Can a Human BEAT Launch Control?
Вставка
- Опубліковано 30 чер 2022
- Head to squarespace.com/drivenmedia to save 10% off your first purchase of a website or domain using code DRIVENMEDIA.
🔴 Follow us!
📱TikTok - / drivenmediatt
📸 Instagram - / drivenmediaig
🔵 Facebook - / drivenmediafb
Can launch control be beaten by a human? Launch control is amazing, it can use insane amounts of data and computing power to help you get the best start possible - and today I’m going to try to beat it.
I’ve done hundreds of competitive starts over my racing career - and I think I’m pretty good - so it’s going to be interesting to see if I can beat the car’s computer.
We have Scott, a professional racing driver who has done hundreds of race starts over his career - to go up against one Ferrari. Whoever wins is the best, I guess?
🔴 Would you like to be featured in a video with your car? Submit it here 👉 forms.gle/ZWMfzqCyDTBwJsqX8 🏁
Press enquiries: press@driver61.com
#LaunchControl #Ferrari #DrivenMedia - Авто та транспорт
Couldn't heating the car up from the first couple tries be a factor as to why he was faster at the end? Honestly I think it would of been a good idea to use launch control again at the end just to make sure that heating up the engine or tyres didn't create an impact on the results.
I tought exactly the same thing
Yep. Tyre temp did it. Pretty usless tests.
You also keep heat soaking the engine which reduces engine power.
the reason he can beat the ECU is the Human is reacting to what is happening as well as anticipating what is going to happen and thus puts a lot more gas into it as he is confident in controlling the car.
The ecu can only respond to what is happening right now, so even the tiniest of wheel slip its going to let of the gas...not always the fastest way to accelerate but better then a novice trying to launch it.
Do you think they're launching a cold Ferrari? ...
You should've did a launch control test right after you did your most successive manual test for data to be more worthwhile! Change may be caused simply with warmed up wheels for example. But in any case, awesome video, I can't tell how I love your new approach to the channel, keep it up!
Exactly, was looking for this comment.
Good call...and agreed!
If the tires were cold on the first run (with launch control), the time would have been the slowest. It would have been nice to have seen a launch control time right after the best manual launch to get a closer comparison of the performance difference.
And the last run has the least fuel (weight).
I agree. I know for a fact that those michelin cup tires really don't like being cold.
also less fuel
launch control also has to be reactive to the road, typically. as a human you can learn to control the throttle in advance, you know it's the same road you just did etc.
It's not implemented that way, but theoretically you could develop a launch control, where you save the data for a specific track at a specific condition and allow the launch control to use this data.
@@HextechLab or maybe even integrate some kind of sensors that could detect the rough road conditions ahead of the car and the acceleration, pair that with some kind of algorithm that would learn to interpret different conditions and optimise G force
@@HextechLab theoretically, with quick enough sensors and ecu, you could also have "perfect" launch control on any surface without any preset or memory, even with inconsistent grip, with launch control targetting always the most powerful launch possible, and perfect shifts, and only limit power in real time when wheel slip reaches optimum amount (for instance 5%) which would be humanly impossible to replicate. but obviously (Ferrari's) launch control is not perfect. I'd love to see the same on a tesla with possibility to fully disable traction contnrol (like the old dyno mode easter egg) Tesla's TC seems really good, beating the braking performace of the ABS on the plaid even though with the same amount of grip, TC has to fight against drag, whereas ABS is helped by drag. also now with new 3 or 4 motors performance EV's, human will never be able to match independent 4WD traction control.
Theoretically extra sensors and actuators don't weigh anything and teleports can theoretically exist
Exactly, we make predictions. This is how tennis players can return serves even though our reaction time is relatively slow
We are not AÍ we are NI, Naturally Intelligent
This reminds me of Mika Häkkinen getting his best qualifying time at the 2001 Spanish GP, the first race electronic aids were allowed again, after turning traction control *off* because the McLaren's system was set up a bit too conservatively towards avoiding wheel spin and therefore left an amount of potential grip unused.
I think that "slip ratio vs longitudinal force" graph could help showing why you want certain amount of wheelspin during performance driving.
Why is that
It does exist and it's used to calibrate landing brakes for aeroplanes. There is a coefficient that measures the amount of spinning of the wheel and the speed of the vehicle. It equals 0 in burnout and 1 if there is no spinning at all. The optimum is 0.8
@@RENO_K because the grip(force) of a tire increases with little wheel spin then suddenly decreases past the optimum slip ratio
@@kevinc.4575 man you thaught me something right here love it thx
@@RENO_K with a little bit of slip, the tyres are spinning faster without losing too much grip. The engine would be at a higher power band too.
I mean if things happened in the same chronological order as in the video, tyres might've been getting warmer, improving the time in the meanwhile. But still great video, just caught my attention
Hi Richard, we left decent gaps between each run to make sure this was constant!
@@OVERDRIVE.studios Yeah but I think you could have alternated between launch control and full manual or done the launch control after beating it just to confirm?
@@OVERDRIVE.studios I think you guys should have done one more fully automated launch control right at the end to confirm your result. I believe the tires were getting better and better over each run
How much weight of fuel it has burnt in the end? I guess it’s not that significant?
@@OVERDRIVE.studios huh, should've mentioned that in a video ^^" Exactly my thoughts about warmed up tires. Would've been nice to make another launch control test to confirm the results anyway~
“There’s is no such thing as too much power, only not enough traction”
- Carol Shelby
Of course, we have to also take in consideration the consistency.
A driver can always absolutely screw up the launch.
The computer will probably bang on the launch time after time after time
Tesla launch tests for example, its just the same rocketship kind of launch over and over again
Keep in mind certain car manufacturers also put a limit to the number of launch controls you can do with the car before you will need to change the clutch to reset it. Including cars with DCT gearboxes. My Clio 4 RS Trophy is software limited to about 3k runs.
Oh, and generally the launch control mode that gets the publicised accel figures is not actually available on the normal settings, and is only accessable via a secret series of inputs! This is because the optimum launch requires so much clutch slip as to be potentially significantly damaging to the transmission. The launch control you access as a normal driver will always protect the clutch packs above getting an absolutely optimum launch because very few people use LC to do more than try to impress someone at the traffic lights, where the difference between an 11 sec qrtr and a 10 sec one is simply not detectable....... When i set some 0-60 times for a famous japanese 4wd road "rally car" from the early 2000's, the best time was set and on the second attempt the output shaft of the transmission sheared clean off the gearbox. not many owners would be too chuffed with that happening....... ;-)
Depends on the car I'd say.
In a 911 Turbo S or BMW M4? Probably not.
My mk4 Focus ST on the other hand might have the most pathetic launch controll in the history of launch controls. Once I release the clutch the rpm drop a bit, I get shit loads of wheelspin.
BMW has garbage launch control. And AWD drive cars could be driven by a literal monkey.
@@ThunderGoatz I heard it's kinda unreliable on some models sometimes, but when the new m4 launches properly, it launches pretty damn hard. Watched a lot of drag races.
Fwd vs rwd. I'm not sure but I thought that a proper launch in fwd cars have even more slip than in a rwd car due to the weight transfer.
My tuned M140i is faster if launched manually, the launch control is pretty pants.
@@DjNikGnashers heard BMW's are kinda hit or miss in this regard. Then again a tune probably also has a huge impact on how well the car is maintaining grip.
This has to be the best car channel on youtube. Interesting content and good humor. Absolutely love you guys.
Thanks Lukas!
@@OVERDRIVE.studios Thank you too, your videos are amazing to watch!
@@kurtlamprecht93 What's that supposed to mean? If you mean that I'm saying it for likes then you're wrong because this is THE BEST channel out there.
@@lukasmuursepp2267 is this channel your girlfriend or something
@@Twongz_does_anything Not yet 😏
Amazing driving skills! I did not expect you to really outperform it so well.
you can see it starts to spin at 2:12 then grips back up
Nice. Thought it was the edit at first.
A little wheelspin is faster. So is being proactive instead of reactive. This is great to see. Cheers
Love the humor and feel of the channel! Reminds me of early top gear!! Love it
This was a cool video, what would’ve made this even better would’ve been a pedalcam and a last run with launchcontrol to confirm the results
I feel like doing more runs and averaging between them would have been a more accurate display of the differences.
would be interesting to see what would happen if it was a torque converter automatic.
It would be nice if the launch control allowed you to tune some of the parameters to optimize the acceleration. It is probably set somewhat conservatively for safety reasons.
Carwow used the pista in a drag race last year and managed to get a 10.2 1/4 mile time. Though that was on a sunny day with a much higher track temperature
Good job man!
probably just getting better and better because the tires are getting warmer lol
After so many tries ofc he is faster :D tires got a lot warmer so he has more grip. Woould be interesting to see a launch control start right after he manual start.
Serious flex. Well done!
Congratulations!!! It would be interesting, nevertheless, to try the same test on different surfaces. This test has an ideal dry and grippy surface, so your professional skills beat the machine. How quick would you respond and manage such power on slippery surfaces? Please, try it again in another video!!!!!!! Also, it will show "normal people" that electronics provide safety, especially in powerful cars.
I LOVE that color spec! That blue and caramel interior is amazing!
Sounds like the engineers have more tuning to do. You're not doubt a great driver, but given the right input I think the computer will easily beat you every single time.
Excellent driving Scott! One thing computers can't do is make adjustments to launching. They only react to limit power output. You can make adjustments on how to launch the car.
Well done, mate
Presuming you didn’t fill up after each run, could the difference in speed be the fuel burning off? Added to the fact the engine and tyres were warmed up as other people have mentioned.
Gotta remember launch control is only as good as the sensors, DAQ system, Engineer(s) doing the programming, and how much time they were given. Lots of variables.
If you love cars and driving.... this is just the best channel. Keep doing your great work.
Nice job Mate :)
Bro salute to you for getting times like that on manual
@Driven Media What series do you race currently?
That’s impressive, great vid! But I wonder how much the fuel weight affected the time tho, like you must’ve try couple runs before setting the sub 11 sec time. Or thats actually negligible?
There's so many other factors I'd worry about before the weight from the fuel. I think it's rather negligible.
Nice exemple of throttle control 🙂
These uploads keep getting better. Scott, these videos are amazing, everyone unamed behind the scenes, you're doing God's work.
I would think the first 2 launch control runs heated up the tires and allowed for more grip on the driver controlled launch
"Give it the beans" love it.
I shouted donut media as soon as you said give it the beans lol great video!
4:50 I think what he's missing is that professional touch. a trained human does react slower than a computer but we still have a pre calculated coefficient of friction when we get use to a car. now its not as simple as us having a dedicated friction meter in humans, its just one of the beautiful wonders of our brain. but feel free to comment if you think im wrong or build upon my argument. (especially because I don't have any sources to quote)
in 'race' the traction control is fully on, it does only allow very limited sideways if in the right conditions, cutting very quickly; 'ct off' is a bit more flexible and it does allow you to actually spin the wheels longitudinally and 'drift' if matching the right conditions, leaving esp fully on at the same time. you should have used 'esc off' to give this video a sense, that's the only mode with everything off a part of the abs.
The reason a human can beat the ECU is anticipation. The human is anticipating what will happen, the computer is *reacting* to what is happening.
Nope. If a Human can beat the computer, then it's one dumb system. Look at German cars. No human can even get close to matching their launch control. Computer uses feedback loop with multiple observers. Observers are sensors to feed data to the computers. So that's like your eyes, the feel and our human reaction, just like 10.000x faster ;)
OP is incorrect. It’s a false assumption that is being made about a humans “superiority”. There is literally zero anticipation that would beat the computer and sensors in a scenario like this if they are designed properly, we are just tooo slow.
No
New aero discovery but by me, you know those side aero on the f1 car, they should have a slant down into the body, to under the cars, so when you turn that air goes into the defuser for better turning.
Friend of mine had one of these, very very fast and I've never been in a car as fast before but it broke down alot! Hence why he doesn't have it anymore.
Lets go humanssss! Good stuff Scott.
How much time was there in between the runs? Could the car do them immediately after each other?
THE BEST CAR CHANNEL.
I don't think its a coincidence that it was faster after every lap. You should have alternated launch control and manual starts. Also it would be interesting to see it tested in different brands of cars to see if any of them have better computers
Cool! There's hope for us yet!
Yesssss another one!!!
The launch control on my Leon Cupra is easy to beat. If you want to sit there and burn rubber leaving long black marks on the road, use launch control. Best times are with traction control on and standard auto mode.
Good job 👍. Worth editing to add a caption saying tyre temps were consistent, fuel loads balanced etc? Might ward off all the comments. Unless that’s the intention to drive interaction ;-)
Wheel speed sensors cut power via the dct transmission?
Thought this power cut was done via engine timing.
Probably should have done one last launch as the tires are now heated up
The cameraman brings out his wife's Tesla Plaid shopping car and lays down a 9.1 just before going to pick the kids up from school.. what a strange world automotive is at the moment..??
Would like to know what fuel was used was it low octane or e85
Did you keep tire temps and pressures roughly the same over all of the runs? Nice job beating a bunch of engineers :)
Really interesting. Some manufacturers are better at engineering launch control than others. The Porsche 911 turbo S must be the best currently - I bet you can’t beat that 😁. There must be a learning phase for each launch to assess current slip vs acceleration vs wheel torque so the algorithm has to be active with this dynamic mapping whilst controlling the throttle vs wheel torque curve. On an ICE and especially a turbo ICE that also is a dynamic arrangement. There’s a building of energy that the ecu must carefully control and the only (limited) energy dump I can think of is slipping the clutch. The ecu could do this but it’s very limited and a bit dodgy. Any chance of more info and vids on all this?
The launch control might also be set so there is a bit less wheelspin to make it more handalable that might not be a word but oh well
Scott, can we have a fuel economy run on the space saver tyres? See how far you can go on them and if they are a solution to rising fuel costs. I genuinely have wanted to know this for a decade or so. Please do it!
It also depends on the car in question aswell
For example an AWD 911 is most probably faster with launch control but something RWD with good throttle response like a Pista or a RWD Huracan could very well be like this
No control at the end? Wouldn't fuel load decreasing as you burn it off and tyre heat play a fairly big role?
Thats a hell of a car :)
how to put the rev up without launch control ?
Yeah im going to have to agree about the tire temps being the main reason it got faster.
Great skills 👍🏻
Try to do that with a car with better launch control aka bmw m 5 or m8 2022, porshe 911 turbo s , rimac nevera etc
Do any cars use a radar to work out wheel slip?
It's a feature in some tractors, you get a percentage slip indicator so you can optimise for traction and efficiency. Compare the actually ground speed to the wheel speed.
on RWD or FWD you could simply compare wheels speed. also, even measuring just one wheel, its not that hard to compute that a wheel is accelering quicker than it is mechanically possible for the car to accelerate,
@@geemy9675 it was the 4WD case I was thinking about but you're right, I hadn't considered that wheel slip in a car is whilst it's rapidly accelerating rather than on a field when it's something approximating steady state.
Thank you :)
@@tomcardale5596 traction control is different off road and on road and you don't want to remove wheel slip completely off road, but it's still the same principle, just with different parameters, like higher wheel slip threshold. detect wheel slip with a sensor (could be the abs sensor that already exists on each wheel) and ECU that compares every wheel speed. I guess it can also use other sensors for more redundancy and to be more precise/reliable like engine rpm/gear/steering wheel position, maybe even an accelerometer. but it's all pretty simple math to detect one or more wheels that start spinning faster than it should. then adjust power (retard ignition/inject less fuel), and/or brake whichever wheel is spinning too fast. diffs will automatically send more power to the wheels that don't spin. if every wheel is spinning, keep reducing power until they lock up again. on a 4wd with all diffs locked, every wheel spins the same speed, so you cant brake wheels individually, only reduce power. if it's electric, you have quicker and more precise control over motor output, thus more efficient traction control
@@geemy9675 no absolutely, I do get that :)
I've got rather more experience with off road traction aids than on road. I drive a vehicle with no ABS but that does have locking and torque biasing diffs and have a fair bit of seat time in various tractors with locking everything.
The radar wheel slip gauge is useful when everything is locked and you're trying to see if you need to reduce the load.
In some modern fast vehicles that have 4wd and limited slip everything I can easily imagine a similar scenario with all four wheels spinning. But then as you say there should be a noticeable engine load shift.
Nice and short video, thanks! Donut media would've thrown a 20 minute piece out
You should do an episode of abs vs threshold braking
i wonder if it would be any different with a NA engine. would the pro be even faster?
Maybe the tires were warmer at the end hence the better times ...? You should have done a lunch control test at the end as well when tires are fully warmed up!
Shows how good launch control is if it takes a professional racing driver to match/better the lc time. Basically, launch control is for the average joe. Scott should’ve also done a third run with an average driver without lc.
Or if you want to be pretty good at first try on new surface/track/weather conditions/new to car....
Try to do the same on wet or icy tarmac
The tires are warm after launches!
Imagine that! You warmed the tires as you went.
The tires are heated up with each consecutive run and will gain grip. Unless you took like hour long breaks between the runs. In that case this would be pretty accurate.
Carwows runway must be a lot better for them to be able to drive 10.4 in the pista
now you won our respect.
what can supercarblondies to do so?
stay tuned to the channels.
Motec has a system similar to this. Mod Bois we up!
Every pass kept rubbering in the used line and cleaning it off, kind of a moot test, but great idea!
now, do this with a Porsche. would be interesting 👍
Our reactions may be slower, but experience and skill can help us predict likely expected scenarios. As said by others, the order of consecutive launches also alters the variables at play.
now let’s see Paul Allen’s Launch control
cant see link for 8 wheeler?
Could u make a video pro vs abs.
Imo it was easy to beat due to MR drivetrain and fresh cup 2 tires even with all that high torque. On FF or FR cars with a simple pair of radial tires it should be more tricky challenge to accomplish
The same happens with abs. A pro gets better brake performance than a system breaking with abs. Not much about the reaction time of the system and more about how much it degrades the performance to reduce wheelspin. When you integrate the brake applied, you get more time near the limit when a pro is doing it instead of when the ECU is apllying and removing the brake rapidly, but many times, in ms. In general, the pro stays closer to the limit, and the system passes the limit and uses a PID controller to remove performance and reduce wheelspin in ms but many times, more in a try and error situation
er, you can't beat ABS, not for modern systems with yaw angle sensing, and certainly not with only 1 foot and 1 brake pedal.
Modern ABS controls all 4 wheel brakes simultaneously based on real time yaw and vehicle trajectory (as measrued by the handwheel angle sensor) As a vehicle dynamics engineer the last time us human testers could repeatibly beat ABS was back in around 2007. it's also worth noting that a race driver can "beat" road ABS systems because they have to be calibrated for the lowest skilled driver and use a tractive effort algorythm called "pick lowest" which means the first wheel to be modulated (braking force reduced) then sets that same lower level of brake for all four wheels, which extends the stoping distance. This is done to maintain driver yaw authority without requring corrective handwheel inputs, because most drivers simply are not skilled or experienced enough to do this.
"race" ABS however is calibrated for the shortest possible stopping distance and uses "pick high" where each wheel is maintained at the individual tractive effort limit for that wheel regardless of the yaw moment that creates and relies on the driver to steer to keep the vehicle going where they want it too. This is most evident on split mu surfaces, where say both left hand side tyres are on ice, and both right hand side tyres on dry tarmac. Here a road system, will drop the "grippy" tyres down in retard to match the force provided by the slippy side one, so the vehicle still goes in a straight line, and faithfully follows the drivers handwheel inputs, but does NOT actually stop as quickly as it could.
This is also important in road cars because frontal impacts are far more survivable than side impacts, so a car that spins and hits something is not good, whereas one that does spin but hits something going faster is gernally actually better - for the vehicle occupants anyway..... ;-) so the system is biased to preventing a loss of directional control.
@@maxtorque2277 thanks! I'm just an engine calibration engineer, with one year of experience only. Thanks for the class, ao much new info for me
I think it's only fair to do a launch control at the end when the tire are as warm as the last attempt. if you still manage to beat it porsche should be next xD
A duel "human VS computer" is actually a duel "human pilot VS human mathematicians" ^^
I think inst the TC , but the Hps losen when TC start to engage . Becouse if the wheel spins , you always are losing time,. So The point is , or TC is not allowing at max 80% to get maximum grip , or the Engine is not gettting the full power of the turbo while the TC is trying to reduce the amount of power to the wheels.
So I just need to spend half my life with a racing career I'll never have, before I can turn launch control off in the car I can't afford, and get slightly better times in a race I'll never be in.
Cool 😎
That pistas spec is 10/10
Could the weight of the fuel left in the tank effected it? Or did you guys fill up between each test?
the reason he can beat the ECU is the Human is reacting to what is happening as well as anticipating what is going to happen and thus puts a lot more gas into it as he is confident in controlling the car.
The ecu can only respond to what is happening right now, so even the tiniest of wheel slip its going to let of the gas...not always the fastest way to accelerate but better then a novice trying to launch it.
The reason of why humans beat the launch control system could be that when the traction control engages it is cutting too much power whereas the driver pushes through it which may be faster
It seemed like the traction control was creating juuuuuust enough “bog” to slow the car off the line.