGOS:
And we have the car,
RODDIS:
I'm sorry, we're talking about two different things here. Originally we were
talking about the one spike on the parking lot.
GERAGOS:
Where on,
RODDIS:
Now, we're talking about a different issue.
GERAGOS:
We're, we're talking about the fact that you have got an explanation for why
something occurs. Your explanation is that when it's on the longitudinal line,
all of a sudden it jumps; isn't that correct?
RODDIS:
Yes.
GERAGOS:
Okay. Now, what I'm showing you is that here at, at 5:53, your explanation is
that the car is parked and then all of a sudden it jumps, correct?
RODDIS:
Correct.
GERAGOS:
Then I'm showing you on 1 double T that at 6:00 o'clock, which you testified to
before, you're saying that at that point it jumped back; is that correct?
RODDIS:
Yes, except for the fact that during that period the car was in motion. It was
moving. You can see it on the track here. It was in motion. Now,
GERAGOS:
How does that,
DISTASO: Objection.
JUDGE:
He hasn't finished his answer, Mr. Geragos. Go ahead.
RODDIS:
I'm sorry?
JUDGE:
No, finish your answer.
RODDIS:
I'm getting confused by all of this; but, truthfully, we have two different
incidences. One incidence where it was parked and it jumped. It was parked on
the line. There was a second instance of a different place where it jumped while
the vehicle was in motion. And this one that you're talking about now is the
second one where the vehicle was in motion when it jumped.
GERAGOS:
You just testified that the, were you confused the last time you testified when
you said that that jump that happened at 6:00 o'clock was the machine correcting
itself?
RODDIS:
No. That is correct.
GERAGOS:
Okay. So you're saying that the car was parked, it just so happened to be parked
--
RODDIS:
No, I didn't say it was parked,
GERAGOS:
right on the longitudinal line?
JUDGE:
Wait, wait.
RODDIS:
I didn't say it was parked,
JUDGE:
Let him finish his question, Mr. Geragos.
GERAGOS:
I'm sorry. Didn't we go through that there were 19 readings out of some 150
readings that were compressed on the data that showed that the car was parked on
the longitudinal line, correct? At 5:53 and change?
RODDIS:
Oh, I'm sorry. Yes, you're quite right, it is. That's absolutely true; that's
correct.
GERAGOS:
Right. And you testified it was parked?
RODDIS:
Yes, that's true. You're quite right.
GERAGOS:
Then you're saying it corrected itself at 6:00 o'clock, lo and behold, and the
car is moving?
RODDIS:
Yes, because during that period of time the vehicle did start moving. And if you
look at it you can see the speed 20 miles an hour, 30 miles an hour during that
whole period of time, until it came back to the 6:00 o'clock where the
compression fixed itself, and it came back, jumped back again. But during that
whole period of time, oh, up to right at the end it obviously stopped again.
Right at the end, the last few readings it was stationary again.
GERAGOS:
Wait; can you just tell me one thing? If the car was parked and then all of a
sudden there was a one degree jump that took you however many miles away, 60
miles away, 69 miles away, whatever it is, then why are you relying on the
machine if you're saying it corrected itself at 6:00 o'clock? You're now
testifying that the data between 5:53 and 6:00 o'clock was accurate when you've
just said that the machine didn't correct itself until 6:00 o'clock, correct?
RODDIS:
I, depends what you mean by "accurate." It was one degree off. The rest of the
decimal places were accurate. It was reading one degree off. But the rest of the
decimal places were correct. And those things showed that the vehicle was in
motion. And, in fact, if you take that move track and put it back one degree, it
overlays roads and things correctly.
<noon recess>
GERAGOS:
Before lunch we had gone through 1 double P, which was the jump where that
38,884 miles per hour had showed up; is that correct? Then we had also gone to 1
double T, which also shows 38,880 at 6:00 p.m.; is that correct?
RODDIS:
Yes.
GERAGOS:
Then we had gone to the next one in order that's in the book would be 1 double
X. Now, 1 double X, do you see anything here on the reported data that appears
to be out of line?
RODDIS:
No.
GERAGOS:
Now, does that appear to you that corresponds to 1 double U?
RODDIS:
No, I don't think it does, if you look at the times.
GERAGOS:
Does it appear to you that the information in 1 double X corresponds to 1-VV?
RODDIS:
I think that's a portion of it, yes.
GERAGOS:
And when you say the times are wrong between 1 double X to 1-UU, you are talking
about on 1-UU the times that have been placed on the various strips on the
exhibit itself?
RODDIS:
Yes.
GERAGOS:
That does not appear to you to correspond to 1-XX?
RODDIS:
I don't think so. I'll check it again fo you. But this one, okay. Hang on a
second. I'm sorry, I have got four hours jet lag catching up. So I have got
45410. 454, no, I don't. There may be a portion of it here. This track, this
1-UU has a lot of data on it. But certainly most of it is not represented in
this, in this table. If I look at it carefully, maybe.
GERAGOS:
So what you are saying is the 1-XX, the 1-UU, do not appear to track; is that
correct?
RODDIS:
Maybe a portion of it, but certainly not all of it, no.
GERAGOS:
Okay. Now, let me move you over to 1 triple A. Does there appear
to
be any problem with 1 triple A, that data that's printed up?
RODDIS:
Not per se, no.
GERAGOS:
Does it appear that this car or the object is moving?
RODDIS:
No.
GERAGOS:
And it's reflected in this 1 triple A that the, whatever object the device is
attached to, not only is there no motion, but there appears to be no miles per
hour except save a couple of entries for one mile an hour, correct?
RODDIS:
Yes. I'm just checking down through this. The last latitude and longitude, some
small variation in the latitude and longitude, not very,
GERAGOS: When you
say there is a small variation in latitude and longitude, that exhibit itself
shows that there appears to be a different latitudinal lock from the GPS device,
doesn't it? As you are moving all the way down this column, that's the fourth
column over?
RODDIS:
Down to this point here where there is a change in this area.
GERAGOS:
The change is when it says unit power up; is that correct?
RODDIS:
Exactly.
GERAGOS:
It is maybe three quarters to seven eighths of the data that's been printed out
shows that there is no motion, that there is no movement of the car, but it does
show slight variations of the latitudinal fix; isn't that correct?
RODDIS:
That's correct.
GERAGOS:
It also tracks differences in the longitudinal fix; isn't that correct?
RODDIS:
They are very small. If you look at it in the fifth decimal place, which is the,
fourth decimal place, which is quite a small number.
GERAGOS:
Now, have you compared the data sets on the earlier diagrams that I showed you,
which is 1-PP, where you are at a longitudinal fix that you said was right at
the precise line, the longitudinal line with other data sets to see if it
recreated
,
this problem recreated itself?
RODDIS:
Not with this set, because I couldn't do it. But I have done it with, or I had,
personally haven't done it. But at the company we have done it, as I said
earlier.
GERAGOS:
That was in,
RODDIS:
In Nova Scotia.
GERAGOS:
I'm asking with this particular data set, have you looked through the, there is
some 3,000 pages worth of data set material that's been produced by the devices
in connection with this case. You are aware of that?
RODDIS:
No.
GERAGOS:
Have you looked through any other data set in connection,
RODDIS:
The only data set is what you have in front of you.
GERAGOS:
Okay. Have you looked in the front of the, through the data set that we have in
front to see if this same phenomenon occurs or didn't occur when you had the
same longitudinal and latitudinal fix as you had on 1 double P?
RODDIS:
I haven't. If you look, there is quite a long way on, I don't see that it would
occur. As I said earlier, it's still a statistical issue. Even if it did show in
there, it wouldn't necessarily occur as a problem. Obviously the problem itself
hasn't occurred, but that would show up in the track.
GERAGOS:
Now, the next location here of data set material is 1 triple F. You see a
problem in the data as it's represented there?
RODDIS:
No. Looks as though it's probably on a motorway or something on the freeway.
But, well, other than the issue we mentioned before about the no-motion-motion
we talked to previously.
GERAGOS:
That's precisely what I was going to ask you. From the times of, roughly we're
talking about 8:00 a.m. in the morning through 8:37 a.m. is what's represented
in 1 triple F; is that correct?
RODDIS:
Yes.
GERAGOS:
Okay. And that would include a time period for 8:12; isn't that correct?
RODDIS:
Right.
GERAGOS:
Now, that would appear to be on a freeway; isn't that correct?
RODDIS:
Because the speeds from that 8-12, that obviously there is a lower area there. I
haven't got the map in front of me for during that period. I don't know why.
GERAGOS:
And for how long? How long a period of time?
RODDIS:
Well, drops right down here, up at the first part, 65 miles per hour, then even
drops right down to 20, 20, 16, 25. And there is low levels. 37. It speeds back
up again to highway speeds. So it's over a fair period. 8:05 down to higher than
that, 8:03, down to, which would be 8:20, or something.
GERAGOS:
8:17, or 8:17 shows up at 67 miles per hour, correct?
RODDIS:
Okay.
GERAGOS:
Now, the next document that's in here is 1 triple I. Do you see that?
RODDIS:
Yes.
GERAGOS:
This shows, this shows no movement whatsoever; is that correct?
RODDIS:
True.
GERAGOS:
Okay. And that shows you from 1:12 to 2:18 with no movement, correct?
RODDIS:
True.
GERAGOS:
1 triple G shows you a map that's plotted from 2:18, and shows you with movement
as plotted on the map; is that correct?
RODDIS:
Right. Except it doesn't tell me that at this point, which does relate to what.
So this, for example, at this point where he's stationary there is a lots of
places there. There is no other reference, I couldn't relate that to any
particular plot.
GERAGOS:
That shows you 2:18 on the exhibit that's marked as 1 triple G, that there is
movement, and there is little arrows here pointing in the direction; is that
correct?
RODDIS:
No. No. Not necessarily. That looks like he is stationary there. It's hard to
tell from that, without actually taking it and looking at which record is when
they are all on top of each other.
GERAGOS:
Is it a fair statement that, based upon the data set 1 triple I that's been
provided, the map that is 1 triple G, you can't make any determination as to
whether or not that's accurate, because there are points that are plotted that
are not reflected in the data set.
RODDIS:
There are points that are not the points are plotted on the map. It's just at
the time that, the time and date are not given against them.
GERAGOS:
If the time and date is not given against them, there is no way for you to judge
whether or not it's operating correctly; isn't that a fair statement?
RODDIS:
On the basis of this file, yes. But I did have, when I say the data we have
given us, we were also given a floppy disc with the information on it so we
could replay this one step at a time.
GERAGOS:
But you didn't bring that to court?
RODDIS:
No. What we did was to check what we saw was the same as what was in the set.
GERAGOS:
This is 1 triple W. Do you see that?
RODDIS:
Yes.
GERAGOS:
1 triple W, do you see some problems with this?
RODDIS:
Yes.
GERAGOS:
And that shows that on 1-27 at approximately 11:11 that he was going 19,442
miles per hour, correct?
RODDIS:
That's right.
GERAGOS:
And it appears that the car was stationary before that; isn't that correct?
RODDIS:
Yes. Same issue we talked about prior to lunch.
GERAGOS:
Well, and then once again, at approximately 11:12, about a minute later, it
shows that the car is now going 6,480 miles per hour; is that correct?
RODDIS:
That's right. That's a one-minute jump.
GERAGOS:
So we have had a one-minute jump, and with, that would that be reflected also on
1 triple J. Do you see that plotted out?
RODDIS:
Goes from there to there.
GERAGOS:
Okay. Now, and do you see that in 1 triple K?
RODDIS:
Yes, that's, yes.
GERAGOS:
Okay. Now, that, how far of a distance is that, is that jump? Can you tell by
the GPS coordinates what the difference is between this jump at 19,442 miles an
hour down to 6,480?
RODDIS:
No. I would have to calculate it. I haven't got it calculated here. It will be
roughly somewhere around fifty miles, I guess, latitude.
GERAGOS:
Now, here is 1 triple P, which appears to be the same data set, correct?
RODDIS:
Yes.
GERAGOS:
This shows basically, doesn't it, that the, what, appears that the car is just
standing still, and that all of a sudden the GPS detects it as moving somewhere
in the neighborhood of 60 miles, then returning. And the car does not appear to
be moved at any time, at least according to the data set for the next ten
minutes. Just appears the car is sitting, all of a sudden it just, the GPS has
it speeding up to 19 or almost 20,000 miles per hour, then coming back, correct?
RODDIS:
Yes.
GERAGOS:
And, in addition to that, there is absolutely no motion generated on the motion
detector, correct?
RODDIS:
Which is probably reasonable.
GERAGOS:
Indicate to you the car is standing still, not moving at 19,000 miles per hour?
RODDIS:
No. I think that's what we explained this morning.
GERAGOS:
Now, you also have an another data set that's in here. That is 1 triple X. Do
you see that?
RODDIS:
Yes.
GERAGOS:
Now, based upon the speeds that you would see here, where would you expect to
see this car on 1 triple X?
RODDIS:
1 triple X has several segments to it. Different segments where he is going at
different speeds. Some areas where he's going 60 miles an hour, that some of it
where he drops right down to slow speeds. Another thing, back again, I think
it's probably a mixed roadway of some kind?
GERAGOS:
Is it safe to say that the between the time of 2:54, and 2:57, for a two-minute
period of time, he was at all times going over, roughly over fifty miles per
hour, except, with a couple of exceptions?
RODDIS:
Yes.
GERAGOS:
Now, the, before lunch I asked you about the, some of problems with the GPS
technology. Do you remember that? Multipathing, one?
RODDIS:
Yes.
GERAGOS:
And poor geometry being another?
RODDIS:
Yes.
GERAGOS:
Now, the poor geometry, could you explain to the Court why that would present a
problem with the methodology, DISTASO: Objection, your Honor. It's been
asked and answered.
GERAGOS:
It has not been asked.
JUDGE:
We have been talking about maps. Again, I think he covered that this morning.
This is different?
GERAGOS:
Yes, this is different.
JUDGE:
You can answer.
GERAGOS:
Poor geometry is not a function of maps, is it?
RODDIS:
No.
GERAGOS:
Okay. Poor geometry has to do with the device locking on to the satellite, does
it not?
RODDIS:
Yes.
GERAGOS:
Now, would you explain to the Court what the problem is involving poor geometry,
or what potentially is the problem involving poor geometry?
RODDIS:
Yes. This is a problem that was a lot worse a few years ago. Now we have a lot
more satellites in the sky. It's a lot less of a problem.
RODDIS:
What happens is, in order to measure the position on the ground, the unit has to
measure the distance to a range of satellites. Now, if you can imagine, if those
satellites were vertically right ahead above, then there would be no difference
in the distances. If they are a sitting beside each other, the calculation is
impossible. Therefore, for a good geometry, you need to have satellites
scattered at different places around the sky, more widely spaced, in a sense it
is better. However, it doesn't mean to say that you can't get a location fix
even with a poor geometry. What it means is that it's liable to be more accurate
if you have a better geometry. In other words, the satellites are further apart
in the sky when you look at them.
GERAGOS:
We talked before lunch about pseudoranges?
RODDIS:
Yes.
GERAGOS:
And pseudoranges are part of how one calculates, or how the methodology is
determined to calculate a position, correct?
RODDIS:
Yes.
GERAGOS:
Pseudoranges can be badly measured, depending on what the geometry is and
whether the geometry is accurate or not, correct?
RODDIS:
Yes, that's basically what I just described.
GERAGOS:
And so one of the things that ideally you want to have is a satellite in each
quadrant of the sky so that you can lock on to hopefully three and maybe four of
the satellites so that you can position, you can get an accurate position,
correct?
RODDIS:
Correct.
GERAGOS:
If you have the multipath situation, and if you have poor geometry, then that
means that the position that's computed is badly measuring the pseudoranges;
isn't that correct? If you have both of those present?
RODDIS:
If that were true, yes.
GERAGOS:
Now, isn't it a fact that you have both of those, multipath situations and a bad
geometry situation, bound to occur when an antenna is hidden on a vehicle, isn't
that one of the problems with hiding the antenna on the vehicle?
RODDIS:
No. Well, you said isn't it bound to occur. And the answer is, no, it's not
bound to occur. It's more likely to occur depending on how the antenna is
hidden. However, you must understand that there are normally about eight
satellites in the sky scattered in the sky, so the unit has a choice. It can
choose. It only needs three or four of those. So it can pick the ones that are
the best position in the sky to use. And so if the antenna is in a poor
position, restricts a view of the sky, what you are doing is limiting the number
of the satellites to choose from. It still can choose from which ones it feels
to get the best reading. The second issue is, that as to, Doctor Loomis may or
nay not have told you, the Trimble receiver actually calculates this PDOP issue,
and makes a decision about whether or not to store a fix or to record a fix if
the geometry is bad. So, therefore, if there is a really bad geometry, the unit
just won't record the position. It's not recorded. It's ignored. It's thrown
away.
GERAGOS:
Doctor Loomis did testify that what they provide is a receiver, and basically
the receiver is then placed into your unit. Your unit then has the software
package, if you will, or that it overlays the receiver, so that the data sets
that are produced are a function of your unit; isn't that correct?
RODDIS:
Well, no. There are two levels. The first the, Doctor Trimble, or Doctor
Trimble. Doctor Loomis's equipment or Trimble's equipment does the first level
of processing in which it decides all those things. And then other equipment
does the second level of processing, which is taking the information after those
decisions have been made.
GERAGOS:
And so if there are decisions that are made prior to getting to your overlay, if
you will, or do you have another?
RODDIS:
Exactly.
GERAGOS:
Overlay of your program, which makes it uniquely an Orion product; is that
correct?
RODDIS:
That's correct.
GERAGOS:
If a PDOP, which is the acronym for the Positional Dilution Of Precision, if you
have a bad PDOP, your overlay is, garbage in garbage out, you are not going to
know; isn't that correct?
RODDIS:
If it's, no. Well, there are two different issues. The first level is that the
Trimble unit will initially reject any really bad ones. If there is marginal
ones, it reports the fact that it has marginal reading, and then we make a
decision based on that. There are three levels of output of accuracy, and we
only accept the best level that it puts out. So, to that extent, I think that
the situation is, is that the unit fixes it. What we store has been filtered to
remove the bad ones.
GERAGOS:.
Is that done by Trimble, or by your overlay?
RODDIS:
Combination of both. The first, the majority of that is done by the Trimble
unit, but we then take some of the data from the Trimble and decide on that
basis. But theirs does most of that.
GERAGOS:
They do most of that. It's one of the reasons the exhibits that I was showing
you, that you had alternate explanations back in July as to what was occurring,
because some of the calculations that were going on, or some of the firmware
errors that were occurring, were from the Trimble unit as opposed to the
overlay, or the Orion --
DISTASO:
Objection, your Honor. Been asked and answered.
JUDGE:
I think it has. But can you give us one more answer to that, maybe get it all
out?
RODDIS:
I think what you are saying is, though, I think I'm not quite sure what you are
saying. The two issues are separate in my mind.
GERAGOS:
Well, the Orion unit decompresses the data, does it not?
RODDIS:
No, that's, the compression issue is related to the way that we send the data
over the air. It's not to do with the actual qualities of the data itself.
GERAGOS:
Well, it does, when you say it doesn't have to do with the actual quality of
the,
RODDIS:
You are going to say it's a one degree jump?
GERAGOS:
Exactly.
RODDIS:
Yes. But the one degree jump is a fixed number, and a fixed, a fixed issue. It's
not the issue of the general quality of the data in terms of where the fixes
were originally stored, or generated, or put in the unit.
GERAGOS:
But that's, but that's what you are saying today, and that has evolved, correct?
I mean it's evolved in the last year. We have had the installation of the
device, we have the data that was recorded. Six months after that we had one
explanation, six months after that we had another explanation, correct?
DISTASO:
Objection. It's been asked and answered.
JUDGE:
Well, these were from two separate explanations as to these anomalies that we're
talking about. In your opinion, based upon everything you have read in the
literature, is there two separate explanations, two types of anomalies, other
than the maps? Let's ignore the maps for the, forgetting about the maps.
RODDIS:
There were two types of anomalies, one of which was an issue we haven't
explained concerning the possibility of radar interference at the airport,
potentially at the airports. And that was one isolated incident that had some,
hasn't affected any of the other data. There was the other instances in the one
degree shift, which is an unfortunate issue. But that relates, that relates, not
to the accuracy of the position in itself, but rather to the transmission of the
data over the air. And I know this sounds a bit difficult to say; but, to me,
bad data would indicate, how do I explain this? I'm sorry, I'm not being very
clear about it. But that's one particular processing problem. And you said that
we had two different explanations of it. Well, frankly, the explanation is the
same, except that one triggered the problem off, was further investigated. But a
one degree shift of good data over so, it's recognizably good, but shifted by
one fixed degree, is not, to my mind, in the same situation as bad data which
would be generated by a bad PDOP, or other things which you would get random
dots around the map, or random places that had no relation to anything.
GERAGOS:
Which is similar to what you described to his Honor happened in Fresno.
RODDIS:
Which in Fresno?
GERAGOS:
In Fresno, meaning the spike that you, we haven't figure figured out what was
going on, except posited the idea it could have been a radar or radio
interference, that just easily could have been bad PDOP, couldn't it?
RODDIS:
No, that's, bad PDOP would never have done that. It's too big.
GERAGOS:
The error that was reflected is too big to have just been attributable to PDOP?
RODDIS:
Because of all the safeguards we built into the system to remove PDOP errors,
yes.
GERAGOS:
Okay. Well, the fact of t he matter is, as we sit here today, we don't have any
explanation as what caused the Fresno situation, if you will, the spike.
RODDIS:
No.
GERAGOS:
And as we sit here today --
JUDGE:
That's the second question. Is that right? There is no explanation, then, as you
sit here to explain this spike in the Fresno data, other than what you have
posited could have been radar interference, or interference from an airport?
RODDIS:
That is true. As I said, my TV signal, you are watching a movie, there is a
burst of interference, you don't know where the interference came from, doesn't
break the rest of the movie bad.
GERAGOS:
You could kind of miss the plot, couldn't you?
DISTASO:
Objection, your Honor. Argumentative.
JUDGE:
Sustained.
GERAGOS:
The bad PDOP, at this point you said there is so many built in safeguards, that
is by Orion, in your overlay, the software has built in safeguards on PDOP?
RODDIS:
No. As I said earlier, it's, there is some small safeguards within the Orion,
but it is, majority of that is built into the Trimble unit. You actually can
set, we can set a level of PDOP if we want to, but we don't actually change. It
we use the Trimble settings.
GERAGOS:
And are you able to tell, based upon these data sets, whether or not the
satellites that were being locked onto by the device were in a single quadrant
or four quadrants in the sky?
RODDIS:
I cannot.
GERAGOS:
Okay. That is the information that one could garner if they wanted, isn't it?
RODDIS:
One could. But, truthfully, the data that you see there is sufficiently
coherent, I would wonder why you would need to.
GERAGOS:
Well, because we have things that have been testified to, at least by you and by
Doctor Loomis, that is either prior, or they have never seen before, or they
have no explanation, and we're in a case that's punishable by death.
RODDIS:
Not that particular isolated incident. That's, you are quite right.
GERAGOS:
We also have the situation where a receiver was not operational, one of these
receivers was, for almost three weeks, and it's been told to you, at that that
was because of an antenna. But if you have not seen the data set as a scientist,
you consider yourself a scientist, correct.
RODDIS:
I'm an engineer.
GERAGOS:
As an engineer, wouldn't you want to see the data set before you just parroted
the explanation? Isn't that what you consider your area of expertise to be, to
be able to look at the data sets and see what was actually going on?
DISTASO:
Objection, your Honor. It's been asked and answered.
JUDGE:
I think so. Given his opinion half a dozen times.
GERAGOS:
I take it you don't want to see more information, you would rather have less or
more?
DISTASO:
Objection, your Honor. Been asked and answered.
JUDGE:
Same objection. Sustained.
GERAGOS:
The positioning of the bad PDOP, is it a fair statement that if the if the
device is picking up satellites out of the same quadrants, that you are going to
have, or it can produce a bad PDOP?
DISTASO:
Objection. Been asked and answered.
JUDGE:
Well, is this the same? Are we back in the same place where we have one
satellite beaming straight down?
RODDIS:
Yes.
JUDGE:
He's answered that question.
GERAGOS:
I'm asking if there is four satellites all in the same quadrant. That's what the
device is picking up on.
RODDIS:
Each satellite being in the same quadrant is not mandatory. I think , but if
they were all very close together in the sky, it would not be a good idea, no.
GERAGOS:
Now, is that a condition where locking onto satellites in the same quadrant
would be more likely to occur because the GPS antenna is hidden?
RODDIS:
I say it again, yes, that is true. But looking at the tracks, I don't see any
evidence of that problem.
GERAGOS:
But you don't know where the antenna was placed, correct?
DISTASO:
Objection. Been asked and and answered.
JUDGE:
He's already answered that several times.
GERAGOS:
Well, can you have a bad pseudorange measurement due to a bad satellite?
RODDIS:
I imagine so, yes.
GERAGOS:
Okay. And can that happen when there is an, I mean most obvious situations would
be if there is a bad clock in the satellite itself?
RODDIS:
Yes, that's true.
GERAGOS:
Are you aware in this case, one of your units was tested by Modesto PD, there
was exactly a one-hour differential between the unit and the actual time?
RODDIS:
I have no knowledge of any tests by the Modesto PD of any type.
GERAGOS:
If I were to give you a copy of something that somebody had done a test on one
of your units, placed it in a car driven in the same locations we have been
looking at, turned out there was a one-hour time differential, would that
indicate to you there was a bad pseudorange measurement?
RODDIS:
No, it would not. It would mostly notice to me that someone forgot to put in
summer time, then forgot to put in the setting was wrong. It's very unlikely to
be a bad pseudorange.
GERAGOS:
What about if they double checked the time when they set it?
RODDIS:
It couldn't be a pseudorange.
GERAGOS:
So what's your explanation for that?
RODDIS:
I don't know anything about it. I haven't seen any results, and I'm not aware of
any problems.
GERAGOS:
Okay. Now, they didn't show, I take it they didn't show you any of that
information as well, correct?
RODDIS:
Correct.
GERAGOS:
Now,
RODDIS:
I don't know that they have done tests or not.
GERAGOS:
Okay. Did you, by the way, did you do any tests in that Fresno area to determine
whether or not you could create that spiking situation?
DISTASO:
Objection, your Honor. Been asked and answered.
JUDGE:
I think he said he did it in Nova Scotia.
GERAGOS:
That was on the longitudinal issue on the one degree.
RODDIS:
On those with the radar, no, I have not.
JUDGE:
You have not. Okay. Go ahead.
GERAGOS:
And isn't it true that the FAA does not allow GPS technology to be used solely
for landing of airplanes?
JUDGE:
I didn't hear what you said.
GERAGOS:
FAA does not allow GPS technology to be used solely for the landing of aircraft?
RODDIS:
That's a very complex question, and I don't think it's fair to answer it. The
answer is, yes, you are right, that they do not allow it solely at this point.
However, it's a complex issue, and a complex problem. It's a lot more to do with
safety than the accuracy of fixes.
GERAGOS:
And the FAA has, or there has been movement, a movement since 1997 to try to get
GPS technology approved for FAA landing; isn't that correct?
RODDIS:
Absolutely correct.
GERAGOS:
And as of today in 2004, it's still not approved, correct?
RODDIS:
Correct.
GERAGOS:
Now,
RODDIS:
But –
JUDGE:
Well, is that because of the methodology, or is it,
RODDIS:
No. It a just takes a long time to change something as important as that. You
don't want to crash an airplane by accident.
GERAGOS:
We don't want to kill somebody by accident either.
RODDIS:
Well, true. But I'm not sure how that applies.
GERAGOS:
Well, this is a death penalty case.
DISTASO:
Objection, your Honor. It's not relevant. It's argumentative.
JUDGE:
Sustained. Facts are still the same.
GERAGOS:
Now, isn't it true that the Coast Guard, when using GPS technology, uses four
receivers minimum in order to navigate?
RODDIS:
I have no idea what the Coast Guard does.
GERAGOS:
Are you aware of the fact that the single receiver is not allowed in a
commercial or Coast Guard Marine applications for navigation?
RODDIS:
To my knowledge that's because of possibilities of failure of one of the
systems, they have a backup.
GERAGOS:
Okay. Failure like,
RODDIS:
Not because of the accuracy.
GERAGOS:
Failure like we had here?
RODDIS:
I don't think you had a failure. But still,
GERAGOS:
January 3rd through the 21st wasn't a failure?
RODDIS:
What was the, which was that one, the Fresno spike?
GERAGOS:
If you had,
RODDIS:
No, the device apparently did not work at all. You didn't fix out in the
tracking data.
GERAGOS:
If you had a second device installed, that would obviously give you the ability
to check the data of one against the other; isn't that correct?
RODDIS:
I suppose that would be true if we had one.
GERAGOS:
We don't know today, as we sit here, you haven't taken a look at that data,
correct, the 3rd through the 21st?
DISTASO:
Objection. Been asked and answered.
JUDGE:
Sustained.
GERAGOS:
Do you know what code phase and carrier phase are?
RODDIS:
Code phase and carrier phase?
GERAGOS:
Code phase --
DISTASO:
Objection. Relevance.
JUDGE:
I assume it has some relevance, he wouldn't be asking. Do you know what he's
talking about?
RODDIS:
I'm sorry, these are technical issues that's related to the fact that GPS
transmission is what we call CDMA, Code Division Multi Application. CDMA. So you
can get phase measurements from different types that are within that system.
It's quite a complex, technical issue. I'm not sure what the relevance is.
Perhaps you can continue.
GERAGOS:
Let me ask you. Code phase and the carrier phase, is the understanding of them
essential to the methodology of GPS technology, one or the other? Is there a
difference?
RODDIS:
Yes.
GERAGOS:
That's the relevance. So I'm asking you.
RODDIS:
No, it's not irrelevant, but I'm not sure --
JUDGE:
He said that's the relevance.
RODDIS:
I see. The relevance to the technology.
GERAGOS:
What is the difference between code phase and carrier phase?
RODDIS:
Well, the carrier is the signal which is transmitted. The code is a code that's
built into, imposed on the carrier. The carrier phase relates to matching the
phase of the code that's imposed on the carrier; and the code phase relates to
matching the phase to the carrier itself.
GERAGOS:
And you, which of those phases is the device that's applied in this case?
RODDIS:
I suspect you have to ask Doctor Loomis that one. But I suspect that would
probably be the code phase.
GERAGOS:
And is the code phase more accurate or less accurate than the carrier phase?
RODDIS:
It is less accurate.
GERAGOS:
And that's what we have in this case; isn't that correct?
RODDIS:
Absolutely. But you haven't asked what the accuracy is.
GERAGOS:
And the accuracy, we're talking about accuracy, talking about the differences
between code phase and the carrier phase, correct?
RODDIS:
Yes, correct.
GERAGOS:
Thank you. I have no further questions.
Redirect Examination by Rick Distaso
DISTASO:
Just, you were finishing an answer, Mr. Roddis, about what's the accuracy
difference between the code phase and the carrier phase. Can you just explain
that?
RODDIS:
Yes. To be truthful, I can't give it to you very well here. But the code phase,
which is what we are using, gives you an accuracy in the order of a few feet, or
tens of feet. And the in present situation, carrier phase would be used more for
survey applications, or where you want to be very accurate, and where you are
interested in the matter of inches, or whatever. And so for this particular
application, I would say what we're doing is sufficient.
DISTASO:
And just one other question. Let me show you People's 2. Just so the, let me
show you People's 2 so the dates are clear. Do you recognize this set of
documents as a repair order that was sent back to you, or to Orion?
RODDIS:
Yes. I'm sorry.
JUDGE:
What are you showing him?
DISTASO:
People's 2. And just for a date, it looks like the date your company wrote
the customer a repair report was on June 12th of 2003?
RODDIS:
Yes.
DISTASO:
Okay. No further questions.
|