Tesla Turns Off AEB In New Cars Produced Since July

[stopcrazypp]

Really? All the advertised features on my iPhones worked right away. My Android (Dell Streak) was no different.

Here's an article just from yesterday. HMD is saying they will have Android P in all the Nokia phones they have now, but obviously they don't have it yet:
HMD promises Android P update for all 2017 Nokia smartphones | Trusted Reviews

Here's Motorola's announcement recently for Oreo:
Say Hello to Android Oreo

Samsung's announcement for Nougat in November last year when the S7 and S7 edge was still their current phones (S8 was not announced until March 2017):
Experience a Sneak Peek of Android 7.0 Nougat through the ‘Galaxy Beta Program’

They are selling phones based on promises of updates to a software version that it doesn't currently have. And you can bet that does make a difference in the purchasing decision.

 

[stopcrazypp]

Wiki is written by the public. How much of the content of that Wiki article was written by those who were paid by lawyers involved in the suit? Normally such articles are often written by legal aids to promote their firm and settlements.

Here's what the problem was, minus the hype:

On all cars, when 12v power is cut, the airbags will not go off. This is a safety feature.

On all key style cars, if you hang enough weight off the key and shake the vehicle violently, the switch will rotate. The fact the switch can be rotated to the OFF position while moving is a design feature on all cars, so you can turn off the engine in case of emergencies such as engine stuck at WOT, or fire.

GM's new switch design had a lower rotational torque than the previous switches. While the key itself could not rotate to OFF by violent agitation, the amount of weight added to key to get it to rotate was now lower. So the incidents of GM cars where the switch rotated to OFF while moving was higher than many other cars. All key style cars can do it with enough weight though.

Note: It is unknown if all 124 deaths would have avoided with higher rotation torque settings. If an airbag did not deploy due to the ignition being off and there was a death, it was assumed in that the death was caused by the lack of airbag deployment, but obviously cannot be proven.

I think he understands this point. The Wikipedia article detailed it, as does the links it gives. The problem was that the ignition could easily be toggled to off compared to other GM cars. Extremely unusually easy, to the point even the first initial test drives of the cars in 2004 had the issue happen. The company was also fully aware of the problem, and held meetings about fixing it in 2005, but ultimately scrapped the idea because it would cost too much money.
The Long Road To GM's Ignition Switch Recall
They quietly fixed the issue in ignition switches made in 2006, but didn't change the part number.
Document: GM Engineer OK'd No New Part Number in Ignition Switch Fix

 

[bhzmark]

I don't particularly want to be sharing the road with folks who are (knowingly or unknowingly) serving as beta testers for products that only kind of work.

now you are just making stuff up. you have no evidence that any Tesla products are unsafe on the road. To the contrary nhtsa has shown that Tesla cars w AP, even in the early stages of its rollout, are less likely to be in accidents. it's amusing to see the psychology of people who manufacture things to be upset about.

 

[Economite]

now you are just making stuff up. you have no evidence that any Tesla products are unsafe on the road. To the contrary nhtsa has shown that Tesla cars w AP, even in the early stages of its rollout, are less likely to be in accidents. it's amusing to see the psychology of people who manufacture things to be upset about.

NHTSA didn't (and couldn't) break out how much of that safety effect was due to AEB (which is definitely a safety feature) from what the effects were of AP (ie the TACC/Autosteer combo). Autosteer is flakey. For all we know, having/using autosteer could be net increasing collisions, but that effect is washed out by the huge collision mitigating effect of AEB. In other words, a car with just AEB may be statistically less likely to crash than a car with both AEB and TACC/autosteer.

 

[EinSV]

NHTSA didn't (and couldn't) break out how much of that safety effect was due to AEB (which is definitely a safety feature) from what the effects were of AP (ie the TACC/Autosteer combo). Autosteer is flakey. For all we know, having/using autosteer could be net increasing collisions, but that effect is washed out by the huge collision mitigating effect of AEB. In other words, a car with just AEB may be statistically less likely to crash than a car with both AEB and TACC/autosteer.

But if you had your way, no one should have had the AP1 hardware from fall 2014 until at least late 2015, when Autosteer/AEB functionality was released, if not later. I love my AP1 car and am very happy Tesla took a different path. I suspect the vast majority of AP2.5 owners are similarly happy to have the latest hardware, even if Tesla is still validating/tweaking the AEB functionality. The other option would have been to hold off installing the new hardware, which would have been against the interest of new owners.

 

[Economite]

But if you had your way, no one should have had the AP1 hardware from fall 2014 until at least late 2015, when Autosteer/AEB functionality was released, if not later. I love my AP1 car and am very happy Tesla took a different path. I suspect the vast majority of AP2.5 owners are similarly happy to have the latest hardware, even if Tesla is still validating/tweaking the AEB functionality. The other option would have been to hold off installing the new hardware, which would have been against the interest of new owners.

Or they could quietly put in the hardware, and not market/sell the AP software/functionality until it is actually ready/available. That's the thing, once software can be installed and turned on and off remotely, there's no practical need to charge for it when it doesn't yet exist based on some promises of what it will eventually do. But Tesla always tries to presell by stating that if the owner waits until the software actually exists to purchase it, the software will cost more. No one else does this. And it leaves Tesla constantly in the position of having to rush *something* over the air to satisfy customers who have already paid for and been promised much more than just *something*.

 

[EinSV]

Or they could quietly put in the hardware, and not market/sell the AP software/functionality until it is actually ready/available. That's the thing, once software can be installed and turned on and off remotely, there's no practical need to charge for it when it doesn't yet exist based on some promises of what it will eventually do. But Tesla always tries to presell by stating that if the owner waits until the software actually exists to purchase it, the software will cost more. No one else does this. And it leaves Tesla constantly in the position of having to rush *something* over the air to satisfy customers who have already paid for and been promised much more than just *something*.

That's a nice fairy tale -- where you wake up and Santa Claus has given you something you didn't expect and you didn't pay for. Not very realistic.

More to the point, none of the complainers/critics in this thread seem willing to live up to the logical conclusion of their criticism, which is that Tesla owners since July should not have been allowed to buy the AP2.5 hardware.

If there are new buyers who believe having AEB off the bat is a deal breaker they can buy an inventory car with AP 2.0 hardware. Problem solved.

 

[lolachampcar]

When I was younger, I saw things in pure black/white. If I had a beef, I simply would not buy.

Age has calmed me and now I'm willing to weigh the pros/cons of different options and pick the best. I just modify how I do business and how I set my own expectations depending on the realities of the chosen vendor. This is what I have done with Tesla and no longer have the previous degree of heart ache.

It really was that simple for me.

 

[bhzmark]

But Tesla always tries to presell by stating that if the owner waits until the software actually exists to purchase it, the software will cost more. No one else does this.

again false. plenty of people do it. they don't pre pay EAP so they can make the decision later. other people choose to prepay at time of delivery and they get a discount for prepaying. in addition to the benefit of the discount, they also get to use the functionality as it is being further developed and refined. But that is a different business model -- a model that is tough for some people to understand. they should probably just buy CPO or a Toyota.

 

[McRat]

I think he understands this point. The Wikipedia article detailed it, as does the links it gives. The problem was that the ignition could easily be toggled to off compared to other GM cars. Extremely unusually easy, to the point even the first initial test drives of the cars in 2004 had the issue happen. The company was also fully aware of the problem, and held meetings about fixing it in 2005, but ultimately scrapped the idea because it would cost too much money.
The Long Road To GM's Ignition Switch Recall
They quietly fixed the issue in ignition switches made in 2006, but didn't change the part number.
Document: GM Engineer OK'd No New Part Number in Ignition Switch Fix

Having driven, bought, rented dozens of the Killer Key cars, I can tell you the switch is not significantly different than other key cars.
I even did a video about the Killer Key Camaros when a lawyer argued you could easily turn off the ignition with your knee. If you'd driven a Kill Key Camaro, you'd understand. Cobalt, yeah, Camaro... BS. It can be done, but easy is not a good description. You can turn on your wipers with your chin. Not easy though.

This isn't fan-boy content. It's reality. Understand there were millions of those cars, not 800k like the article.

For more info call: 1 800 BAD DRUG

 

[SO16]

Wow... That's kind of rude.

Side note: I've never bought or flown a small plane. And to use general aviation as a comparison for a normal consumer product is a bit odd.

Looking back, I would agree. Sorry. Especially since I was thinking it was you who went on and on about this aviation nonsense earlier. But you were not that person. So please ignore. Thanks

 

[SO16]

Having driven, bought, rented dozens of the Killer Key cars, I can tell you the switch is not significantly different than other key cars.
I even did a video about the Killer Key Camaros when a lawyer argued you could easily turn off the ignition with your knee. If you'd driven a Kill Key Camaro, you'd understand. Cobalt, yeah, Camaro... BS. It can be done, but easy is not a good description. You can turn on your wipers with your chin. Not easy though.

This isn't fan-boy content. It's reality. Understand there were millions of those cars, not 800k like the article.

For more info call: 1 800 BAD DRUG

The MAIN issue is that GM hid the issue for a decade. When building new products and moving forward with new technology will of course not be 100% perfect. But how a company handles a known issue is the important piece and that is what a company had full control over. Tesla has done some wrong things in this area and so has GM.

 

[McRat]

The MAIN issue is that GM hid the issue for a decade. When building new products and moving forward with new technology will of course not be 100% perfect. But how a company handles a known issue is the important piece and that is what a company had full control over. Tesla has done some wrong things in this area and so has GM.

All automakers 'hide' dozens of internal reports as to the function of their cars. GM is not unique, it just has the deepest pockets and better documentation. It could be a radio control knob that falls off every 100,000 cars produced. It could be a high number of cracked windshields. Or burned out headlights. Or water leaks. Etc, Etc.

The real trouble was when some GM staffers lied about the reporting. But to this day, there are airbag non-deploys and deploys, which result in avoidable injuries or deaths in all brands. In GM's case it was found that it was easier to rotate the key to the off position than the original design specification called out.

Over twice as many people were killed by excessively violent front airbags which killed 290 occupants in otherwise survivable crashes. The NHTSA stalled the 2 speed airbags after they knew it was killing people. See any lawsuits against the NHTSA for those deaths? Note, the NHTSA was supplied with the data about the crashes that Killer Keys were involved with in fatal crashes. It did not seem to them that it was statistically significant.

It's like a commercial aircraft. There are literally thousands of MRB reports filed with every single tail number on a multiengine jet. Some are corrected, some require redesign, some are considered to acceptable as 'use as is'. There is a whole department and set of procedures for handling this on everything from spacecraft to children's toys.

It's a pretty safe bet that the AEB disconnect was part of a MRB report as corrective action. Unless they did not document it. Which legally, is the safer route.

 

[SO16]

All automakers 'hide' dozens of internal reports as to the function of their cars. GM is not unique, it just has the deepest pockets and better documentation. It could be a radio control knob that falls off every 100,000 cars produced. It could be a high number of cracked windshields. Or burned out headlights. Or water leaks. Etc, Etc.

The real trouble was when some GM staffers lied about the reporting. But to this day, there are airbag non-deploys and deploys, which result in avoidable injuries or deaths in all brands. In GM's case it was found that it was easier to rotate the key to the off position than the original design specification called out.

Over twice as many people were killed by excessively violent front airbags which killed 290 occupants in otherwise survivable crashes. The NHTSA stalled the 2 speed airbags after they knew it was killing people. See any lawsuits against the NHTSA for those deaths? Note, the NHTSA was supplied with the data about the crashes that Killer Keys were involved with in fatal crashes. It did not seem to them that it was statistically significant.

It's like a commercial aircraft. There are literally thousands of MRB reports filed with every single tail number on a multiengine jet. Some are corrected, some require redesign, some are considered to acceptable as 'use as is'. There is a whole department and set of procedures for handling this on everything from spacecraft to children's toys.

It's a pretty safe bet that the AEB disconnect was part of a MRB report as corrective action. Unless they did not document it. Which legally, is the safer route.

Makes sense. And I would agree with your assessment on the AEB disconnect. People were reporting phantom braking events and this could be why.

 

[d21mike]

Makes sense. And I would agree with your assessment on the AEB disconnect. People were reporting phantom braking events and this could be why.

I just got my AP2.5 car last Thursday (9/28) from the factory. I put on > 600 miles now with my road trip back. Not sure I am seeing the same phantom braking events. But during my trip I did notice a few brake events that I was not expecting while on autopilot. Trying to figure it out is was sometimes when going around a bend say to the left and a large truck would be ahead of me around the bend or something like that and it would brake kind of hard. But not so hard that I could not control it especially since I was pretty attentive being my first trip. But I understand the concern of cars behind me which was not the case.

Overall I love autopilot and would definitely not want to be without it. And happy to be on AP2.5. That was the main reason I got the car. And since I watched a lot of youtube videos of autopilot my expectations were pretty much on par with my experience.

 

[Economite]

again false. plenty of people do it. they don't pre pay EAP so they can make the decision later. other people choose to prepay at time of delivery and they get a discount for prepaying. in addition to the benefit of the discount, they also get to use the functionality as it is being further developed and refined. But that is a different business model -- a model that is tough for some people to understand. they should probably just buy CPO or a Toyota.

I didn't mean that no people choose to put off paying for/activating the EAP/FSD software until it is actually available. Many do. And Tesla is just fine installing the hardware in their cars at no cost on the speculation that they will eventually purchase the software upgrade that will (allegedly) make the hardware work.

I meant that no other car company installs hardware without software and gives customers the choice of paying up front for the software that doesn't yet exist, but tells customers that if they want to buy the software after they purchase the car (but once the software is actually available/works) they will have to pay a penalty vs. the pre-pay rate.

 

[bhzmark]

if they want to buy the software after they purchase the car (but once the software is actually available/works) they will have to pay a penalty vs. the pre-pay rate

to you it's a penalty for those pay later after delivery.

to others it's a discount for those that prepay.

half empty or half full.


and yes, thankfully Tesla does lots of things that other mfrs don't do.

 

[SO16]

I didn't mean that no people choose to put off paying for/activating the EAP/FSD software until it is actually available. Many do. And Tesla is just fine installing the hardware in their cars at no cost on the speculation that they will eventually purchase the software upgrade that will (allegedly) make the hardware work.

I meant that no other car company installs hardware without software and gives customers the choice of paying up front for the software that doesn't yet exist, but tells customers that if they want to buy the software after they purchase the car (but once the software is actually available/works) they will have to pay a penalty vs. the pre-pay rate.

Not much different than getting a discount for paying a year's worth of insurance at once. Fortunately people are given a choice and know those options up front. Not sure what your issue is here too.

BTW, I am one who held off on purchasing EAP and FSD.

 

[Economite]

Not much different than getting a discount for paying a year's worth of insurance at once. Fortunately people are given a choice and know those options up front. Not sure what your issue is here too.

BTW, I am one who held off on purchasing EAP and FSD.

Huge difference...

If I pay at the end of 2017 my insurance premiums for the entirety of 2018, the insurance company provides me with the promised insurance for all of 2018.

On the other hand, if I buy EAP and FSD at the end of 2017, there is absolutely no promise about what will be delivered in 2018 or when. And Tesla is incentivized to release something half-assed during 2018, before it is actually ready/tested, since it feels like it has to deliver something.

People who buy EAP/FSD only have extremely unreliable information (ie what Tesla states publicly) about what the software will do and when it will be delivered. Not so with prepaid insurance.

 

[bhzmark]

if that bothers you, looks like the best strategy for you is not to prepay for EAP or FSD. or buy CPO.

Huge difference...

If I pay at the end of 2017 my insurance premiums for the entirety of 2018, the insurance company provides me with the promised insurance for all of 2018.

On the other hand, if I buy EAP and FSD at the end of 2017, there is absolutely no promise about what will be delivered in 2018 or when. And Tesla is incentivized to release something half-assed during 2018, before it is actually ready/tested, since it feels like it has to deliver something.

People who buy EAP/FSD only have extremely unreliable information (ie what Tesla states publicly) about what the software will do and when it will be delivered. Not so with prepaid insurance.