Certifying GPS results

[jginnane_]

Several months ago I was stopped by a PO with a laser unit. He said I was traveling 80+mph. I pointed to him our on-dash TomTom GPS had indicated I was in the low-mid 70s. The officer spoke about his own unit's speed certification and said he was happy to address this in court. We politely disagreed, and I was lucky enough to received a violations ticket (no points, just $$ fine).

QUESTION: How can we establish the reliability of GPS in our dash cams for LE? The TomTom unit mentioned above did not save a record, but any decent dashcam w/ GPS option will record right on the video. In a court, you would (apparently) have to play the video, then measure distance traveled between landmarks versus the clock to show that you were accurate. But I think this could be subject to challenge by the other side, especially in terms of quick-happening events in a multiple car accident.

Is there a GPS certification procedure available? Are some models offered with GPS accuracy certified?

 

[jokiin]

You won't find any that are certified

 

[jginnane_]

You won't find any that are certified

Is there a trade group of GPS manufacturers who would see value in creating such a certification?

 

[jokiin]

Is there a trade group of GPS manufacturers who would see value in creating such a certification?

Would be too hard to do something which would be acceptable in a court of law so would be of little use

Police equipment gets recertification and calibration on a regular basis to be deemed admissible as evidence

 

[niko]

I wrote about this in some other thread.

You dont need a gps data to determine a speed.
Speed can be calculated from any dash camera which records time.
It is not that hard to calculate a speed. You drive back to the accident place, take laptop with you along with evidence video, synchronize some point on the road with video footage point, then forward video lets say 3 sec, again synchronise place on the video with place on the road, calculate distance from A to B and convert / calculate into speed.

You can go to court and win it easily if after your tests / calculations you can be sure you were not speeding.

I wrote about this here and found speed calculation formula.
Speed Calculated by Distance and Time
http://www.csgnetwork.com/csgtsd.html

 

[jokiin]

actually doesn't even need to record time, distance you can go back and measure, video runs real time and time can be calculated via fps for accuracy if you were that keen, no judge is going to let you waste that much of the courts time over a traffic ticket though

 

[niko]

actually doesn't even need to record time, distance you can go back and measure, video runs real time and time can be calculated via fps for accuracy if you were that keen, no judge is going to let you waste that much of the courts time over a traffic ticket though

Yes, on PC you can see FPS ( frame per seconds ) value, so no need time record display stamp on video.

If presenting this data to judge, then you certainly win case even without going into the court. Also before going to judge, you can present same evidence to police station where that police officer works and they will withdraw they ticket. Of course if you are not living in that area where you got a speeding ticket, then you can ask media ( TV / Radio etc ) to publication of police forces radars inaccuracy story, but of course if its your home-town, better resolve this case peacefully ( never "fight" with neighbours )

 

[jginnane_]

Of course if you are not living in that area where you got a speeding ticket, then you can ask media ( TV / Radio etc ) to publication of police forces radars inaccuracy story ...

I'm thinking of the type of case which involves several cars in an accident, and a need to establish a precise timeline of events occurring. (Car A hits Truck B, rolls over, smashes car C into D, etc.) A "certification" would help all involved sides stipulate who hit who first, when, and what chain of events caused event J (for example) to occur. This is made a bit harder because dash cams have differing fields of view, and so a distance traveled of 88 feet, which is 60 mph exactly, may not jibe with what appears to be 88 feet in a crossing, curving, slowing, or diagonal path. And while NTSC systems nominally record at 29.97 frames per second, PAL video is at 25 fps.

Sorry if this is confusing; I'm trying to point out that in a real court of law, some of what we think is clearly evident from a video might turn out to be Jello -- if an opposing side has incentive to try to mess things up. If any of the counters could be made bulletproof, that would help. It's not just GPS particulars (including displayed speed) but even the clock itself that could be slightly off. (To give a simple example, normally 30 fps means a 2-second clock jump, but periodically there needs to be a correction for that 00.03 second difference.)

Why does this matter? Think of incidents involving a truck with flammable cargo; or of a schoolbus with children. In each type of calamity there are millions of dollars of liability to be established. Would any court casually accept an uncertified fisheye image from a $200 camera? It's more likely that as stakes rise, one party would fly an "expert" from the manufacturer to explain technical aspects of a recording.

 

[Quick]

First, the police don't use gps for speed. Usually their radar/lidar/whatever equipment is calibrated and certified. Some places it's required to be calibrated/checked by the officer at the beginning of each shift.
GPS just doesn't have the granularity/accuracy to be used for instantaneous speed measurements -- especially unassisted gps. It will be accurate over relatively long distances for *average* speed over that distance. On my camera, when I brake and stop for a read light I will be completely stopped while the speed on the camera slowly counts down to zero. For example, most cameras are sampling at about 1/second. I can accelerate from 0-60mph in 4.6 seconds and the acceleration is not linear. so... what was my speed at 3.5 seconds? Then you're left with time over distance, but as pointed out above you have to calibrate the processor speed because frames per second might not be per second.

 

[jginnane_]

First, the police don't use gps for speed. Usually their radar/lidar/whatever equipment is calibrated and certified. Some places it's required to be calibrated/checked by the officer at the beginning of each shift.

I'm convinced LIDAR (aka "laser") is far more precise than anything we're going to get in our cars, especially the speedometer. We've had Beltronics and Valentine Ones in our cars for 10+ years, and are fortunate most NJ police have only Ka band (at best) ... once in a while they borrow a LIDAR, then the radar detectors wake only to tell you "You've been Served."

GPS just doesn't have the granularity/accuracy to be used for instantaneous speed measurements -- especially unassisted gps. It will be accurate over relatively long distances for *average* speed over that distance. On my camera, when I brake and stop for a read light I will be completely stopped while the speed on the camera slowly counts down to zero. For example, most cameras are sampling at about 1/second. I can accelerate from 0-60mph in 4.6 seconds and the acceleration is not linear. so... what was my speed at 3.5 seconds? Then you're left with time over distance, but as pointed out above you have to calibrate the processor speed because frames per second might not be per second.

This is a better explanation than most I've seen. Sampling 1X/second introduces enough inaccuracy that if a court opponent wanted to make a point about your indicated speed, at any particular instant your video record could be off 50% or more. It gets harder to argue there's imputed errors if the sampling period (say, 10+ seconds) gets larger, but it's unlikely

So ... what's the best way to present our recorded "evidence" in a court of law? There must be established protocols from CCTV systems.

 

[Street Guardian USA]

Here you go

LoL, J/K
http://www.senteacher.org/worksheet/3/CertificateMaker.html

 

[Street Guardian USA]

I couldn't resist:
http://dashcamtalk.com/forum/threads/amazing-gps-on-dod-ls330w-ls430w.3119/#post-36106

 

[jginnane_]

I couldn't resist:
http://dashcamtalk.com/forum/threads/amazing-gps-on-dod-ls330w-ls430w.3119/#post-36106

Funny! Love it.

Question still stands, though ... if someone's trying to rip apart your dashcam video in a court of law, what are some usual defenses about accuracy, etc.? Or is the only defense that you maintained custody and did not alter the video, and WYSIWYG --?

 

[jokiin]

Funny! Love it.

Question still stands, though ... if someone's trying to rip apart your dashcam video in a court of law, what are some usual defenses about accuracy, etc.? Or is the only defense that you maintained custody and did not alter the video, and WYSIWYG --?

I think it's going to be very much WYSIWYG, the embedded speed info is not real time and will always be off by a bit, if you have video evidence though there's a lot more possibility to explain the circumstances than is possible if you're just talking about an event, what else is happening, flow of traffic and where your vehicle sits within that, traffic and weather conditions, time of day, dusk, dawn, sun in your eyes, can coming at you with driving lights on, that sort of thing is all available for people to see with their own eyes

if you're trying to defend a speeding ticket the embedded speed info is less likely to be any use to you than video that shows you were being overtaken by faster vehicles etc, if that hasn't happened and you got a speeding ticket it's more likely because you were speeding

 

[jginnane_]

... if you have video evidence though there's a lot more possibility to explain the circumstances than is possible if you're just talking about an event, what else is happening, flow of traffic and where your vehicle sits within that, traffic and weather conditions, time of day, dusk, dawn, sun in your eyes, can coming at you with driving lights on, that sort of thing is all available for people to see with their own eyes

All these conditions are excellent points that we tend to forget about. Sun in your eyes or reflecting off the road surface -- oncoming cars with badly aimed headlights -- you really have to show, you can't just tell anecdotally. There's no other way to put a jury into your frame of mind than by showing the actual video.

if you're trying to defend a speeding ticket the embedded speed info is less likely to be any use to you than video that shows you were being overtaken by faster vehicles etc, if that hasn't happened and you got a speeding ticket it's more likely because you were speeding

Had a first-time event yesterday: a speeding Toyota in the passing (left) lane, and the state trooper sneaking up on him while hidden in the slow (right) lane. I was forewarned by my radar detector, so I was strictly at the speed limit. The PO was a real pro using short bursts of instant-on radar. What made this event memorable is that I caught both cars first on my rearview camera, a GoPro Hero2, then the interception and pullover happened right in front of me, caught on the Panorama2.

Incidentally, two miles later there was a standard speed trap set up on the center median, and I've been taught by my radar detector that the first BEEP has to be immediately reacted to. There are days when traffic police like to swarm a section of roadway, and yesterday's weather was right for a bear swarm.