Snipped from the "Aquapac location" thread -> a new thread seemed more appropriate:

quote:

---

Originally posted by Roo

We may not like the accuraccy levels that the WSSRC have set for GPS record ratification but they are the recognised authority representing the International Sailing Federation and any record claim has to have their blessing. It is possible to meet their criteria with off the shelf units but the cost of such equipment is not economically viable at the current time.

---

yes and no. The WSSRC requirements for speed sailing are a little too weighted in their own favour. They are *a* recognised body, but there is no reason why a more appropriate body couldn't be setup to represent the non-million-dollar sailing fraternity.

Some of their requirements are too limiting, given that technology is changing, eg:
- they have a requirement for "positive buoyancy"... What about locations where it is only 3 feet deep across the whole body of water; you can stand up everywhere, you dont need to be positively buoyant.
- they require a "three month pre-notification" of a possible claim. WTF? If it is windy on the weekend, I want to make use of that wind.
- paying them may thousands of pounds/dollars just for the sake of an observer. Is an observer required? sure, but why do they need to fly out to ... (pick your country) - why cant we use a local representative?
- wording of the rules so that only Trimble is allowed to be used is simply pandering to the expensive end of the market. Trimble make great stuff, but $10,000 is a bit much do you think.

Part (most?) of the reason that speed sailing disappeared from windsurfing was that it was useless holding any event while trying to comply with the WSSRC rules. Now that run measurement is easy (removing the obstacle of needing to outlay $$$, stupid rules and video timing), as the last two years have shown, the sport will grow again.

"...there is no reason why a more appropriate body couldn't be setup to represent the non-million-dollar sailing fraternity"

The WSSRC have been doing this before windsurfing existed. They are recognized by the ISAF as the only body responsible for sail speed records. We are talking about speed sailing. If you want to exclude the internationally accepted body from your record attempts .. what crediblity do you have.

"- they have a requirement for "positive buoyancy"... What about locations where it is only 3 feet deep across the whole body of water; you can stand up everywhere, you dont need to be positively buoyant"

Are you serious? The speeds involved are 90km/h. Former world record holder Eric Beal concussed himself in a stack on the canal. It's not very deep but he could have downed as can anyone in 3 feet of water if knocked out and weighted with lead without compensating bouyancy.

"- they require a "three month pre-notification" of a possible claim. WTF? If it is windy on the weekend, I want to make use of that wind"

A reasonable request. Anyone serious about making a record attempt will be preparing long before 3 months. You can book a 1 month time period and have the observer on call for when the wind is suitable.

"paying them may thousands of pounds/dollars just for the sake of an observer. Is an observer required? sure, but why do they need to fly out to ... (pick your country) - why cant we use a local representative"

You obvious haven't thought this out. How can you make any valid claim without an independant observer?

"wording of the rules so that only Trimble is allowed to be used is simply pandering to the expensive end of the market. Trimble make great stuff, but $10,000 is a bit much do you think."

If you are willing to organise the appropriate independant testing of the measuring devise you wish to use like Maquarie Innovation did then perhaps you could use another gps. Or you could just hire a Trimble 5700.

From the PDF:

Incorporating sufficiently long CRC check
(128 bit or longer) or other adequate digital
certificate inside the GPS-Doppler data log
file for the purpose of detecting any data
tampering.


This will never be 100% secure, because the competitor has access to the file. Any key/algorithm can be broken given enough time and resources, so there will always be some element of doubt.
Another problem is the fact that a competitor could change the firmware inside their NAVI, which could easily inflate their doppler data by, say, 10%...

I don't believe that doppler data has sufficient error correction in it either. For example, if you get a bad reading in Lat/Long track it's easy to figure out that it's bad data. But errors will accumulate if you're using doppler data, so it doesn't automatically take care of errors.

I appreciate what the gpsss community are trying to achieve, however I think they need to do a bit more homework.

The criminal mind at work

My sister did say any size estimate from Gestalt
should be treated with caution

However she agrees he is very quick

I don't think I have to worry about this lot for quite a while yet

quote:

---

Originally posted by yoyo

Yeah .. not 'til Friday at least, eh elmo

---

Verkingverkingverking

Work work work

I'm sorry Sailquik, I didn't mean to imply that the team hadn't put in a lot of time and effort.

The "Error of Measurement" idea sounds fantastic, and would work well.

I guess the thing is that I couldn't see how the new system would work, from go to whoa, including how the data gets from the device to the record keepers, without the possibility of it being tampered with. This includes having suspect firmware loaded into the NAVI.

btw at the end, shouldn't "Transverse" be "Longitudinal"?

Doppler is particularly good at measuring speed, which is why it is used by car manufacturers and racing teams rather than a trackpoint location based system. They and we dont need to be particular to the cm about where the speed was achieved. What is important is how fast we went and how accurate those measurements are.

GPS units like the Locosys GT11 generates the doppler data we need to achieve this significant accuracy. The results are output every second, not every ten seconds, but 10 recordings per second does not make the result more accurate. Only more often! However ten recordings per second would be vastly better at capturing fine detail movements like recording turns and jumps, which is why we are keen to get support for 10hz doppler. The SIRF chips measure doppler ten times per second now to calculate the one second outputs, so its technically possible with the current hardware to generate 10 hz output... Yes it will rival electronic timing, yes the performances will be beyond question, and yes the opportunities will be amazing!

I urge everyone to look past the hecklers and make up there own minds about what this technology offers. It is a fantastic opportunity to open up speedsailing and sailing in general with many new ways to have fun, like Alpha Racing, that will enhance our enjoyment of the sport.

You make some valid points Frant. Of course the general (ie non-sailing) community is becoming very familar with the uses and potentials of gps , if not the workings or history of it through the ever increasing exposurure in cars, personal navigators and mobil phones. And this exposure will increase exponentally. Given the minimal cost of having a logging device capable recording a years worth of travel at 1 second interval (Navi with 2G card) it is hard not see such units being standardised/mandated in cars in future.
Goverments will say it's cheaper than cameras for such things as London's congestion charge or Melbourne's City Link tarriffs. Police will point out the benefits to them as will accident authorities (black box). Car makers will imply numerous benefits and other owners (jealous spouses, worried mothers) will see others. Of course the civil liberty unions will complain..and they will be as effective as they were at keeping video cameras out of cities.
The point is, drivers will challenge their speeding fines when they realise the units in their cars are much more accurate than the multinovas. Once in court, the accuracy will have to be officially checked. Once verified in a court of law as an accurate velocity measuring device, a device in common use in the general public. Then a speed claimed in a well publicised/promoted event ( but not WSSRC santioned) may well be accepted as valid in a publics mind now familar and comfortable with the amazing accuracy of these tiny devices. That is something the WSSRC may need to keep an eye on. Technology may sideline them.

quote:

---

Originally posted by Roo

"which is why it is used by car manufacturers and racing teams rather than a trackpoint location based system."
Again you seem to miss the whole point, they are measuring speed not claiming a speed record. In Formula 1, the pinnacle of motor sport, they record the time to complete a lap and an average speed over the distance of the lap.

---

In the several motorsport teams I've worked with, they never used GPS for speed measurement. This is for a couple of reasons, one of which is that it's too slow, another is that you need a clear view of the sky, another is that it would be difficult to correlate this to the other sensors that are being logged.

Just my $0.02.

Doppler is a highly effective system for measuring speed and has the potential to be more accurate than current video timing or timing through eletronic gates.

Even the WSSRC dont care too much about where the speed was achieved because they allowed unfixed gates for MI. Using simple consumer GPS the location can be determined to within 25m. There is no purpose or benefit to determine a more exact location because location is not the basis for the speed measurement. And I doubt if any system could prove an exact location, so it becomes a mute point.

And here is some analysis from NASA on GPS for tracking spacecraft. The speed errors correlate very well with work done by Tom Chalko and Manfred ****hs.
http://geons.gsfc.nasa.gov/library_docs/Fmt97gps.pdf

I just love those suction cups

Out of interest, what difference in readings is one likely to get if one were to put a GPS on each arm? What's a reasonable figure that people normally get?

Yes Nebs, I have done quite a bit of side by side testing with GPS's. The very first thing I discovered was that with the Foretrex the results varied so much it alarmed me. It suggested that if we run a competition with them the variation/error in them could easily cover all the top placings meaning that it was almost meaningless. The difference between units worn side by side was often more than .2knot. That might not sound like much at first, but take a look at the GPS-SS rankings and see how many riders that can cover, especially in the middle rankings.
Most of the problem was from the grid effect because of the way the Foretrex (and Geko) clip the precision of the saved data to save memory space.
This set some of us looking for a better GPS and we found that the Garmin Edge and some others do not clip the results and gave more consistent results. Then we discovered that the Navi/GT-11 is capable of outputting NMEA data which contains extra information on the accuracy and reliability of the signal. Roo had a major part in this discovery and deserves much credit for this.
The result was that we had improved accuracy and the information to be more confident of that accuracy. Side by side testing revealed that the difference was now mostly much smaller and more consistent.
Now we reconed we had a tool that could provide much fairer results for GPS-SS type rankings and especially for speed competitions. As I understand it, Roo then expendeed an enormous amount of time and energy trying to find a way to process the info from the GT-11 in such a way that it would be even more accurate with the idea that it may be able to be as accurate enough to meet WSSRC standards. It seems this was not to be.
Then Tom Chalko suggested that that such accuracy should be able to be had by taking a different approach and using the Doppler calculations rather than trackpoints. He reasoned that this was potentially very much more accurate than the trackpoints since it didn't have the same vulnerability to changing atmospheric conditions and would not require extensive post processing. The accuracy can also be proved to a known level using the underlying science.
Now we have a method that can be used for competitions and we are able to be very confident that close results are ranked correctly. Sise by side testing shows consistency to closer to .02 Knots. For me that is the most important aim. The spin off is that anyone who does happen to go close to or exceed a recognised Record can be very confident of the comparison. The whole question about whether the WSSRC will recognise this method of speed measurement is another matter and is best left up to anyone who wants to put a proposal to them. But the statement that this method of using Doppler is as accurate, or more accurate than Video timing is demonstrably correct and it is also true that the WSSRC sees video timing as accurate enough for their standards. The rest of the arguments and speculation are just that, speculation, and..... politics!

Just a small correction to Daffy and Mal's comments, I had recognised the importance of "doppler" speeds long before Chalko had even become involved in the project or even used the units. I asked Mal early last year to incorporate "doppler" speed and calculations into RealSpeed as it was a much more accurate way of measuring instantaneous speed. My aim was to combine both trackpoint accuracy with the "doppler" speed by boosting the data stream out of the units in binary code to allow post processing to WSSRC record levels. All my research showed the NAVi isn't stable enough to do this and the errors produced by the units aren't constant or provable to allow the necessary levels of accuracy for record claims. For a competition situation where every rider is using the same unit with the same firmware and the same sd card recording nmea data, there is a very high level of confidence in the results. The nmea data they produce provides all the information we require to confirm the results, the binary output is still too unstable to be trusted and has the adverse effect of causing voltage spikes during computation cycles that causes failures with certain sd card that have sensitive voltage requirements resulting in the units shutting down.