Bolt self started/moved, and backed into garage wall/cabinet
- Thread starter SoCalif
- Start date
Help Support Chevy Bolt EV Forum:
A
Anonymous
Guest
What was your displayed range when parking and following when the incident happened? Maybe that could help determine if the car was actually off the whole time (no drop in range) or if somehow it did remain powered up.
Maxters
Well-known member
SoCalif said:EldRick said:
I dont believe my 4 lb indoor cat had access to the FOB, but I will ask.No other animals were in the garage that I saw/found/smelled!
After watching Secret Lives of Pets, I don't trust them. :roll:
SoCalif said:
If what you described is true I predict you will not be able to resolve this with GM until at least two other owners report the same thing here on this forum.
This is because 1) GM will not believe you, 2) GM will point finger at LG Chem, and LG Chem will point finger at GM, 3) then both companies will point fingers at you.
Probably true, but if there was a way for me to get a copy of the diagnostics, it would PROVE the keys were not in the car, and the car was never unlocked, therefore the car did this on its own somehow, or hacked, or some other anomaly which needs to be addressed.
Lets see how transparent GM is, hopefully a little more than the GVT......
Lets see how transparent GM is, hopefully a little more than the GVT......
d2170 said:
This all seems so implausible.
When you say you left the car and the doors locked automatically, did you verify by pulling a door handle to make sure they were locked? How are you certain they locked? Could the car have chirped to let you know that you shut the door and walked away while it was still running?
The way many of these systems work (FCA, BMW, etc) is that if you attempt to put the car in Park and fail to do so, it reverts to Neutral. This is what killed the young Star Trek actor Anton Yelchin.
Yelchin drove his Grand Cherokee to the top of his sloped driveway and with a shift mechanism similar to the Bolt, he thought he engaged Park. However the car was still either slightly moving while he tried to engage Park (imperceptible) or he failed to fully push the lever towards Park and the system put the car in Neutral. There is no visual or tactile cue that an FCA shifter like this has engaged a gear (much like the Bolt). He then walked down his driveway to get the mail and his Jeep was slowly and silently rolling backwards down the grade without him knowing. The Jeep rolled into him and crushed him against the mailbox. FCA has subsequently changed the Grand Cherokee shifter to a conventional PRNDL arrangement.
Can you try parking your car in the garage, putting it in Neutral (don't switch off the car), shutting the door, and walking away? My hypothesis here is that you didn't fully engage Park, forgot to shut off the car's power switch, shut the door, and then you walked away. The car chirped the horn to alert you of the problem and you assumed it was chirping to tell you the doors were locked. Like the Volt, after a period of time, the Bolt might shut off the ignition by itself if left in a running state unoccupied. The car would now be powered down and in Neutral, allowing for either slight movement or other outside influence, all the while you assumed the door was locked and the car was in Park.
Just a wild-ass guess.
When you say you left the car and the doors locked automatically, did you verify by pulling a door handle to make sure they were locked? How are you certain they locked? Could the car have chirped to let you know that you shut the door and walked away while it was still running?
The way many of these systems work (FCA, BMW, etc) is that if you attempt to put the car in Park and fail to do so, it reverts to Neutral. This is what killed the young Star Trek actor Anton Yelchin.
Yelchin drove his Grand Cherokee to the top of his sloped driveway and with a shift mechanism similar to the Bolt, he thought he engaged Park. However the car was still either slightly moving while he tried to engage Park (imperceptible) or he failed to fully push the lever towards Park and the system put the car in Neutral. There is no visual or tactile cue that an FCA shifter like this has engaged a gear (much like the Bolt). He then walked down his driveway to get the mail and his Jeep was slowly and silently rolling backwards down the grade without him knowing. The Jeep rolled into him and crushed him against the mailbox. FCA has subsequently changed the Grand Cherokee shifter to a conventional PRNDL arrangement.
Can you try parking your car in the garage, putting it in Neutral (don't switch off the car), shutting the door, and walking away? My hypothesis here is that you didn't fully engage Park, forgot to shut off the car's power switch, shut the door, and then you walked away. The car chirped the horn to alert you of the problem and you assumed it was chirping to tell you the doors were locked. Like the Volt, after a period of time, the Bolt might shut off the ignition by itself if left in a running state unoccupied. The car would now be powered down and in Neutral, allowing for either slight movement or other outside influence, all the while you assumed the door was locked and the car was in Park.
Just a wild-ass guess.
leodoggie said:
Just trying to eliminate variables here. I think the concept of the car turning itself on without keys, putting itself in reverse, and driving itself into a bench is borderline impossible.
If we can determine that the car could have been left running in Neutral and that the OP would have received the same auditory feedback that they assumed meant that car was locked, that means we have a car with nothing to prevent it from gently rolling or from being pushed without some sort of bizarre electrical gremlin starting the car and engaging a "gear" without keys. That's a pretty big distinction.
roundpeg
Well-known member
How much damage was done? This would tell us how fast the car was actually moving when it hit the work bench. From the one photo posted to another forum, it doesn't look like a lot.
No matter the scenario for how it happened, I suspect GM is going to have make some changes to how the car alerts you to having left it in a state where it may not be secure. As I have already discovered, it's really easy to walk away from the car while it is "running." Part of the problem is the use of obsolete language to describe how an EV operates.
No matter the scenario for how it happened, I suspect GM is going to have make some changes to how the car alerts you to having left it in a state where it may not be secure. As I have already discovered, it's really easy to walk away from the car while it is "running." Part of the problem is the use of obsolete language to describe how an EV operates.
roundpeg said:
It looks like the damaged bench wasn't particularly robust in the first place; a 3500 lb car bumping it at 0.5 MPH could have been sufficient to break it apart as shown in the photo. It appears the car sustained nearly zero damage and the wall behind the bench was undamaged, implying a very gentle impact. Had a car been actively accelerating into that bench I would expect wall and car damage from the impact.
This is tongue-in-cheek guys, but also a little serious.
I'm a software engineer by trade, with 30+ years of experience. Everything with software in it scares me. I'm pretty sure in 30 years I never wrote anything without bugs. The bigger or faster or more complex the thing with software, the more it scares me. Here's a short list of scary software in your life from small to large.
Phone - kills you or somebody else when you're looking at it and not the road
MRI - fries you alive if the software is bad (has happened, I believe)
Car - kills you if the software is bad, or the software distracts you (see cell phone).
Airplane - can (and routinely does) kill you if the crew misunderstands the software mode. I was also an aviator, and I remember several incidents where the crew did not understand what the aircraft software was doing, and flew the aircraft into the ground.
Power grid - sends us all to the dark ages when it's hacked.
True story. Way back in 1993, my Mazda RX7-R1 had a self-trained neural net processor engine controller (really). Shortly after it shipped, RX7s started shutting themselves down at stoplights. End of that experiment. I probably got some technical detail wrong about the RX7 processor, but the basic tale is true.
So if the Bolt has some software vulnerability not detected in testing, color me not surprised. Welcome to the new world. I think that will become the new normal.
I'm a software engineer by trade, with 30+ years of experience. Everything with software in it scares me. I'm pretty sure in 30 years I never wrote anything without bugs. The bigger or faster or more complex the thing with software, the more it scares me. Here's a short list of scary software in your life from small to large.
Phone - kills you or somebody else when you're looking at it and not the road
MRI - fries you alive if the software is bad (has happened, I believe)
Car - kills you if the software is bad, or the software distracts you (see cell phone).
Airplane - can (and routinely does) kill you if the crew misunderstands the software mode. I was also an aviator, and I remember several incidents where the crew did not understand what the aircraft software was doing, and flew the aircraft into the ground.
Power grid - sends us all to the dark ages when it's hacked.
True story. Way back in 1993, my Mazda RX7-R1 had a self-trained neural net processor engine controller (really). Shortly after it shipped, RX7s started shutting themselves down at stoplights. End of that experiment. I probably got some technical detail wrong about the RX7 processor, but the basic tale is true.
So if the Bolt has some software vulnerability not detected in testing, color me not surprised. Welcome to the new world. I think that will become the new normal.
roundpeg
Well-known member
Instinctively we probably all know this from our experiences with technology, but when you put it this way, it's kind if scary.
While some sort of comically-bad software could conceivably start a car without keys, put said car in reverse, and crash itself into a shelf, I think it's extremely unlikely.
I would imagine that the systems and software that operate the key authentication and bootup process are very distinct from those that determine gearshift behavior and also from those that control accelerator position, particularly given the intense scrutiny over such gearshifts (FCA death) and accelerator software (Toyota crashes).
I don't believe there's a single CPU and a single code repository that handles each of these three interfaces. We're probably talking about three systems needing to freak out on their own accord in proper sequence to create this incident. If I were putting money on what happened I'd put it on user error due to a clumsy interface/notification design.
The gearshift is unconventional and the auditory alerts may be easily misinterpreted. Furthermore this could be fraudulent activity for any number of reasons, especially considering the OP wasn't even at home to experience it.
I think the key authentication spontaneously failing (and not failing safe into a no-run state), the shift mechanism failing immediately afterward (and not failing safe into Park) and the accelerator then failing (and not failing safe into a no-acceleration state) is somewhere between preposterous and absurd. This is much more likely to be user error, design misinterpretation, or nefarious activity.
I would imagine that the systems and software that operate the key authentication and bootup process are very distinct from those that determine gearshift behavior and also from those that control accelerator position, particularly given the intense scrutiny over such gearshifts (FCA death) and accelerator software (Toyota crashes).
I don't believe there's a single CPU and a single code repository that handles each of these three interfaces. We're probably talking about three systems needing to freak out on their own accord in proper sequence to create this incident. If I were putting money on what happened I'd put it on user error due to a clumsy interface/notification design.
The gearshift is unconventional and the auditory alerts may be easily misinterpreted. Furthermore this could be fraudulent activity for any number of reasons, especially considering the OP wasn't even at home to experience it.
I think the key authentication spontaneously failing (and not failing safe into a no-run state), the shift mechanism failing immediately afterward (and not failing safe into Park) and the accelerator then failing (and not failing safe into a no-acceleration state) is somewhere between preposterous and absurd. This is much more likely to be user error, design misinterpretation, or nefarious activity.
roundpeg
Well-known member
Blaming bad results on "user error" is becoming the generic excuse technologists use for foisting poorly designed systems on unsuspecting users. I see little practical difference between bugs and design concept errors.
roundpeg said:
I'm not dismissing that this could be a design fault and I'm of the general opinion that automakers should stop tinkering with such an important interface so well-understood by all drivers (PRNDL lever). However I find it extremely difficult to believe that this car started itself from a powered-off state and put itself into reverse with no keys in it.
Something is amiss here beyond just software gremlins.
With regard to "user error", I originally assumed that the Toyota runaway problem was simply another example of "user error" until a close friend, very competent and calm, experienced it on the freeway with his entire family. He managed to bring the car to a halt with the engine still racing, he told me, shut it down, and never again purchased another Toyota product.
As far as I was concerned, this happening to him was as if it had happened to me...not user error.
As far as I was concerned, this happening to him was as if it had happened to me...not user error.
Similar threads
