Brake booster software update - recall

[Anthony121]

Does the ZB have a FBW braking system?
WTF is the brake booster 'software controlled'?
You'd think(being all things GM) it would be simpler, and cheaper to have a conventional brake system...

No But I think the VF V6 engine also had an electric vacuum pump to help with the vacuum for the brake booster. The software might be controlling this.

 

[Skylarking]

Maybe it's as simple as a connector problem, similar to the VF EPS problem?
GM may have decided they can overcome this with a software change, since this would cost GM less money than a component or partial loom changeout.
Kick the can down the road for a couple of years.

Well a simple EPS connector problem took GM/Holden years to come up with a proper solution? They tried grease, epoxy and cant remeber what else before changing some complete racks in customers cars. Then they decided a recall was in order and had customers wait while they build the repair kits to save the cost of a whole rack replacement... Shitfuckers...

Yep, wouldn't surprise me if they are kicking the can down the road...

 

[stooge]

i spoke to holden about it today and it will take a week or two for the recalls to roll out on their system and customers to be notified.
they said they have had no reported issues with it yet but all that "can" happen is you lose boost assist on the brakes, they still work but can become stiff because the software controls the boost pressure and in some instances will not provide enough pressure making the pedal stiff.

cars of the future are going to have a lot more of this sh!t lol, simple software glitches causing all sorts of problems so automotive manufacturers better get their act together and setup secure ota updating so the issues can be fixed quickly.

 

[Skylarking]

they said they have had no reported issues with it yet but all that "can" happen is you lose boost assist on the brakes, they still work but can become stiff because the software controls the boost pressure and in some instances will not provide enough pressure making the pedal stiff.

That’s the same shitfuckery they said around the EPS assistance failure and is simply a lack of imagination on their part. I can easily see bloke or lady driving their car and having a kid run out, where there normally non fault state the brakes would have been enough time to stop the vehicle before reaching that kid, but on finding a hard pedal the driver takes their foot off the brake and reapply which increases stoping distance to 1 meter past the kid that ran out. Either that of the driver just doesn’t press the pedal hard enough to stop… So yeah, no problem according to Holden

better get their act together and setup secure ota updating so the issues can be fixed quickly.

The solution is not to be able to do OTA update of a vehicle and add fresh problems. The solution to to produce fault free software in the first instance which then doesn’t require updating. But such is hard and costly so having easy OTA updates just lowers the bar in terms of design effort and software quality cause the fix is cheap and easy after…

OTA updates can work well with a consumer grade hardware like BD players or similar where failures don't risk lives. However motor vehicle failures can kill people and should require a different mindset to product and certify the software used within these safety critical items. As such, OTA updates aren’t the solution, in fact they can make software quality more elusive and also mask the real problem of poor quality design methods.

On the PC front, we now have persistent UEFI root kit circulating in the wild, something that we were told would be impossible with signed UEFI modules and certificate stores. We were told such shitfuckery couldn’t occur… So what was gained, easy updates that could be marketed as secure. But old school removable BIOS chips were much more secure in one sense as BIOS couldn’t be written to via the OS and either the chip needed to be removed and placed in a programmer or the programmer connected to MOBO header pins and a jumper set to program mode… We supposedly got smarter but screwed the pooch with UEFI but the supplies got a leg up on locking buyers out of their own devices. I don’t want to see the same with cars

Safety critical software within vehicle systems should also be designed using aerospace software development and verification techniques. And lots and lots of testing… Really, such systems should also be physically segregated from the consumer stuff within vehicles like infotainment and such, logical segregation isn’t enough… I don’t want OTA updates of such safety critical systems... I don’t want safety critical systems connected to consumer based systems... I want physical segregation, not logical segregation…

But what I want doesn’t matter so I'm sure as cars get smarter and smarter, the buyers will have to have greater and greater trust in the poorly designed systems that can kill us. Such is life now…

 

[stooge]

The solution to to produce fault free software in the first instance which then doesn’t require updating.

Will never happen.

In simple programming applications it may be possible but the more complex the functionality is the more avenues there are for errors to occur and it can be near impossible to test for every single scenario.

I have developed software for the last 25 years and on many projects i have spent many hours testing only to find a bug few weeks after release etc.
The hardest bugs to find are conditional bugs where it takes one condition to occur to cause another condition to fail.

Safety critical software within vehicle systems should also be designed using aerospace software development and verification techniques. And lots and lots of testing

Even in the aerospace industry there has been countless software faults that have caused all sorts of problems in actual space missions and passenger aircraft which has resulted in downed aircraft and a lot of loss of life.
In many cases with passenger aircraft it took a plane to crash and people to die for them to find the fault and correct it
the aerospace industry methodology is not as good as one would expect.



Vehicles are becoming much more complex and with electric vehicles being the future they are going to get much more complex so for future vehicles the solution is a secure method of ota updating to address the problems quickly.

The update systems can be designed so the updates are received, verified and stored locally and then performed while the vehicle is not in use.
When done properly ota updating is fine.
Recalls for software issues will not be necessary and quickly resolved.


I suspect the future mechanic is going to need some sort of electrical certification and some crazy tooling.

 

[bradp51]

Good reply stooge
Sounds like you actually have some experience behind your opinion

 

[wetwork65]

That’s the same shitfuckery they said around the EPS assistance failure and is simply a lack of imagination on their part. I can easily see bloke or lady driving their car and having a kid run out, where there normally non fault state the brakes would have been enough time to stop the vehicle before reaching that kid, but on finding a hard pedal the driver takes their foot off the brake and reapply which increases stoping distance to 1 meter past the kid that ran out. Either that of the driver just doesn’t press the pedal hard enough to stop… So yeah, no problem according to Holden

The solution is not to be able to do OTA update of a vehicle and add fresh problems. The solution to to produce fault free software in the first instance which then doesn’t require updating. But such is hard and costly so having easy OTA updates just lowers the bar in terms of design effort and software quality cause the fix is cheap and easy after…

OTA updates can work well with a consumer grade hardware like BD players or similar where failures don't risk lives. However motor vehicle failures can kill people and should require a different mindset to product and certify the software used within these safety critical items. As such, OTA updates aren’t the solution, in fact they can make software quality more elusive and also mask the real problem of poor quality design methods.

Absolutely agree in the cover up of incompletely developed systems being let loose in the wild for final QA. And these can kill us as well as third parties. Imagine saying "My brakes didn't work and that's why I didn't stop, officer"?
VF EPS anyone? My steering was too heavy, so I drove into him.
Tesla, amongst others provides OTA updates as part of their marketing shtick "to improve functions". Goodness me, Apple are quite open and comfortable with crippling older products with upates to make them so frustrating that you buy a new one.
It is problematic if the updates don't quite work, similar to Windows updates sometimes being not quite right, despite the programming ability of Microsoft. And it assumes that systems cannot be hacked into to change functions or even kill us - I don't believe there is any system that is truly unable to be hacked by incels (or worse) somewhere in Eastern Europe/Asia/North America etc.

Rave ends here.

 

[mirrabucca]

Absolutely agree in the cover up of incompletely developed systems being let loose in the wild for final QA. And these can kill us as well as third parties. Imagine saying "My brakes didn't work and that's why I didn't stop, officer"?
VF EPS anyone? My steering was too heavy, so I drove into him.
Tesla, amongst others provides OTA updates as part of their marketing shtick "to improve functions". Goodness me, Apple are quite open and comfortable with crippling older products with upates to make them so frustrating that you buy a new one.
It is problematic if the updates don't quite work, similar to Windows updates sometimes being not quite right, despite the programming ability of Microsoft. And it assumes that systems cannot be hacked into to change functions or even kill us - I don't believe there is any system that is truly unable to be hacked by incels (or worse) somewhere in Eastern Europe/Asia/North America etc.

Rave ends here.

And that, to me is the whole point. I am sick and tired of OTA "updates" that disable something I have been using for years, reliably, independant of the internet. An example: Years ago my wife bought me a Chrome key so I could play my collection of music videos on the TV. At that time I had XP on a laptop as my music machine. Great! Worked very well. I could control sound level etc all from the laptop and display my collection on the TV.
Then one day it stopped...... After much Googling, it turned out Google had ended support of XP on the Chrome key, so instead of "if it breaks we wont fix it" , they stopped it working altogether via an OTA update. So Iwas supposed to update my hardware so I could use something that I used quite happily OFFLINE for years? It would be like Sanyo doing an OTA to my microwave so it doesn't cook meat anymore...

Back on topic, my wife's Cruze has an electric vacuum pump to supplement the vacuum brake booster when vacuum is low due to (say) the engine being under boost from the turbo. And its a very old Cruze. Maybe GM had a problem when a random car was in a low-vacuum state and the driver needed brakes in a hurry but the software algorithm hadn't kicked the electric pump in early enough? Or not producing enough vacuum in that circumstance?

 

[VS 5.0]

When done properly

And therein lies the rub when humans are involved.

 

[stooge]

I don't believe there is any system that is truly unable to be hacked

that is the other side of internet connected vehicles too, they need to setup a secure method of updating.

when someone hacks a system they are actually leveraging a bug in the software and again it comes back to the fact that you cant test for every scenario on complex systems.
it is not that they are incomplete it is because they are complex and have many avenues of attack.