Another unexplained acceleration

[andyf1140]

framerateuk said:

TheInternet said:

Safety critical custard. I'd also wager that there are fewer DBW incidents than cable throttle incidents.

I'd agree. I've had the throttle cable snap on my Caterham. In a light car like that it's almost as bad as slamming on the brake.

I had the throttle stick full open on my roadster (Weber 32/36 carb) and that was a frightening experience!

The issue with the DBW arrangement is that its the ECU that decides how much throttle to apply not the driver.

[RobM77]

I couldn't possibly know how this incident happened, but accidents with automatics and someone pressing the wrong pedal, and/or getting confused about the creeping forwards in D are fairly common. As I say though, without seeing all the evidence it's ridiculous to try and judge.

[GroundEffect]

saaby93 said:

GroundEffect said:

You should see the safety validation and FMEM handling with hardware redundancy it goes through. It boggles my mind, and I've been involved in it for years.

I could ask the same here
How would the engine know the difference between a faulty input from a light throttle press and a normal request for max power?
Surely as its within normal operating parameters it would just give max power?

Hardware redundancy and check sum plausibility. As I said, the smartest people in the business have spent 20 years writing software and standards for this stuff. Don't underestimate it.

[RobM77]

GroundEffect said:

saaby93 said:

GroundEffect said:

You should see the safety validation and FMEM handling with hardware redundancy it goes through. It boggles my mind, and I've been involved in it for years.

I could ask the same here
How would the engine know the difference between a faulty input from a light throttle press and a normal request for max power?
Surely as its within normal operating parameters it would just give max power?

Hardware redundancy and check sum plausibility. As I said, the smartest people in the business have spent 20 years writing software and standards for this stuff. Don't underestimate it.

's ability to annoy the hell out of anyone used to driving a cable throttle car.

Just added a bit for you there

[J4CKO]

I did postulate a while back about those with the propensity for unintended acceleration versus say a Tesla Model S P100 D, that really would be immovable objects and pretty unstoppable forces, assuming the clever stuff isnt switch on, those things weigh over two tonnes, do sixty in well under three seconds, have 4WD and 600 lb/ft of torque.

[mk2driver]

Brakes won't always stop a car with engine at full torque, that depends on the vehicle speed. Speed is important due to the fact that all brakes do is convert kinetic energy into heat.

If you are going at a high vehicle speed with engine demanding full torque continuously you might end up in a situation where the brakes fail through overheating before the vehicle stops.

This depends on vehicle speed, mass and the brake system as well as the engines torque capability of course.

In this case it certainly would drop the car given the relatively low vehicle speed and my opinion would be she simply hit the wrong pedal

[saaby93]

durbster said:

Have you ever heard of a case of "unintended acceleration" where the driver has pressed the brakes?

erm no
Ive heard of when youre being towed you have your foot over the brake then suddenly the tow car stops so you do the typically move your foot across and floor the clutch

also there's the brake hard merchants who floor the accelerator by mistake and light up the front wheels

This case is different for two reasons because
her first action is to lightly press the accelerator and the car surges forward - it takes off
- that's not the usual wrong pedal action described by other posters
It sounds like a fault in the fly by wire system somewhere

When it's done that she tries braking and find they don't seem to be working properly
- she doesnt say they're not working as some posters have tried to say
she's explaining the effect on the brakes that the initial fault is causing

And no - I wouldnt expect it to be easily recreatable - it's just one of things that might happen once in a blue moon with electronic gubbins

Edited by saaby93 on Friday 3rd February 15:37

[chryslerben]

andyf1140 said:

The issue with the DBW arrangement is that its the ECU that decides how much throttle to apply not the driver.

With modern driving standards being so low I'd say that was a benefit.

[RobM77]

chryslerben said:

andyf1140 said:

The issue with the DBW arrangement is that its the ECU that decides how much throttle to apply not the driver.

With modern driving standards being so low I'd say that was a benefit.

Yes, it's like clutch delay valves: great news for the incompetent, but a nightmare for the competent.

[Silent1]

mk2driver said:

Brakes won't always stop a car with engine at full torque, that depends on the vehicle speed. Speed is important due to the fact that all brakes do is convert kinetic energy into heat.

If you are going at a high vehicle speed with engine demanding full torque continuously you might end up in a situation where the brakes fail through overheating before the vehicle stops.

This depends on vehicle speed, mass and the brake system as well as the engines torque capability of course.

In this case it certainly would drop the car given the relatively low vehicle speed and my opinion would be she simply hit the wrong pedal

The brakes on any car in current use will absolutely stop that car even with the engine at full throttle, the stopping distance usually increases by 10 metres over a normal emergency stop. If there was a car in the road that couldn't stop then it's brakes would be so woefully inadequate in normal usage as to be dangerous.

[saaby93]

Silent1 said:

The brakes on any car in current use will absolutely stop that car even with the engine at full throttle, the stopping distance usually increases by 10 metres over a normal emergency stop. If there was a car in the road that couldn't stop then it's brakes would be so woefully inadequate in normal usage as to be dangerous.

Unless I'm misreading it can we stop going on about the brakes
I dont think shes claimed brake failure - only that they wouldnt working properly
The issue is with the other pedal - the car surging off

[GroundEffect]

RobM77 said:

GroundEffect said:

saaby93 said:

GroundEffect said:

You should see the safety validation and FMEM handling with hardware redundancy it goes through. It boggles my mind, and I've been involved in it for years.

I could ask the same here
How would the engine know the difference between a faulty input from a light throttle press and a normal request for max power?
Surely as its within normal operating parameters it would just give max power?

Hardware redundancy and check sum plausibility. As I said, the smartest people in the business have spent 20 years writing software and standards for this stuff. Don't underestimate it.

's ability to annoy the hell out of anyone used to driving a cable throttle car.

Just added a bit for you there

I defy any human to actually be able to tell the difference in 'time to torque' of fly by wire vs cable in a blind test. Compared to inertia of the engine and the combustion delays, it's negligible.

[RobM77]

GroundEffect said:

RobM77 said:

GroundEffect said:

saaby93 said:

GroundEffect said:

You should see the safety validation and FMEM handling with hardware redundancy it goes through. It boggles my mind, and I've been involved in it for years.

I could ask the same here
How would the engine know the difference between a faulty input from a light throttle press and a normal request for max power?
Surely as its within normal operating parameters it would just give max power?

Hardware redundancy and check sum plausibility. As I said, the smartest people in the business have spent 20 years writing software and standards for this stuff. Don't underestimate it.

's ability to annoy the hell out of anyone used to driving a cable throttle car.

Just added a bit for you there

I defy any human to actually be able to tell the difference in 'time to torque' of fly by wire vs cable in a blind test. Compared to inertia of the engine and the combustion delays, it's negligible.

That's a very specific thing that you've chosen because it's hard to spot. There are plenty of other DBW characteristics that are obvious in many current production cars, such as the much written about lag at the top of the pedal - BMW especially can't get this right and most of their petrol engines are dreadful because of it (it's why I drive a diesel). With my E46 330ci you could count to three before the throttle pedal did anything - in fact I could tap my foot to the music in the car and the car would just serenely continue without any modification to torque supplied at all. Then there's the three effects I listed in my post earlier on - those are pretty scary. You've also got the weird relationship between throttle pedal position and torque, rather than what you get with a cable, which is throttle pedal position and throttle opening on the engine (as I'm sure you know, at low revs a small throttle opening can deliver more torque than a large one, DBW manages this for you, meaning the pedal is a torque request lever, not a throttle opening lever).

Don't get me wrong, DBW is a great technology. In my 2-Eleven it was fantastic and a noticeable improvement over my Elise 111S. I've not driven a single seater with DBW, but I should image they're good too (otherwise they'd be undriveable). In most fully homologated production cars though, especially petrol powered ones, it's utter st.

[SonicShadow]

saaby93 said:

the car surging off....

.... because she stamped on the accelerator pedal instead of the brake.

[GroundEffect]

RobM77 said:

GroundEffect said:

RobM77 said:

GroundEffect said:

saaby93 said:

GroundEffect said:

You should see the safety validation and FMEM handling with hardware redundancy it goes through. It boggles my mind, and I've been involved in it for years.

I could ask the same here
How would the engine know the difference between a faulty input from a light throttle press and a normal request for max power?
Surely as its within normal operating parameters it would just give max power?

Hardware redundancy and check sum plausibility. As I said, the smartest people in the business have spent 20 years writing software and standards for this stuff. Don't underestimate it.

's ability to annoy the hell out of anyone used to driving a cable throttle car.

Just added a bit for you there

I defy any human to actually be able to tell the difference in 'time to torque' of fly by wire vs cable in a blind test. Compared to inertia of the engine and the combustion delays, it's negligible.

That's a very specific thing that you've chosen because it's hard to spot. There are plenty of other DBW characteristics that are obvious in many current production cars, such as the much written about lag at the top of the pedal - BMW especially can't get this right and most of their petrol engines are dreadful because of it (it's why I drive a diesel). With my E46 330ci you could count to three before the throttle pedal did anything - in fact I could tap my foot to the music in the car and the car would just serenely continue without any modification to torque supplied at all. Then there's the three effects I listed in my post earlier on - those are pretty scary. You've also got the weird relationship between throttle pedal position and torque, rather than what you get with a cable, which is throttle pedal position and throttle opening on the engine (as I'm sure you know, at low revs a small throttle opening can deliver more torque than a large one, DBW manages this for you, meaning the pedal is a torque request lever, not a throttle opening lever).

Don't get me wrong, DBW is a great technology. In my 2-Eleven it was fantastic and a noticeable improvement over my Elise 111S. I've not driven a single seater with DBW, but I should image they're good too (otherwise they'd be undriveable). In most fully homologated production cars though, especially petrol powered ones, it's utter st.

Then it's kind of down to driver preference, right? Because I design the pesky things, I do think of the throttle as a 'torque demand' not a 'butterfly valve position control'. So I adjust my own inputs to suit.

I haven't worked on Driver demand control (the wky term we use in the industry) directly but I presume there is a reasonable amount of damping in the system at the first few %age of travel to keep things smoother. My M3 in normal pedal map has that very thing. It makes causing backfires more bothersome!

[RobM77]

GroundEffect said:

RobM77 said:

GroundEffect said:

RobM77 said:

GroundEffect said:

saaby93 said:

GroundEffect said:

You should see the safety validation and FMEM handling with hardware redundancy it goes through. It boggles my mind, and I've been involved in it for years.

I could ask the same here
How would the engine know the difference between a faulty input from a light throttle press and a normal request for max power?
Surely as its within normal operating parameters it would just give max power?

Hardware redundancy and check sum plausibility. As I said, the smartest people in the business have spent 20 years writing software and standards for this stuff. Don't underestimate it.

's ability to annoy the hell out of anyone used to driving a cable throttle car.

Just added a bit for you there

I defy any human to actually be able to tell the difference in 'time to torque' of fly by wire vs cable in a blind test. Compared to inertia of the engine and the combustion delays, it's negligible.

That's a very specific thing that you've chosen because it's hard to spot. There are plenty of other DBW characteristics that are obvious in many current production cars, such as the much written about lag at the top of the pedal - BMW especially can't get this right and most of their petrol engines are dreadful because of it (it's why I drive a diesel). With my E46 330ci you could count to three before the throttle pedal did anything - in fact I could tap my foot to the music in the car and the car would just serenely continue without any modification to torque supplied at all. Then there's the three effects I listed in my post earlier on - those are pretty scary. You've also got the weird relationship between throttle pedal position and torque, rather than what you get with a cable, which is throttle pedal position and throttle opening on the engine (as I'm sure you know, at low revs a small throttle opening can deliver more torque than a large one, DBW manages this for you, meaning the pedal is a torque request lever, not a throttle opening lever).

Don't get me wrong, DBW is a great technology. In my 2-Eleven it was fantastic and a noticeable improvement over my Elise 111S. I've not driven a single seater with DBW, but I should image they're good too (otherwise they'd be undriveable). In most fully homologated production cars though, especially petrol powered ones, it's utter st.

Then it's kind of down to driver preference, right? Because I design the pesky things, I do think of the throttle as a 'torque demand' not a 'butterfly valve position control'. So I adjust my own inputs to suit.

I haven't worked on Driver demand control (the wky term we use in the industry) directly but I presume there is a reasonable amount of damping in the system at the first few %age of travel to keep things smoother. My M3 in normal pedal map has that very thing. It makes causing backfires more bothersome!

The torque vs throttle opening issue is indeed merely a driver preference. For me it felt weird at first, but seems natural now. I suspect most people prefer the pedal controlling torque.

The other aspects that I listed (delays etc) are just bloody annoying - I've lost count of the number of otherwise superb cars ruined by effects like that. For example, the problem with top of pedal lag is mainly in entering a corner; if a car takes half a second to respond when you transition from brake to throttle then that may not seem a lot on paper, but in a 50mph corner that's 11.1 metres, which is the entire corner with no throttle control at all.

Please don't get me wrong about your technology - in theory I think it's utterly wonderful. My SVA Lotus 2-Eleven was wonderful and way better than a cable such as I have in my Formula Renault (I own a 2008 FR, the last year made with a cable throttle). In practise in a road car though it's awful and a deal breaker for me on lots of cars.

[AJB]

GroundEffect said:

I defy any human to actually be able to tell the difference in 'time to torque' of fly by wire vs cable in a blind test. Compared to inertia of the engine and the combustion delays, it's negligible.

Depends on the car. In my current car and anything implemented at all well it's just fine and I don't notice it. I once had a Skoda Fabia courtesy car though, and kept almost stalling every time I pulled away. Eventually I made it to work, and experimented in the car park to figure out what my problem was. I realised it was just that the throttle response was so laggy that it hadn't even started to open by the time I'd brought the clutch to the biting point. I could blip the throttle and be off the pedal again before the revs even started to rise, let alone die down again.

I suspect it was early days of the technology and they'd just specified a woefully inadequate motor on the butterfly. Or maybe, but less likely, the control loop software was rubbish.

[GroundEffect]

RobM77 said:

GroundEffect said:

RobM77 said:

GroundEffect said:

RobM77 said:

GroundEffect said:

saaby93 said:

GroundEffect said:

You should see the safety validation and FMEM handling with hardware redundancy it goes through. It boggles my mind, and I've been involved in it for years.

I could ask the same here
How would the engine know the difference between a faulty input from a light throttle press and a normal request for max power?
Surely as its within normal operating parameters it would just give max power?

Hardware redundancy and check sum plausibility. As I said, the smartest people in the business have spent 20 years writing software and standards for this stuff. Don't underestimate it.

's ability to annoy the hell out of anyone used to driving a cable throttle car.

Just added a bit for you there

I defy any human to actually be able to tell the difference in 'time to torque' of fly by wire vs cable in a blind test. Compared to inertia of the engine and the combustion delays, it's negligible.

That's a very specific thing that you've chosen because it's hard to spot. There are plenty of other DBW characteristics that are obvious in many current production cars, such as the much written about lag at the top of the pedal - BMW especially can't get this right and most of their petrol engines are dreadful because of it (it's why I drive a diesel). With my E46 330ci you could count to three before the throttle pedal did anything - in fact I could tap my foot to the music in the car and the car would just serenely continue without any modification to torque supplied at all. Then there's the three effects I listed in my post earlier on - those are pretty scary. You've also got the weird relationship between throttle pedal position and torque, rather than what you get with a cable, which is throttle pedal position and throttle opening on the engine (as I'm sure you know, at low revs a small throttle opening can deliver more torque than a large one, DBW manages this for you, meaning the pedal is a torque request lever, not a throttle opening lever).

Don't get me wrong, DBW is a great technology. In my 2-Eleven it was fantastic and a noticeable improvement over my Elise 111S. I've not driven a single seater with DBW, but I should image they're good too (otherwise they'd be undriveable). In most fully homologated production cars though, especially petrol powered ones, it's utter st.

Then it's kind of down to driver preference, right? Because I design the pesky things, I do think of the throttle as a 'torque demand' not a 'butterfly valve position control'. So I adjust my own inputs to suit.

I haven't worked on Driver demand control (the wky term we use in the industry) directly but I presume there is a reasonable amount of damping in the system at the first few %age of travel to keep things smoother. My M3 in normal pedal map has that very thing. It makes causing backfires more bothersome!

The other aspects that I listed (delays etc) are just bloody annoying - I've lost count of the number of otherwise superb cars ruined by effects like that. For example, the problem with top of pedal lag is mainly in entering a corner; if a car takes half a second to respond when you transition from brake to throttle then that may not seem a lot on paper, but in a 50mph corner that's 11.1 metres, which is the entire corner with no throttle control at all.

I'd be very surprised if the delay was actually 0.5 seconds. Time to torque on the engine itself will be around 0.3 seconds, of that maybe 0.05 will be electronic/actuator delay. I would be very interested in seeing data you have to show otherwise.

[yellowjack]

Anyone got a link to that track day/driving experience video?

The one where the instructor is yelling "Brake! Brake! ...BRAAAAKE!!!" and the chap driving it is yelling back "I am, I am, I am" as they leave the track and drive at high speed over the grass at the side of the circuit.

The bloke driving is absolutely convinced that he's braking, even though he's clearly hard on the throttle and the poor bloke beside him is begging him to use the brakes...

[AJB]

saaby93 said:

Unless I'm misreading it can we stop going on about the brakes
I dont think shes claimed brake failure - only that they wouldnt working properly
The issue is with the other pedal - the car surging off

But if you're heading towards something or someone fast, then as long as your foot isn't on the wrong pedal every subconscious instinct will make you press the brake pedal really hard.

I don't think it's at all likely that the car surged, and she reacted by getting on the brakes but only pressed the pedal gently and crashed at speed.

To my mind either she was on the wrong pedal so didn't press the brakes hard, or the brakes didn't overcome the engine's torque despite her pressing hard. And that'd mean mechanical/hydraulic brake failure as well as throttle failure.

If she'd said "the car surged forwards so quickly that I didn't even have time to get to the brake pedal before crashing" then I'd be more inclined to accept that it could have been the car's fault (although human error is still MUCH more likely). But the fact that she thinks she's moved her foot to the brake pedal and pressed it, but couldn't slow/stop the car with the brakes, almost 100% rules out car problem in my mind.