So wheel speed sensors drop out and the car will accelerate uncontrollably? I love EVs, but with all this complexity I wish there's some kind of mechanical disconnect or a big red STOP button somewhere.
https://x.com/PicturesFoIder/status/1832940173400699255
(apologies -- not sure of the best Twitter passthrough to use)
If not, what's stopping a "traditional" (ICE) car from (mis)behaving in a similar fashion in some catastrophic circumstance that would damage its computer?
An ecu has a far more complicated control algorithm than a electric motor controller. If it were suddenly damaged it’s more like that the engine would fail to run at all then run out of control because the ecu needs to control the airflow, fuel and spark position for the engine to run, if any of those fail to work, or stop firing at the right the exact time they are required the engine will just stop or run very poorly. I actually think this is true of electric vehicles too, it’s far more likely to stop the motor working than to have it run out of control, unless a wheel speed sensor is damaged or something.
A petrol car can be placed into neutral if all else fails, the engine will run out of control but the car wont. Also the gearbox controller is typically a different computer from the ecu.
The brakes on any car should be able to over power the engine. This is not a challenge for 99% of petrol cars because the torque they output is tiny compared to what a brake system can apply to the wheels. If you slam on the brakes the engine doesn’t even come close. Idk about other countries but in Australia this also applies to electric cars that are road legal, it’s a requirement.
the ecu is usually located in the passenger cabin or sometimes next to the battery quite deep inside the engine bay.
The only thing that would cause a petrol engine to really go out of control would be if it was fly by wire throttle and that throttle position sensor was broken in the particular way that it’s reading as full throttle. Idk if manufacturers do this but it wouldn’t be hard to design a fly by wire throttle that when it fails the ecu will see it as closed not open.
Anyway I don’t think it’s much of a concern for electric cars either tbh.
More like 25 years ago, at least in France. The 2001 Renault Clio 2 I'm driving has throttle-by-wire, the newest car I personally know of with a mechanical throttle is a 1998 Peugeot 205, the last model year of a car that debuted in 1982. I doubt any European car manufactured after 2001 has a mechanical throttle, if only because of European emission standards.
> The only thing that would cause a petrol engine to really go out of control would be if it was fly by wire throttle and that throttle position sensor was broken in the particular way that it’s reading as full throttle. Idk if manufacturers do this but it wouldn’t be hard to design a fly by wire throttle that when it fails the ecu will see it as closed not open.
On the Clio 2 car, there are two redundant linear potentiometer tracks. If the dual measurements don't match or if either sensor is disconnected, the ECU will default back to a slightly higher than idle throttle.
That’s really neat. I was thinking something along the lines of having a microcontroller with a good adc right next to the throttle pedal and its angle transducer (hall effect would be better) and having that microcontroller send a digital signal. If the ecu can’t make sense of the digital signal or it is missing, just set the throttle to idle. That’s a solution that would be reasonably cheap today though but not 25 years ago.
Anyway in my experience most Japanese and Korean cars didn’t have fly by wire until the 2010s
Same for the steering. BMW for example has a method where the steering wheel is physically connected but the computer can add corrections to it via a clever set of gears. See here: https://en.wikipedia.org/wiki/Active_steering
If the computer (or electric steering motor) fails the steering wheel still works.
No, I don't think any of my bench tests suggest that for this car. The torque is minimal throughout, so if the motor was pushing an actual car then it might not move at all. If it did move, even a light touch on the (mechanical) brake pedal would stop it.
My problem is that I had a bench setup with no load on the motor and no mechanical brake. I could have pulled the safety interlock (all EVs have these for emergency first responders) but this stops the motor at all costs - wasn't sure of potential damage to the motor controller circuitry from back EMF.
Also the issue experienced in the post wasn't an issue with a sensor per se.
However, most relevant regulation (IEC61508, ISO26262, DO-178X) requires that systems controlling machines in automotive, rail or aerospace have a possibility of dangerous faults lower than 10^-9 (over the expected lifespan).
Many critical control systems like this are formally verified and/or extremely well-tested and have redundancy in both software and hardware.
https://en.wikipedia.org/wiki/2009–2011_Toyota_vehicle_recal...
'Michael Barr of the Barr Group testified[30] that NASA had not been able to complete its examination of Toyota's ETCS and that Toyota did not follow best practices for real time life-critical software, and that a single bit flip which can be caused by cosmic rays could cause unintended acceleration. "
If you've driven rattletrap manuals for most of your life, and have worked on cars, rebuilt engines, replaced clutches, rewired things that failed, yeah. That's an obvious conclusion, and I expect some people without doing that will have enough sense of what's going on to consider a drop to neutral (and letting the rev limiter handle keeping the engine intact).
But go ask most people, even in technical fields, about the details of a car, and you'll struggle to get much beyond "I press the gas and it goes." You run into this constantly if you're a "car guy" and people ask you questions about why their car isn't going. "It turns over but doesn't start!" can mean anything from "the lights are barely on and nothing happens" to "the starter relay clicks but nothing happens" to what I would consider that to mean, "the engine is rotating under the starter's power but is not engaging in sustained internal combustion."
Neutral isn't a thing most people even think about, unfortunately. Park, Drive, Reverse, and some oddball other positions that you don't want to end up in accidentally. Yes, they're useful, and yes, they solve problems, but it's not something that a lot of people would consider. Neither do they seem to consider "Stand on the brakes until the car comes to a stop. No, really, stand on them!" - because I've yet to meet a moderately well maintained vehicle that can't come to a stop with the gas floored and the brakes applied firmly (yes, I've tried, it's a standard test of mine after brake work). But if you only apply partial brake pressure, or have a vacuum brake booster, you only get a few attempts before the booster has lost vacuum (won't get any more, because wide open throttle), and if you've heated up your brakes trying without succeeding, you may very well have no usable brakes left. Passenger car brakes are adequate, but you can easily overheat them and fade them if you try, or boil the fluid, or... etc. Again, not something you'll find many people aware of these days.
I wish it were different, but "magic box I put gas into and it goes" is closer to the reality of how many people consider cars these days.
I locked myself out of my house recently, and it was only after scaling to the 1st floor, breaking in through an open window, and breaking through a locked interior door (the house had been secured as I was going on a trip, and the only things I forgot were my keys and that window), that I remembered that there was a spare key in my car (which was open). This moment of clarity coincided with the stress going away.
My daily driver has an archaic manual sequential transmission (2005 Ural - sidecar motorcycle sort of thing), and I select neutral at every stoplight I'm likely to be at for a while to avoid wearing the clutch bearings. Also, I have to most of the way double clutch my shifts on that bike (pause in the false neutral between gears) to avoid too much clashing. If I had a runaway throttle condition (certainly possible), I have at least three instant methods I'd use (kill switch, clutch, and rock it into a false neutral). But I've spent most of my driving career with such things, and vehicles that don't have those are a bit of a novelty to me.
1st will hold the car by itself if the parking brake slips.
Video from immediately after it flipped over https://m.youtube.com/watch?v=DChDTGIciNI
Think of how much effort it takes to pedal a bike from standstill in 1st gear vs the highest gear. The force required is the same if it’s being applied to the wheel and turning your legs (the engine)
And turn the wheels such that even if it got going it would be stopped by terrain? That should be especially easy on unpaved African roads.
Using the terrain is my preferred option, but then I didn’t have that choice
I agree, in theory it should be fine.
Driving a lap around the African continent is not the place to test that theory.
I just looked it up, my Jeep has the NSG 370 6 speed box behind the old 3.8ltr V6.
1st is 4.46
Reverse is 4.06
I agree. During a period of huge storms in my region I kept mentally preparing for getting caught in a flood: engage first gear, press the gas and go, slow and steady.
Then my fears materialized and as soon as the flood started pushing my car, I pressed the clutch and engaged the second.
Thankfully I realized it fast enough, kept pressing the gas and engaged back the 1st, so my mental training may have helped, but that was a close call.
If people think they are hitting the brakes (but are accidentally hitting the gas), then hitting the 'brakes' harder will make the problem worse.
You need training to guarantee correct reaction when things go wrong.
Here's an anecdote: I used to drive a car with a standard transmission in Los Angeles. In quite a few places, the parking spots have a "trench" in the front for drainage. So, you can place your car in reverse, release the brake and have your car roll forward quite a bit (the trench makes an even stronger downward slope on a hill that is already pointing downward) before the clutch engages. A bit surprising but nothing that weird for someone who drives a stick.
Now, have that sequence happen to someone driving a car with an automatic transmission. They shifted to reverse, the car is rolling forward more than they expect and is on a hill, they hit the gas to arrest the roll, the transmission engages and the car shoots in reverse. Pray that there isn't anything close behind them or they're going to run over a pedestrian, put their car through a wall, etc.
2) Because the majority of the people who had "uncontrolled acceleration" were old.
The vast majority of the cases were very likely driver error by older drivers who had incorrect habits ingrained. Toyota probably would have won the case if this was the only issue.
Alas, Toyota lost the case because their processes for safety were such a complete shitshow that they were going to get destroyed in court.
They have so much torque at a standstill that hill assist is a given and going up or down a hill slowly forwards or reverse is not an issue. Absolute gamechanger.
Not sure what you mean by standard transmission, but in a manual you'd also have to use the parking brake while using the clutch to go up in reverse. Otherwise the motor would stall, or you'd roll forwards.
Car starts accelerating out of control? You have several options that you can try in no particular order (except the first one, which should always be tried first).
Lift your primary foot to confirm you're not accidentally pressing the accelerator instead of the brake (this is surprisingly common). Dump the clutch with the other foot at the same time.
If the engine is still going nuts, shift to neutral and/or turn the ignition off (DON'T remove the key yet; that will lock the steering wheel, which is a bad idea when you're still moving). Coast to a stop somewhere safe; your brakes will also still work for a while. You won't have power steering, but you won't need it. Remove the key -- this WILL kill the engine (if you didn't already switch off). You're done.
My driving instructor did this to me in an empty supermarket car park. Then he did it again until I got the hang of it. It's a valuable learning experience.
At the time this was in the news, I was never able to find a coherent explanation of whether any such bit-flip affecting some piece of software on some module, would also inhibit the interpretation and implementation of a neutral drive mode selection.