Intro to Data Analysis presentation available

[924Guy]

After many hours of work, review, editing, and feedback, I've finally completed an Intro to Data Analysis presentation. This was created to help explain how to analyze data from your data acquisition systems (regardless of brand) with an eye to improving driver technique. It's good reading whether you have been using a system for a while, just got one, or are still trying to decide if you want one.

Here's the direct link...
http://vaughanscott.com/AIM/Data_Analysis_Intro.pps

And a reminder - we will be covering this in full presentation mode at an upcoming Waterford Hills monthly meeting (as mentioned in the other thread)...

[jlucas]

Vaughn, That is excellent. I'm sure there will be many people vary appreciative of your sharing.


Couple of questions & observations:

pg 3) why do you think wheel speed/beacon type has better repeatability/accuracy? I do both and I had not heard that before.
pg 4) "resolution?" What are you referring to? are you talking input logging rates? speed trace resolution due to GPS update rate? GPS position resolution? (side note: Many people do incorrectly understand the absolute positional accuracy of GPS data available in this price range.)
pg 3) somewhere, here or later in the pg 26-27 range you might want to mention the costs to have real time in car predictive lap time feedback, or maybe even which systems are capable of having it (or upgrading to it).
pg 14) are you saying it's sawtooth because he's hunting for the the right corner speed (too slow, then to fast, then to slow....)? or is there some other reason?
pg 25) what is the answer to your question? Hard to tell from just speed.
pg 26-27) as the driver becomes more consistent the best theoretical time moves closer to something that is achievable

If I can help in any way, please let me know.

[Jeremy Billiel]

Great into and info! Thanks for sharing

[924Guy]

Vaughn, That is excellent. I'm sure there will be many people vary appreciative of your sharing.


Couple of questions & observations:

pg 3) why do you think wheel speed/beacon type has better repeatability/accuracy? I do both and I had not heard that before.
pg 4) "resolution?" What are you referring to? are you talking input logging rates? speed trace resolution due to GPS update rate? GPS position resolution? (side note: Many people do incorrectly understand the absolute positional accuracy of GPS data available in this price range.)
pg 3) somewhere, here or later in the pg 26-27 range you might want to mention the costs to have real time in car predictive lap time feedback, or maybe even which systems are capable of having it (or upgrading to it).
pg 14) are you saying it's sawtooth because he's hunting for the the right corner speed (too slow, then to fast, then to slow....)? or is there some other reason?
pg 25) what is the answer to your question? Hard to tell from just speed.
pg 26-27) as the driver becomes more consistent the best theoretical time moves closer to something that is achievable

If I can help in any way, please let me know.
[/b]

pg 4 (first 2 comments) - I've heard that the (positional) accuracy of the GPS units can (not does, but can) leave something to be desired. I don't use one, I use an AIM, so I can't comment directly.

I've not dealt with a system with predictive feedback - so can't make any comments there, I'm afraid...

pg 14 - Sawtooth not in the wheel speed blips on corner exit (that's likely just wheelspin), but rather that it's a saw-tooth, or triangular shape overall to the corner - as compared to a more bowl-shaped speed trace for the other/better examples. Of course, there's no hard fast answers here - this all has to be taken with a little grain of salt, as suits your track. It's a lot easier to get a smooth entry and exit on a big sweeper than on a short quick little jab of a corner. YMMV!

pg 25 - I don't know what the answer is, but it doesn't matter for this case specifically. If these are spec cars - is the other guy cheating? Or maybe your motor is tired, that's all? Maybe you're overweight, or poorly geared? Maybe he's driving differently in some other way, like steering inputs? Maybe your brakes are dragging? Maybe he didn't run out of NOS and you did??? It's meant to stimulate discussion and thought, not provide all the answers. This is an introductory presentation, not by any means conclusive...

pg 26 - quite true! The point here, especially for the beginners, is to take that number with a grain of salt (especially if, as often happens, some segment times are corrupted and not accurate/realistic. If I look at that, see that I can theoretically drive a 1:18 (but I'm doing 1:25's right now) - there's a 50/50 chance I could end up pushing harder than I'm ready for and stuffing it bad! So while it can help provide some motivation to push harder - yes, your car is capable of more than you're achieving right now - it can't be taken too literally, too quickly.

Good questions and observations, thanks!

[Daryl DeArman]

I'd like to soilicit opinions regarding the merits of GPS vs. Wheel speed/transmitter combination.

I see faults in both of them and would like to hear opinions regarding their strengths and weaknesses, especially if someone has used both types.

I use a low end model garmin gps when jeeping and rarely, if ever, see accuracy less than 14 feet. I can't see how this would be good at line evaluation or looking for consistency if it is possible (however improbable) to get two identical traces with a line that really varies by as much as 14'?

I have experience with an older AIM My-chron system on my prior race car (Formula Ford)...if I try a line that varies by 2', for example, through turn 8 at WSIR the difference between a 900' radius and a 902' radius over a 100 degree turn is about 4 feet. All the traces will reference me to a point 4' further ahead on the track than I really am and then suddenly rectify themselves when I cross the s/f transmitter beacon. What good is this when trying to evaluate time vs. position. What if I lift a wheel, get wheel spin or mommentary lock up? It thinks I am ahead/behind where I really am.

I am not looking for seconds, I am looking for 10ths and consistancy...

Thoughts?

[Greg Amy]

I'd like to soilicit opinions regarding the merits of GPS vs. Wheel speed/transmitter combination.[/b]

For reference only, since I have not yet taken the time to read the above presentation, my laps times derived from the GPS-based DL1 are usually within THOUSANDTHS of a second to the official transponder-based times. As in, the DL1 shows a lap of 1:02.909 and the official time is something like 1:02.912.

Further, any time I get two (or four!) wheels off on any lap, the DL1's GPS-based tracings clearly show (to the delight of my crew chief and to my own consternation) that it happened.

Pretty damned accurate, I'd say. Certainly accurate enough for what we do.

[924Guy]

I see a similar level of accuracy with my beacon/sensor setup - usually within thousands of the official transponder times.

And yet there can also still be problems with that; if you end up with multiple beacons, with an AIM system, and aren't expecting them, you can get laps split up incorrectly. It can all be fixed later when doing data analysis, but it can be a PITA and annoying. Worst case, your lap times end up getting displayed incorrectly while on-track - can be annoying when trying to set a pace, whether it's for qualifying or during a race. I also find that the beacon recorded location, in spite of 100% consistent placement from weekend to weekend, seems to drift a little, and has to be adjusted so that I can overlay laps from different weekends correctly. OTOH, if there's more than one user with an AIM system there at the track, we can group together, place multiple beacons, and get split times (as well as the overall lap time). Though I have to admit, my dash won't display them (I have a MyChron3). Perhaps the more expensive AIM systems will?

[924Guy]

One more thing... note that this presentation is intended to be a discussion of how to gain benefit out of the systems - not so much to sell one or another. We felt it would be incomplete to not say a few words about some of the factors to consider in buying one type or another, and include some vendor links, etc., but it's not intended by any means to be a serious discussion of what to get. I suspect the various magazines can do a better job of that. We wanted, more than anything else, to cover what to do with them once you've got one. Your vendor should be able to help you get it installed and up and running...

[Daryl DeArman]

Vaughn,

I really appreciate all the work you've done on this. I pointed your link to a fellow open wheeler (who'd never stumble across this site) that I've been trying to convince to get a simple DA system.

I don't have a vested interest in either type of system, currently I don't own either.

I believe they can be great tools--I am just not so sure at this point that finding out what difference small changes in line and apex points make can be discovered with either type of system.

Traction circle use, once you are on the gas do you stay on the gas? Are you turning in too abruptly and then having to correct just after turn in? Are you taking too long to build to maximum g's in a corner? Are you really carrying more speed wound tight in 3rd through that sweeper or 'bogging' in 4th?

[924Guy]

Cheers... actually, Daryl, I find the detailed kind of questions you pose to be more easily answered with a trace analysis (plot vs. time or distance) rather than a g-g plot (friction circle). But that may be because that's what I'm used to on the day job. Bottom line, use whatever is easier for you to understand...

[Daryl DeArman]

My poor writing skills didn't make it clear....I meant "valuable tools such as a friction circle, as well as answering questions such as....." I agree, the time vs position and postion vs. speed plots are easier for me to interpret.

[lateapex911]

I must be the only guy not to have Power Point.

Are there free readers availble, or do i just need to suck it up on this one?

[dickita15]

open office . org

[stevel]

I must be the only guy not to have Power Point.

Are there free readers availble, or do i just need to suck it up on this one?
[/b]

Jake, MS has a free powerpoint viewer. Just google powerpoint viewer, it comes up first. You don't have to dl'd the whole openoffice stuff.

s