Control Shaping!

I have wondered for years!

Does the driver drive the car he has, or the car he thinks he has?

In other words, is the driver an Active or Reactive control system?

What defines the best drivers, are they prescient!....?







Charlie

A good car will be predictable, thus allowing the driver to be active. A bad car will force the driver to be reactive.

Very bad cars, of course, largely ignore the driver altogether.

E.g. 1999 Audi A4 2.8 tip quattro:

Enter corner too fast. Result: understeer.
Turn wheel more. Result: understeer.
Turn wheel less. Result: understeer.
Stand on accelerator. Result: understeer.
Stand on brake pedal. Result: nothing. And understeer.

I swear that car understeered in a straight line. And it had no, but which I mean none at all, zip, zero, made-of-cardboard, brakes. I've never driven anything before or since in which the brakes faded on their first application. I very nearly killed several friends making this unexpected discovery on a French D-road.

Deep. I am not sure if this forum is appropriate, but I'm sure the boys will give it a good shot anyway. I think Fatboy will give a good accounting.
I think a driver is an active, reactive and pre-active component.
I think that there may be some here who won't be able to give an opinion and here's the reason why.
I was an instructor in the airforce and some guys couldn't fly. Sure, they had been thru FPS and got their wings but they got flamed everytime they tangled with a hot shot pilot in combat training. If God didn't program the pilot with the right stuff, he would never be able to learn to fly on his own. Suppose such a guy was giving his opinion here, how would you know what he was saying was valid?
My wife can't drive to save her life and to watch her back out the garage is a moment to behold. There are some guys with an engineering degree like myself who haven't any aptitude for engineering. A brilliant engineer in a lot of cases is a hopeless mechanic. Ross Brawn couldn't do Michael Schumacher's job.
Daryl Waltrip was talking about driver aptitude last week, and he mentioned that some drivers could get results from a less than perfect car where as others required a perfect car. Ryan Newman(engineer) can request specific adjustments to his car and Mark Martin couldn't and so on, but the result was the same. What understanding of the machine must a driver have and does this make him a better driver?
Obviously, some drivers have an additional skill and that is to drive round a less than ideal car. They are able to calibrate and modify their skills to accommodate the shortcomings of the car. Montoya had a 5th place car so he always came, well, 5th.
What is going on inside Newman's and Martin's heads? Apparently, complete understanding of the machine is not required.
Hitting a ball with a bat requires no knowledge of physics, but somehow all creatures can calculate and compute and execute complex corrections in motion, reactively or actively.
Raikenen enjoys feedback from Schumacher at Ferrari as Schumacher did from Barachello. He would still manage without it.
I flew gliders and often encountered various raptors soaring near me or more correctly, I was flying near them. The eagle and me both did it at will but our instincts took care of the details by initiating input and reacting to outside input.. Was this magnificent creature deriving the same pleasure as I was? I knew more about the theory of flight than he did but at some level we shared a certain instinct and exploited it and shared the enjoyment. With experience and repetition, and aptitude(instinct), I think the input is active and pre-active. The eagle and me connected at this level.
Rally driving is more reactive than F1. The best drivers have a multitude of skills and aptitudes that can't be learned. Schumacher lacked spatial orientation and situation awareness that would have gotten him killed in a dogfight and his chops proved that. In a fair fight he would loose every time. I said fair fight. His mechanical aptitude and speed and a powerful competitive, aggressive determination and self belief won him championships and maybe those are the most important components of all. Will.
Kyle Petty stays at the back of the pack because he has no skills to speak of.

Originally posted by Charles E Taylor
I have wondered for years!

Does the driver drive the car he has, or the car he thinks he has?

In other words, is the driver an Active or Reactive control system?

What defines the best drivers, are they prescient!....?







Charlie

All drivers are active. They all provide feedback and the crew respond. That's the easy part. Case in point: test drivers.

The best (in terms of championships) drivers are reactive. They feel what the car is giving them, and use tricks to squeeze the most out of them.

In Schumacher's case, his tricks were mostly strategic--timing his pit stops and understanding the implications of the in- and out-laps and such.

Jeff Gordon and Tony Stewart in NASCAR are other examples of people that seem to be able to "extract more performance" out of a car than most other drivers. In their case, tire management plays a big role.

In summary:
--any driver can say it's understeering or it's oversteering
--good drivers can get the most out of a good car
--great drivers can get the most out of a midpack car
--no driver can win in a backmarker car no matter how good

Originally posted by phantom II
Fatboy will give a good accounting.

If this means I'm supposed to agree with everything you wrote, then I don't.

A circuit driver uses active, reactive, and predictive components to his driving and it varies widely depending on the car, conditions, the track, who is paying for crash damage, etc.. Largely, they drive the car they have, especially at the higher levels of racing. At the lower levels where car control, bravery, stupidity, and experience widely vary, then this isn't as clear. No one is driving a really 'good' car and very few if any are really driving the car right, so you can get quite a spread in the field.

PII spouts about this driver and that like he knows them. He doesn't. His views are myopic observations by someone who has never done it. Salt to taste.

What FB said. At the top levels with real professional race drivers you will tend to get the same performance every time out (within normal human variation of course.) That is pretty much the definition of a real professional race driver, come to think of it. As FB says, they drive the car they have. If you make them drive faster they will crash. However, it is still enjoyable to see the veteran have the special day once in a while.

But in the lower and developmental series the mental component is far more crucial -- often more crucial than any single adjustment on the car. So it is important to keep the driver's head in proper adjustment at all times, just as with the car.

Old crew chief trick: first test day with the new driver, give him the car with something broken on it. Closed-up spark plug gap, dead shock, something like that. Easy, but not too obvious. When he comes back in to tell you about the problem you lean in and listen intently, ask probing questions, nod sagely at his astute observations, in short treat him like Gil de Ferran. Then swap out the part in two minutes and send him back out. Fixed it the first time, and in moments... why, you must be a mechanical wizard. You have now convinced him that you are a genius and together you are destined for great things. Of course this feeling will last only until you run into something you can't fix, but it is important to get things off on the right foot. I can't believe I am giving away this stuff for free.

The flip side to that unfortunately, is I've always taken a dim view of 'intentionally mis-set a corner to see if he can find it' because to me it establishes you don't trust me. If you want to test my sympathy, tell me something is changed (but not what) and I'll try to figure it out. If I can't feel it, I'll be honest enough to tell you.

The idea here is not to test or challenge the driver's diagnostic skills but to give him confidence in them, and in you. This is set up so that he identifies the problem and you fix it. It's like a corporate trust-building exercise at the company retreat except you are not passing out corny photocopied scripts LOL.

I totally agree: No engineer or crew chief should ever do what you describe. That guy should be fired and then shot. He is SUPPOSED to know 10^6 more about the car than the driver and if he is that insecure he does not belong in the business. The goal is to establish communication and trust, not prove who is the smartest by playing games. I know some teams use such ploys in driver evaluations... usually it is a pretty fair acting-out of how the relationship will eventually fail.

What did I tell you, Charles? My best friend FB would give a good accounting and so he has. I can always count on my buddy. He is my Stradivarius.

Originally posted by Fat Boy


If this means I'm supposed to agree with everything you wrote, then I don't.

A circuit driver uses active, reactive, and predictive components to his driving and it varies widely depending on the car, conditions, the track, who is paying for crash damage, etc.. Largely, they drive the car they have, especially at the higher levels of racing. At the lower levels where car control, bravery, stupidity, and experience widely vary, then this isn't as clear. No one is driving a really 'good' car and very few if any are really driving the car right, so you can get quite a spread in the field.

PII spouts about this driver and that like he knows them. He doesn't. His views are myopic observations by someone who has never done it. Salt to taste.

See my thumb? Gee, you're dumb.

Control Shaping

An interesting term. Used a lot in engineering as a way of describing certain types of control laws.

In the context of Motorsport, further discussion of this topic might be relevant. In a simplistic way it could be used to explain the different ways a driver drives as his car changes characteristics throughout a race.

But there is much more to it than just adaptive driving.

There are many instances with drivers’ post event descriptions of their cars performance which fail to identify even gross problems. Do they “Control Shape” their way around these defects?

This is particularly evident when good results are obtained despite the deficiencies. Some of these deficiencies are very easy see; (Stuck in gear, no brakes, front wing missing, collision damage etc) some are very much more subtle. (Broken engine mountings, disconnected anti-roll bars, leaking shock absorbers etc).

Sometimes the treatment of this condition is very dismissive and sometimes psychological, but it does not resolve the fact that the car is often a lot less than optimal.

Control shaping (adaptive driving) is one possible contributory element in the driver not picking up these defects. Perhaps he is not aware he is doing it. It is probably instinctive.

It is often found that on another day the same elements (Car, driver, engineering) will combine to a situation where the slightest deviation from optimal will produce a very depressing result.

Perhaps there is much more to know about this effect. Can the very best divers adapt more widely than others?

Perhaps also it would be interesting to consider a second placed car which has just been beaten by a deficient car. (perhaps with a collapsed front wing.)

The usual response from the winning driver, asked what his (sub-optimal) car was like, is often; “I don’t know, I was too busy winning the race”

But what is the response of those beaten by an obviously damaged or crippled car? They do not seem to have much scope to blame the equipment.

I would like to explore how "racing engineers" deal with these issues.


Charlie

Is this an example of adaptive/reactive driving? The vehicles are certainly suboptimal.

http://www.nascar.co...et.exclude.html

My experience as a retired trouble kid driving underpowered and much less than optimal handling cars is that, close to the limit, it's like curling (the winter sport). You throw the thing into the corner, oversteering as hell ,that's active control, and 90% of the deal is done. You can rub the ice to try making it go into the right direction, and you have 10% to fix as a reactive control.

Originally posted by benrapp


1. A good car will be predictable, thus allowing the driver to be active. A bad car will force the driver to be reactive.


2. E.g. 1999 Audi A4 2.8 tip quattro:

Enter corner too fast. Result: understeer.
Turn wheel more. Result: understeer.
Turn wheel less. Result: understeer.
Stand on accelerator. Result: understeer.
Stand on brake pedal. Result: nothing. And understeer.

1. I am lucky to have driven many comp cars due to my knowledge of Weber carbs that put me in demand until injection took over, as well as spending half my working life running workshops and test driving many cars on a daily basis. A good car is one that you can tell what to do, it reacts to your input and will do so predictably and I have always called them action cars. Of course a reaction car is one that takes the lead and you have to just point and sometimes hang on (not a good feeling either sometimes) but by no means does it mean a slower car. History has proven that there are drivers to suit both, Keke Rosberg and Alain Prost in the McLarens probably a reasonable example. The McLaren was a designers car and needed a smooth accurate driver like Prost who was totally capable of being a reaction driver but near the end of the year when Keke, who was an action driver, was finally given freedom to loosen the car up, he flew.

2. What do you expect when you hang great lumps of metal so far foward of the front axle line, watch any rally Quattro and see what the fronts are doing.

Originally posted by phantom II

Rally driving is more reactive than F1.

Dont believe that for a moment, thats just what it looks like from the outside.

Rally cars are actively and intentionally put out of control to be in control. Driving a reactive handling car in a rally (dirt) is a nightmare and likely to fall off the road very quickly.

Originally posted by cheapracer


Dont believe that for a moment, thats just what it looks like from the outside.

Rally cars are actively and intentionally put out of control to be in control. Driving a reactive handling car in a rally (dirt) is a nightmare and likely to fall off the road very quickly.

Seriously. That's why they have notes! They plan ahead!

Your comment reminds me of the ol' "Scandinavian Flick." Looks crazy, but I guess it's effective, at least on turns of increasing radius.

Let's examine that 'moment'. The rally car is in slow motion compared to if it were on tarmac and fitted with slicks.
Like drifting, during the slide, the driver reacts to conditions and his inputs commensurate. Sure he has initiated the turn 100 yards in advance, but during the time that he arrives there, the input is vastly different to the tarmac situation. Like driving on a skid pad, if you are going for speed, you will have to have sharper and quicker inputs to react to any slippage that might occur than if you drift on the same radius. The input and skills required are completely different. I take my 500 hp Vette around my favorite turn which my best friend, Fatboy knows, either in a powerslide in a lower gear or fast in a taller gear. It requires vastly different approaches and planning. Be it show or go, both methods are immense fun but I say, the drifting is purely reactionary even though you are aiming the car purposefully toward a pre determined point.. You don't have that option in a rally car except for France. Maybe its a question of semantics because I don't know what a reactive handling car is.

Originally posted by cheapracer


Dont believe that for a moment, thats just what it looks like from the outside.

Rally cars are actively and intentionally put out of control to be in control. Driving a reactive handling car in a rally (dirt) is a nightmare and likely to fall off the road very quickly.

An interesting topic.

Does anyone remember Andy Green and the rear wheel steer Thrust 2 record breaker? He had to avoid making control inputs within a certain frequency band. He wrote that he either made rapid and abrupt steering inputs or very slow, very gentle ones- nothing other than that. He seems to have figured out that he had to do this after a few "moments" experienced driving Thrust 2. A very brave, thoughtful and skillful man.

As an aside, consideration of what Phantom2 reports, about some pilots not being successful in certain situations, and applying it to race cars leads to the intriguing possibility of constructing race cars without a human driver at all. Why not race those against race cars with human drivers? At first the humans will win but soon enough they won't. Perhaps this would lead to technologies that could be useful outside the racing arena.

Regards

Gerald

Originally posted by Charles E Taylor

This is particularly evident when good results are obtained despite the deficiencies. Some of these deficiencies are very easy see; (Stuck in gear, no brakes, front wing missing, collision damage etc) some are very much more subtle. (Broken engine mountings, disconnected anti-roll bars, leaking shock absorbers etc).

Sometimes the treatment of this condition is very dismissive and sometimes psychological, but it does not resolve the fact that the car is often a lot less than optimal.

Control shaping (adaptive driving) is one possible contributory element in the driver not picking up these defects. Perhaps he is not aware he is doing it. It is probably instinctive.

First thing that comes to mind for me is there is a wide variation in technical expertise among drivers.

Also in sensitivity. Some seem to have very little but are very fast all the same. At the other extreme, some are like the squirrel with the oversized ears -- gone mad from hearing every noise in the forest.

Originally posted by Gerald Ryan
the intriguing possibility of constructing race cars without a human driver at all. Why not race those against race cars with human drivers? At first the humans will win but soon enough they won't. Perhaps this would lead to technologies that could be useful outside the racing arena.

Regards

Gerald

Big Blue II?

We're getting there--the US DOD has sponsored a yearly event that pits 100% independent vehicles against each other over some predetermined area of desert. Basically the Paris-Dakar rally for robots, on a much smaller scale.

The first year they had it, no contestant even came close to finishing. But apparently they are finishing with some regularity these days.

Oh yeah, and I think it's been proposed with some degree of seriousness (as a safety thing) to make all racing cars remote control.

Yeah but no.

Reminds me of Mansell, I honestly think that he was one of the fastest drivers of all time but only when the car was perfect and when he thought something was wrong, holy cow, what a reversal!

I would say that Benrapp's observation is the most correcr. I do think however that since a driver initiates any of the the motions that would result in over/understeer etc. that makes him/her active, and when they are making corrections to however the car reacts then the driver is reactive. A good car needs him/her to never need to react. And a good driver can always react appropriately.