Data Acquisition

[Bill]

From the recent tech session and my conversations with other drivers, there appears to be a growing interest in data acquisition. After using a G2X for about 3 years, IMHO where this can be most helpful is where you can compare your results with other, usually faster, drivers. Through the generosity of some drivers like Steve Grosekemper, Paul Young, and Don Middleton in sharing their data, I have been able to do this for myself. Considering the apparent, more widespread, interest in the club I was wondering if it would be of interest to broaden the scope to make this information more widely available. The easiest approach might be to add a new menu item to our PCA website under the <Results-Time Trial>, maybe called <Data Acquisition>, which would provide a list of those that have data loggers and are willing to share information. The format could be:

Track: driver: car(class): fastest lap time: email: Unit (G2X,IQ3,etc):

This would allow anyone interested in a particular track to directly contact someone in the group (via email) to get a copy of one of their (in the case of G2X,IQ3) .DDF files for their fastest laps. With this format you allow some control over the dissemination of your information in the event you didn’t want to send it to a close competitor .

Alternatively, if everyone agreed, the website could just have a download link (instead of the email link above) to allow direct download of the appropriate file. Individuals could submit just their fastest session (.DDF file) for a particular track to be included in the list.

Interestingly, you actually could use this information without initially having the logger itself. The software (at least in the case of the G2X,IQ3) I believe can be downloaded free. Although without your own data you couldn’t compare your results with others, having access to these data files you could at least see what information is available from these units. It could be viewed as an armchair instructor! Maybe that would convince you to get one of your own.

Having such a list of fellow data loggers might encourage exchange of other information and provide sources for help. I have written software that uses the raw data from the G2X to do analyses in ways other than what is available in the G2X software. For example, it’s possible to compare driving lines through turns, braking profiles, detailed turn analysis, shift points, and other info. It would be interesting to hear from others what other info they would consider useful which might foster some productive collaborations.

On the other hand, maybe all this is ‘over analysis’ and the best approach is ‘seat of the pants’.

Bill

[Robert Joe]

Does the software you wrote spit out textual data or graphical data?

[gulf911]

Bill, the main benefit would be if you were in the same car , or same class of car. While the units I have seen arent accurate enough to show accurate lines through a turn without a view of the track underneath. Too many variables that would not necessarily help someone in a lower class car or tires / suspension etc. All of these variables affect turn in , braking etc. IMHO. Although , I myself wouldnt need to see Steve G's data... When I chase him I usually see the passenger door , rather than the rear of the car...

[Bill]

by Robert Joe on Fri May 31, 2013 7:45 am

Does the software you wrote spit out textual data or graphical data?

The programs are written in Python and produce both text and plotted graphs. They currently don't have a GUI and are not particularly user friendly - I wrote them for my own use. If there had been more interest I would write a user interface but from the lack of any responses I doubt if I will take the time to do that.

by gulf911 on Fri May 31, 2013 9:50 am
Bill, the main benefit would be if you were in the same car , or same class of car. While the units I have seen arent accurate enough to show accurate lines through a turn without a view of the track underneath. Too many variables that would not necessarily help someone in a lower class car or tires / suspension etc. All of these variables affect turn in , braking etc. IMHO.

Dan,
That is why I suggested the car, class, etc. be listed (see the original post) with the deposited data. The G2X and IQ3 use an external GPS antenna and map the track to 0.1 ft (1.2"). It is also possible to drive the track (presumably at lunch) on the inside and outside to get the width down to the same accuracy. You would only have to do this once for each track. If your racing line is less than 1" then you're right, it wouldn't be accurate enough for you. With regard to 'help', I would think it would be as much help as an instructor who wasn't familiar with your car, tires, suspension, etc. but who still gives helpful guidance and advice. The advantage of having the data is you could study it later rather than trying to remember in detail what your instructor was telling you in the 'heat of the moment'. Since it deals with the specifics of your particular car and how you're driving it, it should be more helpful than the multitude of books that describe how to autocross and race in general terms. Even though you might be comparing your results to faster drivers in different cars I think you can learn a lot from the comparisons - at least, I have. The intent would not be to 'program' you or your car to act identically as someone else but to broaden your understanding of the dynamics of driving fast. All that said, from the lack of interest in following any of this up the majority apparently feel as you do.

-Bill

[Jad]

FYI, I highly recommend the data acquisition systems and learn a lot from them, BUT don't believe the 1.2" accuracy. I have used several different systems and the 'drift' that occurs during the day can be several feet. That is not to say the data isn't useful or can't be adjusted, but even our AIM system which is much better than our G2X was, isn't perfect.

[mrondeau]

FYI, I highly recommend the data acquisition systems and learn a lot from them, BUT don't believe the 1.2" accuracy. I have used several different systems and the 'drift' that occurs during the day can be several feet. That is not to say the data isn't useful or can't be adjusted, but even our AIM system which is much better than our G2X was, isn't perfect.

+1 on the accuracy. The biggest benefit is being able to see your speed, lateral g-forces as well as acceleration and braking through various sectors. I've found that just knowing my speed at a particular apex point is a huge help and will allow me to pick up time on the track.

[gulf911]

Dan,
I would think it would be as much help as an instructor who wasn't familiar with your car, tires, suspension, etc. but who still gives helpful guidance and advice.

-Bill

Hi Bill,
I would have to disagree. The instructor is in your car and if he knows 911's and your tire type, suspension etc. he can tell you what to do and what is wrong and how to improve based "on your car". Its like me riding in my car , then using that data to show your braking points , turn in etc. I do see your point of using same class and mods to help somewhat on a faster driver. If you could get an instructor/faster driver to drive your car for a few laps and overlay the data this is where you would benefit most IMHO.

Whats the specs of your car?

[Steve Grosekemper]

All good points.
I have gained the most from my G2X by driving the same car with different drivers as well as comparing my own data against itself at different sections of the track.
For example if you have two similar heavy braking sections and one has a 1.2G braking force and one has a 1.5G braking force, the problem is probably the driver.

Several years ago I compared all my braking sections at Button Willow. (The track I have the most number of laps)
They varied from 1.0-1.5g. using the G2X speed/g-force graphing and the on board video.
I found out I was just not braking to the cars full potential in all braking zones.
The G2X told me which ones I was doing well with and which ones I was being a big fat baby with.

The next morning I went out with an unchanged car and took off 2.5 seconds from my best previous day lap times.

At Spring Mountain several years ago I compared speeds graphs with a car much faster that what I was driving.
I noticed transition sections that I was coasting and he was full gas.
so I stole his better line and picked up a second on a track I thought I had nailed.

However you do it comparison is your best teacher with data aquisition.

[Bill]

Re: Data Acquisition
by Jad on Mon Jun 03, 2013 8:15 am
FYI, I highly recommend the data acquisition systems and learn a lot from them, BUT don't believe the 1.2" accuracy. I have used several different systems and the 'drift' that occurs during the day can be several feet.

Re: Data Acquisition
by mrondeau on Mon Jun 03, 2013 10:41 am
+1 on the accuracy.

It is very interesting that Jad and Mark appear to have a way to assess the absolute accuracy of GPS units in data loggers. The GPS location of the car on the track for any GPS-based logger is relative to the car and its absolute position on the track down to the error limit of the unit. Simply running laps and comparing that data is not sufficient to have any idea of accuracy. For that to work it would be necessary to have the car follow exactly the same line to within a few inches or whatever the error limit was, on repeated laps and then compare the data. I don’t think that’s actually possible. I'm not sure what the 'drift of several feet' Jad is talking about - the whole track, parts of the track, or what. Also, how was that determined - were there absolute reference points? I would be very interested in the methods they used.

-Bill

[Jad]

Nothing that fancy, but when you overlay a morning session and an afternoon session and all the laps in one track session are a car width off all the laps in another session, including driving down the straight setting up for the next corner, I am assuming it is GPS accuracy, but it could be something else I use the speed and g's data more than the position data as it seems more accurate and useful for me.

Either way, as Steve said, every Saturday night I carefully go over the data and compare speeds run to run and event to event between my dad and I and we both improve significantly overnight. The systems are extremely valuable to improve your driving and times. It is work and takes time to analyze, but well worth the investment if you are inclined to use the data you get.

Best instructor there is, even better if you share your car with another driver for a session to compare.