Data Acquisition Tools

After test-driving a 944 Spec a couple of weeks ago, the owner of the car sent me the TraqMate data it recorded. I am now reviewing my sessions and trying to see what information I can learn about my driving and where I can make improvements. Obviously good data acquisition programs offer a lot of value, but it's a bit daunting to use.

I would love to see a GRM article on this topic. Explain how to get the most out of information provided by data acquisition software. What to look for on where to make the biggest improvements.

Any of you other smart folks who are more experienced, I'd love to hear advice on what best to examine when reviewing data.

Did the car owner give you data from one of his fast laps for comparison? This is where a lot of bang for your buck can be found by looking at entry/exit speeds of segments where he's faster, then figure out how he did it (sometimes you can get this from poking at the data, sometimes video, sometimes you have to ask for the super secret line).

If it's just your data from a session, you can look at all of your times and look at variation to give yourself a rough idea of how consistent you are. You can also look at variation in total lap time, but if there's a moderate amount of traffic you won't get laps that are within tenths of seconds. Instead, look at your segment times and throw out obvious outliers, or even better, use video to identify the throw away data.

Finally, trackpedia has some ok info about this stuff: http://www.trackpedia.com/wiki/Data_Collection_and_Analysis http://www.trackpedia.com/wiki/Traqmate

For all you guys with a GPS in your tow car: if it has a decent battery, you might want to consider zip-tying it into your race car for a few sessions.

I do that with mine and after each session I review the data screen. It records top speed and distance traveled. Then I zero it out and see how the next session differs. It's not real useful, but it's especially interesting to see what your top speed is between sessions. This Summer my Sprite recorded 101 MPH on my GPS (at BeaveRun/PVGP). That's the fastest (recorded) I've ever driven it. I'm going to The Glen in Oct. I'm hoping to see 102

The key to understanding all data aquisition is knowing what it's supposed to look like. And I do mean ALL data.

Like what I do for a living- I take many, many MB of data running various tests for a new car- primarily emissions- and while there's some degree of experimentation, there's a lot more knowing what I want it to do. So I look at the data, and make adjustments based on that.

IMHO, same thing for driving a corner- you need to know what you want the corner to look like on the data- speed changes, brakes, gas, G's etc. There are a ton of book out there on how to take a corner properly, but in addition to that, having a fellow driver to compare to is a big deal. With multiple pieces of how to take a corner, you can change things that will hopefully speed things up.

Eric

The best value to a GPS tool to me is definitively answering the "Which line is faster" question. I'll use the outer loop at WGI as an example...

Using the tach alone to measure exit speed indicates that a wide entry and middle of the road line at open throttle yields a faster exit... and for years I drove it that way until a friend loaned me a Traqmate and helped me take a lap apart. When you look at the GPS you see distance, time and speed. If you enter from the middle and keep tight you exit slower but travel less distance.... and as it turned out the inside line was worth a bunch of time over carrying more speed further around the outside. Working thru a whole lap and experimenting with different strategies took me from low 2:16s to mid 2:14s in a single weekend with no changes to the car.

Question everything about how you drive and then answer the questions.

I'm pretty sure GRM did run a feature on this a few years back. I've just started playing with it and even took a class on it last weekend.

I've got two datalogs of my own that have been instructive - same car, same track, but one's over a second slower because I'm rusty. So I've been poking around that. Having two (or more) approaches to compare as walter has done is really helpful.

You can also look at your friction circle. You should be dancing around the edge of it all the time. If it's a big blob, that's a sign that you're not consistently using the car's full ability.

For the DIY folks, this can't be beat for $25 USD:

http://www.dataq.com/194.htm

Get yourself a cheap laptop, decide what 4 things (or 8 if you use 2 of them) you want to capture, hook up the sensors, and go.

The software is great, like a 4 channel oscilloscope, except you can record, playback, stop, etc.

NFI, etc.

Carter