How important is the feedback from a F1 driver?

I would say driver feedback is next to worthless.
Data aquisition remembers all. Drivers I know actually change their styles to mirror data on where they are driving slow or fast according to logged data.
The Mark Donohue days are over.
The fact is if you want to be fast in the lower formulas, getting a data logging car and comparing ones styles to faster drivers is where its at.
As an example F Continentals are fast.
On paper, Tube F Mazdas should be slower, yet are almost always faster because most have Data logging. Drivers learn and compare to squeeze every bit of speed from that info to go faster.
The horse is now following the cart.

 

Telemetry can make a car "generic quick". But it takes feedback to make the car suit the driver.

Every driver has his quirks. Some are hard on tires, some work the brakes too much, some push the revs higher. Making an F1 car suit the driver's individual quirks will make the car/driver combination faster than putting every driver into a generic "spec" car.

Otherwise you're in Formula Atlantic.

Whatever makes Kimi quick is something he does with the car. Can the car stand up to it? Does even Kimi know what it is? And could that have been the reason for Kimi's poor "luck" with McLarens?

I suspect one of the reasons that Schumi went so long without a mechanical failure is that he drove like a robot -- every corner the same each time, lap after lap. Once you dial the car in for that, it doesn't break from it.

 

I think that driver feedback still has a place in Formula 1 racing. Telemetry and lap data can only go so far. I still think that it takes at least some input on what the driver actually feels in the car that the track data isn't going to be able to communicate to the engineers.

 

To take it to the next level, you really need the driver to be able to communicate and tell the engineers not only what is wrong, but how to make it better.

 

Hi,

My take is that you've got to have a driver who can reliably, and within very few laps, be able to take the car to the limit, but *not* over it [You don't last in F1 with a reputation as a wrecker, as we know]

*How* they reach the 'limit' (aka lap time) varies from lap to lap, and any car setup changes between runs don't change things consistently around the lap - Sure, they have the 'raw data', but I like to believe the driver is still the guy saying "doing x should have been an improvement in corner y, but I screwed up" for example....

However, I think the *real* driver feedback comes in those epic 'driver debrief' sessions. - They're looking at printouts/displays/who knows what (ie, drowning in data) with details from the session - The good ones know how to analyze it and improve - The 'bad' ones get confused and drown in the data.....

As usual, just my 02c,
Cheers,
an

 

I remember reading that they get many many gigs, possibly terabytes of data from each test session. So much information, and that it is pretty worthless if they do not know what to look for. I think that is where good feedback comes in.

 

I think Michael Schumacher showed that drvier input is quite important. The car today still has remnents of his influence.

 

Why don't we ask a current driver and engineer?

 

Tell that to Schumi.

 

Give me a second while I call Jean Todt. He said that the feedback is very important, and that Michael is very good at giving feedback and thats what made him the best.

 

Not at most tracks. Check the results from this years run off's or any of the previous three or four years worth of data from Mid Ohio. The FC's were a full 2.5 seconds per lap faster. The fastest qualifying FM in 2005 would have placed 13th on the grid for FC that same year. Look at lap times from Watkins Glen Nationals, Lime Rock or VIR and it's the same. Fastest FC's were still 2 to 3 seconds a lap faster over the past few years.

Andreas: You hit the nail on the head. The value of driver feedback is that the driver can give INPUT about how to INCREASE the limits of the car. Any set up has a 100% usable limit (not the driver but the car). Telemetry, even as advanced as F1 can only show if the driver has reached the car's limit.

It CAN'T make recommendations on how to improve and increase that limit. The driver can make "subjective" recommendations on how to increase the limit on a particular corner. Some driver's at the F1 level are very good at this other's don't even come close.

Second, telemetry can't show how a car will wear tires out over the course of a given period. For example, the perfect set up at the beginning of a stint may mean fastest lap times at the onset, but toward the end of the stint a driver may loose valuable time once the tires are on the downward slope of wear.

Perfect case in point is the 03 season where Rubens sometimes had better setups and was setting faster laps early on in his stints but then would be badly behind in terms of lap times on one set of tires. Neither telemetry nor Ruben's knowledge clould keep him ahead of MS on the same set of tires on the same stint. Some of this was put down to varying fuel loads, but Badoer in an interview back in 03 said something to the effect that MS would pick a better set up in tesing that was a tenth or two off Ruben's pace at the onset but over the same period for the full stint MS was quicker overall by a few seconds because his setup was better for the entire stint.

The idea that driver input isn't valueable at the F1 level is rubbish. It's even more valuable at this level because the difference between 1st and 5th is so close.

 

Insight and feedback of true quality - along with the abilty to extract performance from the car - is what separates good drivers from great drivers.

The key point about telemetry and data is that it can only tell you 'what'. The driver, who is most closely linked to the machine can look through this data and add the important 'why'. 'What' without 'why' is next to useless. Context and interpretation is everything.

 

Driver feeback is still vitally important. They are the only person that really knows what the car is doing. When a computer drives the car then you no longer need that driver feeback. I think what has changed is that the driver now has less input in how to fix whatever problems the car might have. Back in the day the driver might suggest a change to be made to the car and the team would go in that direction. I think that has gone away and this might be the reason why guys like Kimi won't have too much of a problem. All they have to do is tell the team what the car is doing, that's pretty much all, if Ferrari ever loses it's direction setupwise however, they might be in deep trouble.

 

Telemetry shows you what the car is doing and what the driver is doing at any given point on the track. It does not show you if your wires are lose or if the car is lose if the car has over or understeer. It will show you speeds, shocks, brakes, gas, Gs, etc.
Telemetry is very good and very important but it is useless unless you get the drivers input as to what the car is doing that then you can look at telemetry and make adjustments to the car.

In Daytona we had a pro-driver, Mike Johnson running in one of our GrandAM cars. His driver feedback was excellent. We were talking for 10-15min after a session. Sometimes longer than he was actually in the car. Based on his input and explanations I went to our data guy and we looked at the car data. From there we made adjustments to the setup. Tried different things and bingo...solved his bus-stop problem.

To give you an idea, telemetry will tell you where a braking point is that you use in a corner. You then overlay that to another car and driver and see where that guy is braking in the same corner, how much more speed he carries etc. From that you conclude YOU can go deeper into a corner. Well this is where telemetry ends. You try that and find out that your brakes are not the same as the other car and therefore you ...bang hit the wall but by simple telemetry you should have been able.

 

This pretty much sums it all up.

 

I think that maybe the biggest roll in feedback is in determining the driver's set-up preferences. With different driving styles, they need to interpret the feedback and data and tailor to best suit their abilities. I think that's why telemetry will always be considered "binary" or "generic". Because until the cars are completely automated, all the data in the world still has to be processed and put into action by a human being.

 

i think nothing beats human feedback...

 

The part of the communication between the engineer and driver is the way how driver can drive. I don't mean now the talent, but rather style:
to get accurate, comparable logging data, you have to be very systematic driver: as to say, to be able to do precisely same driving result lap after lap.
There is no point if you are terribly fast, but each lap you do is different than other. This is where it comes why Schumacher was excellent. Kimi can do that one as well. The way he does it differs from Michaels way, but the tolerance when driving fast precisely identical laps one after another is the part of the development talent. Which obviously differs from racing talent. These skills can of course be included to the same person. He is the one we should call professional.
True is, that if you can do systematically and then after discussion with engineer change systematically your lap routing and rhythm, engineer can help. There is no point of wandering allowed.

 

Andreas:
it is the age old question if a robot can ever "feel". All the information you can get will not tell you how the driver "feels" the car. That is something only a human can tell you. If it were simple as telemetry all we have to do is eliminate the drivers in the first place and let the engeneers run the cars from the computer behind the pit wall.

To tell you on the problem I had in Daytona.
All the telemetry could not tell you there was a problem in that turn. It did not feel right for the driver. Of course you can get all the information from sensors but what do those actually tell you? Nothing. You need the input of the driver.

If you go on the Shell website you will find an interview with the race engineers. It actually tells you a lot about how much and what the driver tells them and what they do with the information.

 

Here is the link:

http://www.shell.com/home/Framework?siteId=ferrari-en&FC2=/ferrari-en/html/iwgen/downloads/video_clips/pitstop/zzz_lhn.html&FC3=/ferrari-en/html/iwgen/downloads/video_clips/pitstop/ferrari_pitstop.html

 

I don't think it matters if you don't care whether you win or not.


One person I would like to have answer this question: Michael Schmacher.

 

Telemetry will tell you which way to go in adjusting a car or driving style if you let it.
As far as car set up with an infinate number of adjustments as in a modern professional formula car, good data analized by a good engineer is paramount.
A driver can tell if he/she likes more oversteer, wing, whatever. But the data will tell you if that made the car/you faster and hopefully why.
I would be interested to see the results of a modern formula car set up with driver feedback only.

 

I will always remember an article reveiwing the driving style of Jody Schecter with others and I believe one of those was Mark Donohue; the great feedback/analysis expert. They measured corner entrance, midpoint , and exit speeds. They tracked turn-in, apex, and exit points (the line) The final ANALYSIS was that JSs speed (faster) was because of a different driving style.
Within a year all competive drivers adapted to that style.