Here are a couple of academic papers on the use of GPS for establishing speed records. These are not for the mathematically challenged. One is 9 pages the other 29 pages.
http://www.google.ca/url?sa=t&rct=j&q=use%20of%20gps...

http://www.google.ca/url?sa=t&rct=j&q=use%20of%20gps...

Quote:

Originally posted by Scudley These are not for the mathematically challenged.

x approaches infinity, where x = understatement

ATB,
Sam

I'm not even going to look. Can someone give the Short version? My gps has never given me a reading that didn't make sense to how the run felt. I have a very high level of confidence in it's readings.

Both links are for the 9 page version. Not that I want to read the 29 page version.

I tested my Garman eTrex H against two cars I've owned in the past 6 months and each came within a maximum of 2mph outside the GPS at 60mph. I was curious as how accurate the car speedometer was.

My friend has an iphone app that allows him to get 70mph by walking down to the bottom of his garden, but I think the phones work differently.

I noticed that he mentions Sandy Point in Victoria Australia , a popular buggy beach.

Quote:

My friend has an iphone app that allows him to get 70mph by walking down to the bottom of his garden, but I think the phones work differently.

Dude, that is one FAST gardener!!

I've checked mine in the car quite a few times and find it quite believable, but I have seen spurious readings on my older magellan PIECE OF GARBAGE (oops, did I type that out loud?) while both in a buggy and in a sailplane. According to the Magellan, I had a top speed of 2000kph in the glider.... and another reading of 35kph in the paddock when I know I was only tootling along at 20 something.

My verdict is: usually reliable but can have the occasional glitch.

How accurate is the location finding of GPS?
From natural Resources Canada (http://www.geod.nrcan.gc.ca/edu/geod/gps/gps13_e.php) single point measurement is good to about +/- 5m.
How fast is a kite buggy ~ 40m/s.
GPS calculates speed from displacement between locations and time traveled. If measurements are ocurring every second, the distance calculated is about 40m, and because we only know the positions of the two points to +/-5m, +/-10m. That is 25% accuracy on distance. Are you going to base a world records on measurement that are that bad.
Even if your GPS is good to 1m, the speed accuracy is going to be +/-5%.
Given that you know speed to +/-5% is there any difference between 83.1 mph and 85.2 mph. Mathematically speaking there is none as you really no the speed to 1 sig fig.
There is also discussion on how some of the sources of error are environmental which means increasing the number of GPS units does not improve the situation unless the runs are made under the conditions.
This is why the author is suggesting that the doppler shift of carrier frequency is a better source for velocity measurements.
The papers are discussing how to improve these number either by changing the method of measurement (9 page) or by statical analysis of the track data to provide to increase confidence (29 pages).
The nub is that we might as well draw names from a hat as use GPS.
S

Does anyone want the link to the 29 page doc on analysis of GPS data to establish a speed over a distance record?.If you think the 9 page doc is dense math, you don't want to see this one.

So the speed records set to date are incorrect or not valid?

If you were to use significant figures, the record speed is around 80 or 90 mph. Anybody with a speed in that neighbourhood is a contender. Scientifically, I would have to say the current record is fun, but meaningless. Validity is up to the sporting body.
S

I know that when I'm cruising along and watch my current speed, and I feel an extra pull from the kite, I can watch the speed go up smoothly and predictably to match the extra pull I felt. Likewise, a steady speed across the lakebed, shows a steady speed on the gps. I would think that with all that variability, the speed reading would be erratic all the time. Likewise, even with the electro stuff in a car, the gps will consistently read within a mile or 2 per hour the same as my speedo in the car. I have more confidence in the gps than my car speedo. I just don't see any readings from my gps that make me question its accuracy or reliability. If it did, I wouldn't bother attaching it and would go to a bicycle computer that measures speed with rpm of the wheel/tire considering its circumference.

Bob, what is positional accuracy of your GPS? To be able to say your max speed is 85.1 mph, your gps is going to have to have positional accuracy to about 6".
Those who are paid to do this crap for a living would say your confidence in your GPS is misplaced.
Sorry, please don't blame me. I feel like I told you there was no Santa.
S
ps Your bike computer, if properly calibrated, is good to about +/-2% or about 10x better than gps.

Stole this from GPS explained.
Accuracy Values by Garmin Receivers
Accuracy of position determination Accuracy of position determination

The declaration of the accuracy by Garmin GPS receivers often leads to confusion. What does it mean if the receiver states an accuracy of 4 m? This readout refers to the so-called 50 % CEP (Circular Error Probable). This means that 50 % of all measurements are within a radius of 4 m. On the other hand, 50 % of all measured positions are outside of this radius. Furthermore, 95 % of all measured positions are within a circle of twice this radius and 98.9 % of all positions are within a circle of 2.55 the radius. In the given example, nearly all positions are within circle with a radius of 10 m. The determined position is in the worst case accurate to 10 m.

I'll add my anecdote:

I have a Garmin (GPS) in my car, and usually watch my speed on there instead of my speedometer. I know while accelerating, I've seen my speed on the GPS jump erratically both high and low of what my actual speed is (comparing it to my speedometer). I'm not sure how it compares when cruising simply because I'm not watching the speed on the GPS very often with the cruise control set.

More information is always better imo.

Even though it may be a bit of a buzz-kill for people who use GPS to track speed records, it's worthwhile to know how accurate they actually are.

Maybe they need to change the output display from that precise "83.2 mph" to "Hmmm, I reckon 'bout 80".
You still need to be going damn fast to get that reading in the 80s.
S

Quote:

Originally posted by Scudley Maybe they need to change the output display from that precise "83.2 mph" to "Hmmm, I reckon 'bout 80". You still need to be going damn fast to get that reading in the 80s. S

He was...and we as a buggying community trust the accuracy of our GPS devises

I think there was 3 + gps's attached to one buggy and all read the same top speed. Maybe we should try multi brand gps to compare accuracy.

Oh Scudley, no shooting of the messenger here, its just that my observations make it seem much more accurate than that. I will admit I've never had a max speed reading over 80 mph............... Yet.

Who cares ? I've seen speeding tickets beaten down with math too.

It's an exercise for braggin' rights. Very precious, highly dangerous to achieve, life-long pursuit, summiting Everest, Holy Grail kinda braggin' rights. It's "extraordinary", meaning most of us will never see it. It is a gauge for those who keep score. "Lead Dog" is very special and the skill, talent and giant nards required to achieve it in this game are worthy of the highest praise and admiration on any scale.

I happen to like this scale because it's cheap and accessible. If you're looking for accurate, go to Elmer's during speed week and try run the lights with the 200mph cars. Now THAT's internationally recognized accuracy... unless there's a lawyer in the crowd who specializes in beating traffic tickets.

Quote:

Originally posted by popeyethewelder He was...and we as a buggying community trust the accuracy of our GPS devises

RYFM and check the specs Popeye. You will see your trust is misplaced. Sorry to pee on the parade.

S

And another country heard from...

Thanks for tearing down my illusions and exposing the sham that is wind power sport as measured by a de-funked and egregiously flawed globular possibility simplex. Thank GOD we don't use GPS for commercial Naval, Aviation or Land navigation.

By the way, what it your suggestion to resolve the problem ?

My suggestion... There ARE other windmills to tilt at. Now would be a good time to saddle up and ride out, big boy.

greasehopper,
all I can say is read the paper. The link is at the top. If you have a problem with his work let him know.
GPS is in used the fields you list for position, not speed, and they do not use it for speed for the reasons listed.
Tell us your gps model and we check the specs and go over the math to calulate the speed and the error.
Now would be a good time to read the paper, big boy.
Then time to put up or shut up.
S

Yeah, I read it and I get the joke. My point is, who cares ? I certainly have no need to pick fly poo from pepper. It's "ball park" close, internationally available and CHEAP.

I noticed you presented no reasonable solution to this supposed "problem". I suppose one might deduce an ingenious plan to market a "certified" speedo system is afoot. Outstanding !!! Drive On !!! Best of luck, let me know how that works out. I doubt us "also rans" will be stepping up to the counter but I don't doubt there's a market out there.

98% confidence on speed of 85.2mph +/-50%
25% confidence on speed of 85.2mph +/-10%

Big ball park.
If you want, we can set pi equal three cause it's easier.
S

Okay, last time, I promise.

I'm not trying to be an arse here. A cynic, perhaps, but not an arse.

My point is, we can point to the flaws all day but, in the absence of superior means, it still rules the day.

Who cares how bad it is on paper if it's the smartest choice going ?

What is your suggestion for fixing the problem of the grossly inaccurate ground speed indication of GPS ?

You did not read the paper, or if you did you did not understand any of it, as it was a suggestion as to how to fix the problem using available technology.
Bike computer is a quarter the price and can be 10x as accurate so GPS is not smartest or cheapest- but GPS is so sexy.
Well if you want to know if one guy is really faster than another, GPS does not rule the day unless one guy is going less than half the speed of the other.
The guy who wrote the paper was concerned about measuring boat speeds, so hopefully the much larger marine market will come up something soon.
I am really sorry the news has upset you so much. Please do not be pissed off with me cause I told you about the problem, take it up with Garmin or whoever it was that sold you something that does not do what you thought it did.
I really look forward to seeing my critics provide the accuracy of their GPS so we can go over the numbers together and see whether the numbers are any better than I say they are.
Afterward maybe the group can come up with a means to measure buggy speed which is reasonable and acceptable to us all.

S

I was curious about this a few years ago when hunting. Using a gps that was accurate to about 250 feet. If the readings are done out of a circle I get the problem. There is no easy fix and also using several gps and getting an average is about as good as it can get. setting up a section of a track and doing a precision time run would be about the best we could do. Problem then becomes how do you aline the track with the wind.... Maybe a laser gun like police use or how they check baseballs being thrown....I agree the system isn't perfect but Its close very close.

I am not picking sides but am curious as to what people can come up with to maybe counter the problem.

The bottom line is not whether or not GPS is "this accurate, or that accurate".

Currently it is the chosen method to validate speeds. Not raining on Scudley either.
But it will take a whole lot of convincing to change the entire buggy community. Not saying that, that is what Scudley was/is trying to accomplish.

Thing to remember, keep in mind is how is any tech evaluated/calibrated and do you trust it?
Remember for years speed was measured by how many times a crude gear went round and round.

:D

No one has mentioned the accuracy of a paper written in 2007 by a college boy. If it were published in the New Scientist magazine then fair enough.

Something else to consider. Change never occurs based on one paper. Typically a theorem needs to be tested multiple times before it becomes a postulate.

The paper does show promise. For some reason I understood about 75% of what the author was saying. However I do have one problem with one reference site the author used. Wikipedia? Almost all major universities have stated that Wikipedia is not a reputable reference site since anyone can log in and put their two cents in with out a shred of truth to what they are citing.

Having said that, it will be interesting to see where this goes.

Why has no one sent in the specs of their GPS? Is it because it is as bad as I say?
Wake me up when someone sends in their GPS unit's specs.
S

Quote:

Originally posted by csa_deadon However I do have one problem with one reference site the author used. Wikipedia? Almost all major universities have stated that Wikipedia is not a reputable reference site since anyone can log in and put their two cents in with out a shred of truth to what they are citing.

Thank you Sean for having read the paper.
He cited wikipedia as source for definition of Margin of Error and Central Limit Theorem. AFIK neither article is subject of raging controversy. It is common cite such a source as it means everyone is singing from the same page.

the 29 page papers info. do a google search and it will turn up:
SURVEYORS, GPS AND THE
WORLD SAILING SPEED RECORD
R.E. Deakin, W.N. Cameron, D.M. Sil#@%$#! and K. Zhang
Department of Geospatial Science
RMIT University
GPO Box 2476V
MELBOURNE VIC 3001
I got a headache when they started integrating vector cross products

For those did not bother to down load the 9 page paper here is the intro.


Abstract.
This article demonstrates that speed
measurement with accuracy approaching 0.01
knot is possible by using GPS Doppler data.
The method is illustrated with measurements
made by GT-11 GPS unit made by Locosys.
Introduction
Typical approach to using GPS for speed
measurement today is to consider a series of
“trackpoints” that record position estimates
(latitude and longitude) determined by the
GPS at regular time intervals.
Each GPS trackpoint is determined with some
error that is variable and difficult to
determine. Hence, speed values computed
from a series of trackpoints have unknown
accuracy and cannot be considered reliable. It
is virtually impossible to prove the accuracy
of speed computed from a recorded series of
trackpoints.
Modern GPS devices implement digital PLL
(phase-lock loop) receivers to continuously
track carrier frequencies of a number of
satellites. For example, GT-11 tracks carrier
frequencies of up to 12 satellites
simultaneously. The frequency tracking has to
be continuous simply because each receiver
has to be always ready to receive data from its
satellite.
The very fact that data is read from any given
satellite is proof that its carrier frequency is
tracked with high accuracy.
The difference between the known satellite
carrier frequency and the frequency
determined at the receiver is known as a
“Doppler shift”. This Doppler shift is directly
proportional to velocity of the receiver along
the direction to the satellite, regardless of the
distance to this satellite.
The most inaccurate is the method that tries to With multiple satellites tracked it is possible
estimate an average speed over some to determine the 3D velocity vector of the
“accumulated distance” between trackpoints. receiver. In general, the more satellites are
Due to trackpoint inaccuracies, the line tracked – the better the speed estimate.
connecting all track points is a zig-zag, even
if the real path of a speed competitor is a
smooth or straight line. Since the length of
this zig-zag is always longer than a
smooth/straight line, the “average speed”
determined with the “accumulated distance”
method always overestimates the real speed.
The less accurate are trackpoints (the less
accurate is a GPS unit) – the larger the
estimated “average” speed and the more
impressive is the “achievement”...
Doppler
An alternative to measuring speed from series
of trackpoints is using the Doppler effect.

Does this mean everyone who won a beanie or some other swag during the Kitestorm Speed Rankings has to surrender their winnings? (Bummer on them shutting down).

Scudley's points seem reasonable. If folks want to make some sort of 'Official, Guinness BooS-P-A-M-L-I-N-K- sort of claim for a speed record, then there should be a means of validating the claim to a verifiable degree of accuracy and precision. The difference between two contending speeds would have to fall outside the margin of error of the measuring system in order for one to 'officially' be deemed faster than the other. I didn't have to read the paper to figure that's what he's getting at.

And all the folks in the 70-80 mph range on buggy or skis with a kite? I'll never come close, but I'd love to watch them go head-to-head!

krumly