F1 "space map" concept

I'm reading the current Top Gear magazine, and they have an article on Schuie that has some info I've never heard before. Apparently the Ferrari F1 car (and presumably all F1 cars) has a "space map" concept. The car knows where it is on the track at all times with GPS inputs, and it is programmed to adjust the throttle map, the E-diff, and traction control to be optimized for each corner. The parameters of the space map can be controlled from the cockpit in real time (those multiple manettinos), and Schuie is constantly tweaking to get the best possible lap time. Ross Brawn and the team are consistently amazed at how much attention he pays to this process in the middle of qualifying and racing, no other driver they have worked with has the spare brain cycles available to analyze and then make changes under those conditions.

Amazing stuff, and I've never heard it mentioned before.

Gary

 

A little off topic, but Ross once before commented about being unsettled by how chatty Schumi is on the radio while putting in hot laps.

 

I have always wondered about the space map concept when they used to show the displays of where everybody is on the track. I always wondered, whether that was just based on calculations using the S/F line or whether that was real live data. Apparently they seem to know where the cars are although I don't really see how that works, unless it is differential GPS.

About MS tweaking the buttons: We heard that before and I can imagine him doing that in a hot qualifying lap where he has to squeeze every bit of time out of it. However if you watch him on the in board camera you rarely see him (or anybody else) tweak anything. They adjust brake bias with depleting fuel amounts, but not much else me thinks.

 

This is interesting and also amazing how the gps data can be used to make adjustments on the car on milliseconds basis going into a corner or exiting a corner. Is this done by the driver or computer? This must not be the same satellite gps we are familiar with as delays are enormous to communicate with the satellite. I still don't see how it is possible. But if it is possible through some local cellular base network it may be illegal since they make those adjustments through the computer especially if differential and brake are being tweaked. Isn't that technically traction control in conjunction with something else that controls the car's behavior...

 

Another question I have pertains to whether this is constantly in use during the race, only during qualifying, or toggled on/off in relation to racing situations. On an open track, when an all-out run is needed, it would be pretty simple to program something like that. However, racing situations are very fluid, which leads to variable racing lines, changing entry/exit speeds and so on. That would seem pretty difficult to program for.

 

Would it be possible for you to scan this article and upload it on to the internet?

Thanks

 

The implication was it used during race as well as qual.

Gary

 

Matt -

I don't think it has to be that fine. When the GPS senses it approaching turn 1, it selects the map for that corner, and I would guess keeps it until turn 2 is imminent. I would bet they have some software that keeps track of position of the car by speed from a known position and doesn't need constant updates from GPS, just periodically looks to GPS to validate the car's position. I'm guessing here, but I bet it is something like that.

Gary

 

I'll give it a try.

Gary

 

PM me with your email and I'll send the scans to you...they are unreadable if I shrink them to the max file size for FChat.

Gary

 

I would agree with your guess, but it still doesn't explain it fully: The point of this system is to squeeze out the last ounce of performance from the car by tweaking it in certain ways. That requires a high degree of precision in where the car is on the track. However the approach you mention lacks precisely that precision.

Even differential GPS still has a resolution of about 2 meters I believe, so it would still be a crude approximation for breaking points, apex etc.

I can't help it but think this is hype. It makes more sense to me to build traction control and other goodies into the engine management etc and feed the board computer by sensors already in the car. I might be wrong, but it reminds a lot of the speculation mid season about Renault's secret launch control that "senses" the lights change etc. whereas Ross Brawn once commented he believed that one to be "simply" the sum of all the parts of a great design.

 

I don't have any insider knowledge, but I know someone who may. I'll run it by him, it's possible this is all B.S.

Gary

 

A friend of mine who has been involved in some major-league racing teams (including CART and IRL) has this to say, although I just emailed the scans to him, so maybe more later.

>>Sounds like a wee bit aggressive interpretation of what they are using (from the description, it borders on almost being an active system). All such ECM and data systems are driven by positioning; trackmaps are the core of said systems, and in general terms, are created by wheel speed and g sensors.

As for a "space map," it would require a TON of inputs to make it effective--the position of the car on the track has little to do with any interpretive work on the computer system's part (Schumacher could be 15 feet offline and scrubbing tires--the system wouldn't have much to optimize solely on track positioning)--feedback from a wealth of steering and suspension strain gauges, laser ride height sensors, and infra-red sensors/tire pressure sensors would be required to generate all of the "feel" a computer system would need to make such a system work. Complete tire modeling and physics info from the tire manufacturer would also be required (Firestone used to provide us with said info to use for Pi's Sim software).

Granted, with a mapping system in place, and making use of the variety of sensors I listed (I'm sure there are more I know nothing about) and tire info, you could build a system that could use all of those inputs to deliver a certain percentage of "feel" for the driver to manipulate (in essence, it would tie a number of individual controls into one--things like percent traction control, how harsh/soft the traction control is activated, diff ramping aggression/passivity, brake bias, etc,) into one integrated control measure.

As it is now, most of those controls are set individually, and are as I've known the Ferrari system to work. Maybe they have come up with a better mousetrap--knowing them, they'd still provide both individual and integrated controls for maximum variables.

I'm not saying this is correct, but if I had to think about how it was done, this is the only way that makes sense.<<

 

I've read the article too and it sounded very convincing to me. Why would Ross Brawn make something up like that at all? That doesn't make sense to me.

 

To send the competition into the wrong direction.

If you read the complex response from tdf360's friend one realizes how much effort and resources it would cost to build such a system. If Ferrari's competition tries to do that, they'd be busy for the winter without getting anywhere.

 

I spoke to a couple of friends today morning who are involved with F1 and they were not willing to disclose too much information. This makes me think that there is some truth to the article.

However, I too have my doubts about the degree to which such technology exists and is used. It has been known that after 2000 Ferrari developed several new technical devices with development partners such as Infenion Technologies, Acer, Magneti Marelli, Momo and Bridgestone.

These were the break throughs in electronics development that gave Ferrari the edge in tight situations. It is very possible that this "space map" concept is one of them.

 

did the article specifically say it was satellite based positioning?

there are plenty of ways to do this in a constrained geography like a race track without a satellite.

never mind that they could perhpaps (perhaps) do it on dead reckoning if the feedback re: wheel angle and speedometer were accurate enough, right? once a lap you could re-sync with a local or satellite triangulation?

doesn't seem that impossible - just money.

though i think it's cheating

doody

 

hmm, if it were an article in Autosport I could understand your point. But Top Gear magazine?

My impression from the article was that they don't use the system to be up to date constantly, rather get some presets for every corner and then work from there (with the manettinos). Also when conditions change, Schuey changes the settings and these changes are recorded into the space mapping for the next time around.

Reminds me of the button "SC". That's probably a preset for a Safety Car situation, correct?

 

In this interpretation it would make sense to me and sound believable. But it also means that it isn't that revolutionary either.

 

No, the GPS theory came from a friend as we discussed the article, I should have been clearer on that. I thought it sounded plausible. I'd be glad to email scans of the article to anyone who wants them.

Gary

 

Gary,

Please shoot me a copy, I may be able to reduce and post it.

Matt