7 ms·
> So the controller only ever increased the torque request, or kept it at the same level. Even when I simulated pressing the brake it was like "Nothing needs to
by cwalv 2y ago
> So the controller only ever increased the torque request, or kept it at the same level. Even when I simulated pressing the brake it was like "Nothing needs to change, we're not even moving!"
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.
- trhway 2y ago>the car will accelerate uncontrollably? https://en.wikipedia.org/wiki/2009–2011_Toyota_vehicle_recalls 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. "
- cwalv 2y agoYeah, maybe not particular to EVs. I do remember wondering back during the Toyota "uncontrolled acceleration" epidemic why people wouldn't just put the cars in neutral
- Syonyk 2y agoBecause the vast majority of people driving cars don't have enough of a mental model of the vehicle systems to consider such a thing, especially in a panic situation such as 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.
- Arch-TK 2y agoEven people who know about the solution might find it drops out of their brain in a moment of panic. 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.
- Syonyk 2y agoAgain, it depends a lot on your experience with vehicles. I expect someone who had driven a manual for a long while (or even learned on one but hadn't driven one recently) would be radically more likely to come to "Oh, select neutral" as a solution than someone who has only ever driven automatics. "Neutral" is far more part of "life with a manual" than it is with automatics - I would be willing to bet that a substantial fraction of automatic transmissions have never been deliberately put in neutral. 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.
- mkesper 2y agoAlso my Toyota Auris Hybrid has a weird kind of joystick for changing gear and putting it into neutral position requires holding that position for some time. Gave me an unpleasant (but luckily harmless) event in a car wash where you're required to have power on but use neutral position with automatic gearing.
- 01HNNWZ0MV43FF 2y ago
- idunnoman1222 2y agoPeople that did put their car in neutral didn’t make the news
- bsder 2y ago1) Because panic is a thing. 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.
- jval43 2y agoEVs (and hybrids) solve that problem so elegantly. 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.
- aaronmdjones 2y agoThis is why I'm so glad I have a manual transmission with a physical ignition key. 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.
- myself248 2y agoI'm still not sure whether it applies to this case, but on the Prius at least, "neutral" is a software concept. There is no physical linkage from the drive mode selector ("gearshift") to anything mechanical, and there are no mechanical components that could uncouple the powerplant from the wheels. Putting the powertrain in neutral is accomplished by removing torque commands from the electric motors, at which time the engine and wheels can free-spin. 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.
- thescriptkiddie 2y agoI feel like if you can't come up with the idea to try pushing in the clutch, shifting into neutral, turning off the key, or applying the handbrake within ten seconds you should have your license taken away. and if you design a car where those things wouldn't work you should go to prison.
- rocqua 2y agoThis was on automatic shifting cars.
- thescriptkiddie 2y agoI don't have a lot of experience with automatic shifting cars, but I believe that they still have neutral, keys, and handbrakes. If they don't then their designers should go to prison.
- Doxin 2y agoIf the wheel speed sensors drop out they put nothing on the CAN bus, not a "0 speed" message. I think it's pretty safe to assume the controller logic here has a fairly strict timeout on how often it wants to see wheel speed messages.
- Retr0id 2y agoThere is presumably still some possibility of them failing in an "always reports 0" way
- lostlogin 2y ago…Like when dismembered and spread across a bench.
- Arch-TK 2y agoIt's not unthinkable that whatever transducer takes rotation and turns it into a signal which is processed by an MCU and translated into CAN messages could get stuck producing one signal and trigger such a scenario without the car's control system needing to be on a bench.
- 0_____0 2y agoThere are some parts of engineering in safety systems where you have a single thing that could go wrong that would have serious consequences, and the result of the FMEA is that "it has to not do that".
- zardo 2y agoI don't think you'll find a wheel speed sensor without a few "outputs incorrect speed" failure modes.
- 0_____0 2y agoDetectability is one dimension of an error, and "bad wheel speed" has decent detectability I reckon - either through redundancy, grey codes, index pulse checking, bound checking. Also the issue experienced in the post wasn't an issue with a sensor per se.
- qingcharles 2y agoThis is a video of a driverless car getting rear-ended, and doing who-knows-what damage to its electronics; it then goes rogue at max speed through the streets smooshing whatever is in its path: https://x.com/PicturesFoIder/status/1832940173400699255 https://x.com/PicturesFoIder/status/1832940173400699255 (apologies -- not sure of the best Twitter passthrough to use)
- croisillon 2y agoEspecially since each link to Xitter forwards a couple times to itself and messes the browser history… One popular option is to use the Nitter instance xcancel.com
- blashyrk 2y agoI am completely ignorant of all things automotive. I was under the impression that any relatively recent (15 or so years) ICE car also operates by way of a car computer, and that stepping on the gas pedal is just a way to politely instruct the computer that you would like it to apply throttle. And that for even more recent cars this also applies to braking (since the newer cars can brake automatically). Have I got it all wrong? 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?
- left-struck 2y agoYou’re sort of right in your assumption but there’s a lot of context missing. First, the time period is more like 30 years for cars having engine control units (ecu), but most car up until 10 years ago or so had hard physically wired throttles where you stepping on the throttle pulled a physical cable. If that cable doesn’t get pulled the engine doesn’t get enough air to go really fast no matter what the ecu tries to do. More recent cars have fly by wire throttles meaning they are like electric cars in that sense. 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.
- angusgr 2y ago> So wheel speed sensors drop out and the car will accelerate uncontrollably? 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.
- foobarian 2y agoHuh. It just occurred to me that our EV has no mechanical mitigation at all. No physical handbrake, no clutch, no mechanical key to power off the engine. I'm not sure how to feel about this.
- speedgoose 2y agoYou have brakes.