-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Run-away I-PACE, really?
- Thread starter Pink Duck
- Start date
Lord Farquad
2022 MYLR W/B/20s/Tow bar. Previously 2021 M3SR+
Um. Used to have a 2010 Toyota Prius; SO was driving a 2008 Toyota Sienna. And there were All These Reports about runaway Toyotas.
Toyota claimed, early on, that it was Floor Mats. Sliding up and jamming the accelerator pedal. They went so far as to, at their cost, provide a retrofit driver's side floor mats for all their vehicles that clipped onto hooks so they wouldn't slide. But.. there were Still Reports of runaway Toyotas.
Finally, a really good bunch of lawyers teamed up with an excellent expert on automobile controllers. Against every attempt of Toyota's to prevent People From Looking At Source Code and Schematics, the expert and one of his team got into a clean room and were allowed to read the source code. No notes. But this expert and his friends had reverse-engineered the schematic of the auto controller, went looking for errors, and found.. a huge number.
The problem turned out to be serious spaghetti code and very little attention to good engineering practice and standards. Toyota had a controller that had been put together about the time they were ditching hardware carburetors. It was barely adequate at the time for what they were doing. As more and more complex vehicles were being built, the software "engineers" and their management kept on adding code but not doing any further analysis to see if there was adequate real-time to do the processing that they were asking this underpowered controller to do. Further, there were major problems with the code, related to the stack and heap, where, given enough interrupts, the variable space (holding things like the throttle position value) would get overwritten. To any of you with controller experience, this is beyond a no-no: This is absolutely verbotten.
Worse, there's this idea of a Watchdog Timer. If the software doesn't come by and decrement a value from time to time, it underflows (or overflows, but you get the idea) and causes a reset. The Huge Pile of Code that was the problem also included the watchdog timer. And an external watchdog timer was mis-designed and Wouldn't Work.
If one was lucky, the bollixed-up watchdog timer would go off; then the CPU would perform a reset in about a second or so, and the driver and testers at Toyota wouldn't notice a thing. If one was unlucky, the throttle position (cars were drive-by-wire) would freeze and no messing around with the pedal would change things.
Fault codes? The same pile of bollixed up software was also responsible for generating fault codes.
With this information, the expert and his buddies put a Modern Toyota onto a dynanometer, ran it at speed, and went looking for errors. They were reliably able to get the car to take off and not stop.
The brakes? Easy. Slam on the brakes, hard, and leave the pedal down, panic stop all the way. Toyota misdesign or no Toyota misdesign, the car would stop.
But normal humans don't do that. Step on the brakes, the car slows a bit, come off the brakes, car speeds up, hard on the brakes, lather, rinse.. and now the brakes are too hot to stop, no matter how hard one presses on the brake pedal. And if one come off the brakes long enough for them to cool, well, now one is at some unreasonable speed and Crashes R Us.
The victims in this case were a couple of ladies driving, I believe, a Camry, coming off an interstate onto an off-ramp ending at a stop light. The car did its thing, the lady had maybe ten seconds to realize that the End Was Nigh, and didn't. The car went through the light and crashed; I believe the passenger died.
I read the deposition, in detail, when all this was going through trial at the time. And, yeah, I am a EE with microcontroller hardware and firmware design experience. I've never done automotive, but all the terms in the deposition were very familiar. Thinking back on it even now, I'm still furious, but you couldn't get near me for days after reading that deposition. The jury (and, yeah, it was a jury trial) had all the terms explained, in detail, to them by a person who was both a prof and a better expert on this than any of the Toyota so-called experts.. and none of the actual code writers showed up from Japan. No surprise there.
The Toyota Prius, interestingly, was a white-paper design on the controller and wasn't subject to this problem. The Sienna was. And there lies a tale: We got a monetary settlement. But no replacement of the subject engine controller. And, to my knowledge, no firmware update for a late 1990's controller design. As far as I know, and I could be very wrong on this, there's 2000's-era Toyotas still running around with that engine controller.
So: I'll actually agree with the posters who say that problems with runaway cars are 99%+ about people who mistake the accelerator for the brake. It happens, literally, all the time.
But, if there's one place where the phrase, "The exception proves the rule.", this was it. It is possible to misdesign code in a modern automobile so it does a runaway.
As far as older cars: Back before we got married, I was the passenger in the SO's Dodge Colt (motto: Orange Pumpkin) when, when accelerating into a turn from a stop at a red light, the car Took Off. Being young, with good reflexes, I reached over and turned off the key, and she successfully drove the car to a parking position on the street. Popped the hood and discovered that, somehow, the accelerator cable had gotten hooked over a protuberance in the engine compartment. Unhooked it. I think (it's been a long time) some hose had gotten out of place by falling out of its clip. So, yeah, older cars were subject to this, too. The car in question was a manual, so she could've just hit the clutch.. or the brake.. but I happened to react first.
Toyota claimed, early on, that it was Floor Mats. Sliding up and jamming the accelerator pedal. They went so far as to, at their cost, provide a retrofit driver's side floor mats for all their vehicles that clipped onto hooks so they wouldn't slide. But.. there were Still Reports of runaway Toyotas.
Finally, a really good bunch of lawyers teamed up with an excellent expert on automobile controllers. Against every attempt of Toyota's to prevent People From Looking At Source Code and Schematics, the expert and one of his team got into a clean room and were allowed to read the source code. No notes. But this expert and his friends had reverse-engineered the schematic of the auto controller, went looking for errors, and found.. a huge number.
The problem turned out to be serious spaghetti code and very little attention to good engineering practice and standards. Toyota had a controller that had been put together about the time they were ditching hardware carburetors. It was barely adequate at the time for what they were doing. As more and more complex vehicles were being built, the software "engineers" and their management kept on adding code but not doing any further analysis to see if there was adequate real-time to do the processing that they were asking this underpowered controller to do. Further, there were major problems with the code, related to the stack and heap, where, given enough interrupts, the variable space (holding things like the throttle position value) would get overwritten. To any of you with controller experience, this is beyond a no-no: This is absolutely verbotten.
Worse, there's this idea of a Watchdog Timer. If the software doesn't come by and decrement a value from time to time, it underflows (or overflows, but you get the idea) and causes a reset. The Huge Pile of Code that was the problem also included the watchdog timer. And an external watchdog timer was mis-designed and Wouldn't Work.
If one was lucky, the bollixed-up watchdog timer would go off; then the CPU would perform a reset in about a second or so, and the driver and testers at Toyota wouldn't notice a thing. If one was unlucky, the throttle position (cars were drive-by-wire) would freeze and no messing around with the pedal would change things.
Fault codes? The same pile of bollixed up software was also responsible for generating fault codes.
With this information, the expert and his buddies put a Modern Toyota onto a dynanometer, ran it at speed, and went looking for errors. They were reliably able to get the car to take off and not stop.
The brakes? Easy. Slam on the brakes, hard, and leave the pedal down, panic stop all the way. Toyota misdesign or no Toyota misdesign, the car would stop.
But normal humans don't do that. Step on the brakes, the car slows a bit, come off the brakes, car speeds up, hard on the brakes, lather, rinse.. and now the brakes are too hot to stop, no matter how hard one presses on the brake pedal. And if one come off the brakes long enough for them to cool, well, now one is at some unreasonable speed and Crashes R Us.
The victims in this case were a couple of ladies driving, I believe, a Camry, coming off an interstate onto an off-ramp ending at a stop light. The car did its thing, the lady had maybe ten seconds to realize that the End Was Nigh, and didn't. The car went through the light and crashed; I believe the passenger died.
I read the deposition, in detail, when all this was going through trial at the time. And, yeah, I am a EE with microcontroller hardware and firmware design experience. I've never done automotive, but all the terms in the deposition were very familiar. Thinking back on it even now, I'm still furious, but you couldn't get near me for days after reading that deposition. The jury (and, yeah, it was a jury trial) had all the terms explained, in detail, to them by a person who was both a prof and a better expert on this than any of the Toyota so-called experts.. and none of the actual code writers showed up from Japan. No surprise there.
The Toyota Prius, interestingly, was a white-paper design on the controller and wasn't subject to this problem. The Sienna was. And there lies a tale: We got a monetary settlement. But no replacement of the subject engine controller. And, to my knowledge, no firmware update for a late 1990's controller design. As far as I know, and I could be very wrong on this, there's 2000's-era Toyotas still running around with that engine controller.
So: I'll actually agree with the posters who say that problems with runaway cars are 99%+ about people who mistake the accelerator for the brake. It happens, literally, all the time.
But, if there's one place where the phrase, "The exception proves the rule.", this was it. It is possible to misdesign code in a modern automobile so it does a runaway.
As far as older cars: Back before we got married, I was the passenger in the SO's Dodge Colt (motto: Orange Pumpkin) when, when accelerating into a turn from a stop at a red light, the car Took Off. Being young, with good reflexes, I reached over and turned off the key, and she successfully drove the car to a parking position on the street. Popped the hood and discovered that, somehow, the accelerator cable had gotten hooked over a protuberance in the engine compartment. Unhooked it. I think (it's been a long time) some hose had gotten out of place by falling out of its clip. So, yeah, older cars were subject to this, too. The car in question was a manual, so she could've just hit the clutch.. or the brake.. but I happened to react first.
Last edited:
yeah.. but in this case - just reckless driver
Looks like the driver has now been arrested... What a plumb! https://www.itv.com/news/granada/20...jaguar-electric-car-brakes-failed-on-motorway
Depends if the police had to consult Jaguar Land Rover specialist, and how much that individual was paid 
Well if he was not guilty then you already would have jis lawyer or family members trumpeting all over the place.Depends if the police had to consult Jaguar Land Rover specialist, and how much that individual was paid
Imho as I wrote, bellend was speeding, maybe already had 6 or 9 points, got flashed and decided to invent that story.
The amount of data black boxes record with every car these days is such, that I am sure it sees and knows everything
Driver might have been bargaining on his local Jaguar dealer being so useless that they'd just report nothing on the diagnostic logs again.
Might not have counted on the police impounding it for independent forensic analysis.
I still can’t believe the YouTube interviewer didn’t ask during the call whether he tried putting the car into neutral the second time, and if not, why not.
From 2nd vid: JLR spokesperson said: “We're looking into this matter with urgency. A full review is underway to determine the cause of the incident which is still yet to be established. The safety of our clients and vehicles is JLR's highest priority.”
Might not have counted on the police impounding it for independent forensic analysis.
I still can’t believe the YouTube interviewer didn’t ask during the call whether he tried putting the car into neutral the second time, and if not, why not.
From 2nd vid: JLR spokesperson said: “We're looking into this matter with urgency. A full review is underway to determine the cause of the incident which is still yet to be established. The safety of our clients and vehicles is JLR's highest priority.”
Last edited:
“I think if he's if he’s done me over and 880,000 people have watched it then I just think this story gets better”
Furious back-pedalling by the interviewer, who was sympathetic to whatever the iPace driver said all through the interview (as said, he's obviously anti-EV). These luddites just can't stop believing their own nonsense. Look at the comments in that last YT video. Some of them think the arrest is a Government conspiracy FFS!
Plenty of red flags in the interview:
My take - he didn't want the car or can't afford it after changing jobs but couldn't give it back (leased) or resale value has plummeted (purchased).
No doubt his mates have been ribbing him about driving a milkfloat and he wants to get back in a gas guzzler to appease them.
Possibly he's just an attention seeker. Mission accomplished.
I almost wonder if the Police watched it and then realised they'd been had...
Plenty of red flags in the interview:
- Speeds/times/distances/route on both 'runaways' didn't add up
- No explanation of how brake pedal felt when he tried to use it
- No explanation of what accelerator pedal was doing, if anything
- Didn't even mention the EPB (handbrake) which like a Tesla can be applied in an emergency
- Wouldn't go into neutral then suddenly did
- Range estimates supposedly wrong but no specific numbers quoted
- Police couldn't open door from outside (perfectly normal in any car these days)
- Mention of being flashed by cameras and what result of that might be
- Problems with car and dealer not responsive (only 1 EV technician shared between several dealerships - Really?)
My take - he didn't want the car or can't afford it after changing jobs but couldn't give it back (leased) or resale value has plummeted (purchased).
No doubt his mates have been ribbing him about driving a milkfloat and he wants to get back in a gas guzzler to appease them.
Possibly he's just an attention seeker. Mission accomplished.
I almost wonder if the Police watched it and then realised they'd been had...
I think they already knew it.
I am sure they checked his driving record, saw multiple speeding fines, saw motorway cameras (which probably showed that he was breaking before speed cameras before his voyage was supplemented with call to 999) and connected the dots.
And then got the forensics to check the car and it was clear as day
I think he just realised he'd been pinged by a van on a bridge, or had actually seen a Gatso flash, at over a tonne and was already on 6-9 points, and needed his licence for his job.
As said very early, as did others, it never really added up. Losing brakes is one thing - hard to believe in a modern car to be honest - but an EV with regenerative braking via the motors? If you're not pressing the accelerator you're slowing down. Even if none of that was working if you could get it into neutral the wind resistance would slow you down from motorway speeds pretty quickly I think.
His contention was that he lost the hydraulic brakes and the accellerator stuck at 90 (not top speed, for some reason), and couldn't switch to neutral, all simultaneously.
People actually believed that, and not just people looking for a reason to hate EVs... we truly are in the silliest timeline.
People actually believed that, and not just people looking for a reason to hate EVs... we truly are in the silliest timeline.
Similar threads
