They are reasonably accurate these days if you get a 10hz system but its still +-3m or something like that. However the DL1 etc claim they use accelerometers to correct this. i.e. the relative position of the car on a lap is reasonable, but the offset from the track edge may be out a bit (and change over the day).

As R+C says you can use differential gps but this is expensive (requires a pit lane beacon), or now you can get correction info and apply it later to correct the positioning further. I think the DL2 software will do this, but feature is locked out the 'budget' DL1. Not sure whether its worthwhile or not though.

 

I am a bit ( to say it softly) confused .

The -+ 3m error will be consistant ?

a) Does this mean that stopping on a lap at exactly the same place allways the GPS will show two,three or more different positions at consecutive laps with an error of +-3 m? or,

b) Does this mean that stopping on a lap at exactly the same place allways the GPS will show two,three or more different positions on different days ?

I certainly preffer option a)

Being a) and/or b) true...What is the use of those systems if I can not compare different laps or different days ?

Help ! Please !

 

thats more than 1.5 car widths.

For the Supercup we used the ADL2 from Motec. Its inertial system is impressive. If required we can actually adjust the track layout without much difficulty.

With professional drivers, the exact 'where' on the track is not that important. If one driver is faster than the others through a section of track, a quick look at steering angles, braking points, acceleration points usually shows what is happening as we overlay the different drivers. The discussion quickly resolves any different line issues as the question 'how did you manage to do that at that point' is answered, more often than not we discover that resurfacing has created an anomaly in these circumstances.

During the race itself, the line is often of minor importance, as drivers defend their position and try and get past their opponents. One of my preferred passing techniques requires that one learns to drive fast right off the line so that you are in the best position at the following corner and then the other guy just eats your weeds. Its very hard to defend against.

R+C

 

No, the GPS system can be in error at any time and the error varies continuously. So a stationery car could vary in displayed position through the day such that the distance between the two most distant recorded locations were 6 metres apart, unless the system has 'smarts' to ignore sat error induced movement. But you cannot rely on these systems as there is no certainty as to the accuracy of any baseline position and a race car sliding around on track can be quite tough on the inertial system.

I drove once from Weissach to Winnenden when the Sats blinked out, the inertial system was only 7 K out after about 45 minutes built up area driving.

R+C

 

Thanks Nordschleife !,

Now it seems to me that the best use of a GPS system will be the comparison against a "BEST" run ( mine or elses..) . As a mater of fact, this morning it was a "resistance" race ( three hours race ) with three drivers in the same team . They were using the GPS to compare their relative technique at the same turns.

If the GPS system is inconsistent, the only way to compare would be one against the other during the same lap ? Should the reading be erratic, how could I compare my ( braking, acceleration, etc..) driving against someone else if I can not rely on the fidelity of the position ?

Sorry to keep you busy. I am aware of your experience in such activities,,

 

Even on a Speed vs Distance graph , the distance will be measured from a unique starting point ? If this distance is measured by the GPS, and the GPS is reliable up to +-3m ( As stated in the DL1 sheet ), how can one compare his technique against another driver or against his own previous performance ?

If the error is inconsistent, it could go - in theory at least - up to -3m one lap and +3 m the other, this is...6 m !? The difference between braking 6 m later on the straight line at my nearest track ( Jarama) is the difference between taking the next turn or finishing out of the track...

 

there is some misinfomation here especially from our Euro buddies-Most Euro military people are pushing Galileo (the Euro equvialent of GPS) because it galls them that the US military has complete control of GPS as it is now(Glonass is the Russian equivalent and totally unreliable here) So if you get a Euro opinion of regular GPS take it with a grain of salt. No insult intended just 19 years of GPS experience talking and having worked with MANY countries on different weapons systems and presenting and hearing a bazillion papers at GPS conferences all over the US. (the international GPS technical meeting is in the US every year in the later part of Sept-see ION.Org for info-Insitute of Navigation of which I am a member)

As a GPS engineer for the Navy way back when we would only have 2-3 satillites visible yes the GPS solution could be more than 3 meters-but now in the US and usually the whole world there are so many SV's(space vehicles) visible your solution will not drift much because most GPS receivers today ( I am not familiar with the GPS recievers in these tracking systems) know how to switch out SV's when the solution becomes inaccurate. Most receivers today are at least 12 channel and sometimes there are 11 SV's visible here in the states. You need to be tracking 3 SV'sfor position in 2D and 4 SV'sfor a really good 3D solution. Most of the time there are 7-8 SV's visible. This allows modern recievers to get an overdetermined solution which is better than 2 meters. At these speeds in cars there is no problem for a good receiver to get a good solution. In fact many GPS manufacuturers have figured out how to beat US government control of the solution.
Using both L1 and L2 signals and they are getting in the less than 1 meter error and even in the centimeter resolution for well placed receivers. The US government turned off Selective Avialbility(S/A) in 1999 or 1998. So the bulit in drift is not on right now. And probably wont be turned on with the New SV's going up and a new civilian and military frequencies going up. With these slow speeds if these companies are using modern multi channel receivers it should be easy to get more than enough info to help the average driver. Differential is not too expensive for more well funded teams and you can get survey quality kenimatic receivers and the real time service is like $100 a year (raycal)-- if you post process you can actually get the Differential station info off government websites for free.(find one close to your track) the processing software is the only cost besides the car receiver and antenna. If you dont have a crew I dont see how real time data would help a driver, so the post processed data would be more than enough. just ask how many channels the receiver can receive at once and if its more than 8 or 12 you should get more than enough accurate data to help the weekend DE driver at the track.

ANd yes I would trust the GPS to map out the track if you drive one lap outside and one lap inside to get an idea. This 3m error is worst case and it should be alot more accurate than that. as example I found a flat contrete marker in the desert using an old magellan 4 channel reciever using only the lat and long of the marker that was 2 feet. It was buried in sage brush so that solution that day was better than one meter because there is now way I would have found that thing.
You should be able to repeat the map after the session and duplicate the map to with in enough accuracy to make you feel comfortable of your system

 

In the days before GPS, we did it with good old fashioned inertial navigation and star tracking, still yielded a pretty damned good CEP.

GPS is cool, but nothing can beat a really good spinning mass inertial sensor. The Peacekeeper (MX) gyros are an absolute work of art, as is its accelerometers. Fully floating beryllium sphere, CEP of better than 100m
http://nuclearweaponarchive.org/Usa/Weapons/Airs.html
.......God, how I miss those days.

 

The only problem with our INS sytems these days is the manufacturers and developers started to let INS systems get worse than they were in the 80's because now they use
GPS to update the INS and get a combined INS GPS solution-they let the mirrors on the Ring Laser gyros get smaller and less finished than before because supposedly they didnt need a good INS since you had GPS updating. The truth is that GPS can be denied and then what do you do with a substandard INS? they are coming around though and building better and smaller IMUs for these systems and if you can build a 1inch by 1inch INS system the world will beat a path to your door. It seems as the silicon rate sensors and the accellerometers are not good enough-small enough for cars but not accurate enough for the military. Its a good thing it only took 10 years to figure out GPS is not the end all do all for navigation, and INS research started to ramp back up. Youre right I love a good INS system.

 

search the racing forum for this topi. a lot of good information.

 

I have a traqmate and LOVE it. Definitely the best performance per $ I could have spent (outside of a day of instruction from a factory driver). I have the RPM hooked up as well as a brake sensor. They provide much needed info. The one I have not yet conquered is the throttle position sensor. Maybe this year. PM me if you want hook up info.