Apple Car

[ooid]

Well it looks like, Apple just re-considering their position on the car they might produce/sell

http://www.nytimes.com/2016/09/10/technology/apple...

[anonymous-user]

ooid said:

Well it looks like, Apple just re-considering their position on the car they might produce/sell

http://www.nytimes.com/2016/09/10/technology/apple...

Making ANY (decent) car is hard. Making an autonomous one before existing car companies is really hard. If anyone is going to make autonomous cars work, the technology will come from the established companies in ADAS, like Bosch or Valeo.

[2gins]

Have either of you actually read the story? As is so often the case, the headline is attention grabbing 'Apple re-thinks car plans' or whatever but the actual article could not be more lacking in detail about what Apple are, or are not doing or not doing. A notoriously secret project remains notoriously secret. fking wow.

Eta not an Ifanboy here, really couldn't care less. Just read the link before jumping up and down guys.

[ooid]

2gins said:

Have either of you actually read the story? As is so often the case, the headline is attention grabbing 'Apple re-thinks car plans' or whatever but the actual article could not be more lacking in detail about what Apple are, or are not doing or not doing. A notoriously secret project remains notoriously secret. fking wow.

Eta not an Ifanboy here, really couldn't care less. Just read the link before jumping up and down guys.

Yes I did. I did not jump into conclusion, basically a news addition to conversation. But here we go another one, if you wanna jump and down so much

http://www.macrumors.com/2016/09/09/apple-car-reth...

[k-ink]

How many of you will trust your lives and those of your family to robotic cars? I will wait until a few thousand guinnea pigs have died ironing out the bugs. So maybe when the tech is 100% proven and there are no robot induced deaths for ten years, I will consider it.

As for electric cars I will wait until the aerated battery cores are out giving 2000 times more storage and recharge speed.

[RobDickinson]

People are not getting safer at driving.

Self driving cars will.

Tesla have a very high minimum standard of safety before their self driving functions are not beta. Far far safer than people

[k-ink]

anonymous said:

[redacted]

So I assume you'd be happy to be one of those killed by the Tesla software bugs then? If so crack on fella. They wont be on my radar until others have died fixing the bugs for me. Only then, and if I am really bored of driving, will I be interested.

[ChasW]

Like many giant industries before it the auto sector has been slow to change and pick up on major innovations. Generally it has been forced into major change by regulators, ie safety and emissions. Disruption has come from the likes of Tesla with a stylish and attractive electric cars when the mainstream built hybrid cars shaped like cyclists helmets bought mostly by people needing to be PC or wishing to make an eco statement. I am sure Apple and Google look at cars simply as appliances which can be serve their bigger ambitions which I guess is to "own" you as a customer when you are on the move. The car offers great opportunities for a communications hub that can be monetised with new services that the disruptors are well placed to provide. The mainstream manufacturers have picked up on this. My mate has a new C Class. Get in it and your eyes are immediately drawn to the screen and its controls. To me it's a computer on wheels and has no appeal whatsoever.

Where PHers feel challenged is in the new thinking, myself included BTW. For me cars are about fun, excitement, the joy (and pain) of ownership and improving them. They were aspirational. For many in the future they will be appliances to move you around that users rent by the hour, day or year. I can't image in 10 years time that a young schoolboy will have a full size poster of an Apple car on his bedroom wall where his Grandad once had a Countach or Porsche Turbo!

[RobDickinson]

k-ink said:

So I assume you'd be happy to be one of those killed by the Tesla software bugs then? If so crack on fella. They wont be on my radar until others have died fixing the bugs for me. Only then, and if I am really bored of driving, will I be interested.

What software bug is that then?

[silent ninja]

k-ink said:

How many of you will trust your lives and those of your family to robotic cars? I will wait until a few thousand guinnea pigs have died ironing out the bugs. So maybe when the tech is 100% proven and there are no robot induced deaths for ten years, I will consider it.

As for electric cars I will wait until the aerated battery cores are out giving 2000 times more storage and recharge speed.

I guess you've never flew in a plane? I high level of autonomy is already existing. The next step is natural and will reach a head in the next two decades

[k-ink]

We all have to deal with other people. That risk is a constant.

What you are adding with robotic cars is another layer of added risk on top - that of the software not recognising certain situations which have led to deaths.

For those who don't care and just want to be the first with the new toy, they will be the ones to test whether it is safe with their own children. Good luck with that.

I respect what they are attempting to do, as I have said many times on PH. But I don't adopt new unproven tech prematurely. I am not the sort who camps outside of Apple stores.

[ChasW]

k-ink said:

We all have to deal with other people. That risk is a constant.

What you are adding with robotic cars is another layer of added risk on top - that of the software not recognising certain situations which have led to deaths.

For those who don't care and just want to be the first with the new toy, they will be the ones to test whether it is safe with their own children. Good luck with that.

I respect what they are attempting to do, as I have said many times on PH. But I don't adopt new unproven tech prematurely. I am not the sort who camps outside of Apple stores.

Rather be at risk of robot driver than this couple coming in my direction


http://www.bbc.co.uk/news/world-us-canada-37320535

[k-ink]

At the moment there is no choice of: robot car OR the usual maniacs coming towards you.

The current choice is: unproven robotic car AND the usual maniacs coming towards you.



Only after many years will the software be proven to work 100%. There will be no escaping maniacs until every scrote owns a robotic car.

[frisbee]

k-ink said:

At the moment there is no choice of: robot car OR the usual maniacs coming towards you.

The current choice is: unproven robotic car AND the usual maniacs coming towards you.



Only after many years will the software be proven to work 100%. There will be no escaping maniacs until every scrote owns a robotic car.

I don't think the bar needs to be set that high, most human drivers are completely oblivious at the wheel - work 50% maybe?

[k-ink]

By the time enough people are in robot cars to make a difference the software should be fully tested anyway. All I am saying is we are a long way off right now.

[a7x88]

The difference between humans and computers is that computers don't make mistakes. If you want a repetitive task done to a constant standard as quickly and effective fly as possible l, you get a machine to do it.

Note the tesla autopilot accident. The autopilot is currently in beta and still requires full attention/intervention - not only did the tesla not recognise a semi truck across the road but neither did the guy.

A self driving car won't get tired/sleepy/distracted/drink drive etc. It will be improved upon each and every time an accident does happen - human drivers have a particularly poor historical performance in this area.

Then add that statistically a US drivers is killed once every 100 million miles. Tesla has completed 130million miles on autopilot with one death. It's record has already exceeded humans

[k-ink]

Computers do make mistakes because they do what humans tell them to do, via software. A robot car is only as good as the phuman programmer. If the programmer is human the program will have errors.

Driving is not a simple process like adding lids on jars on a conveyor belt. There are millions of variables and millions of decisions to be made. It only takes one special set of circumstances to unfold and it will go wrong. Only time and testing will iron out those bugs.

Ps

I can't even type of sentence on my phone without it thinking it knows best and changing six words. st programming equals st results.

Edited by k-ink on Sunday 11th September 11:08

[a7x88]

You said it yourself - the bugs will be ironed out. Humans however have not improved significantly over the years. The records already show the few bugs in the programme designed by humans (yes there are and will be mistakes in something as complex as this - you are right) still make for a safer method of transport per mile.

Now bear in mind this is beta - tesla have stated it will stop being beta when it's safety record is 10x that of humans. So one fatal every 1000million miles.

Humans have no hope of achieving that. An constantly improving "script" does. It shouldn't make the same mistake twice.

You cannot irradiate the risk if you want to travel by car. The aim is not to irradiate it. But to significantly reduce it. 10x less likely to die - yes please

[frisbee]

k-ink said:

Computers do make mistakes because they do what humans tell them to do, via software. A robot car is only as good as the phuman programmer. If the programmer is human the program will have errors.

Driving is not a simple process like adding lids on jars on a conveyor belt. There are millions of variables and millions of decisions to be made. It only takes one special set of circumstances to unfold and it will go wrong. Only time and testing will iron out those bugs.

Ps

I can't even type of sentence on my phone without it thinking it knows best and changing six words. st programming equals st results.

Edited by k-ink on Sunday 11th September 11:08

There isn't just a single programmer. There will be independence, other engineers will review, scrutinise and verify the software. Its like having a whole car full of backseat drivers that can instantly agree on the correct thing to do.

If your car has ABS you've already got a layer of software between you and an accident anyway.

I'll trust a team of engineers a lot more than a trust most other drivers on the road!

[ooid]

I do not see any problem with Automation or self-driving. I think it will mostly replace trade (vans, bus, or any other commercial transporting vehicles) super quickly. I'm sorry if this has been someone's employment but we would see major benefits in especially city traffic. Many accidents or congestion happens due to large vehicles in cities nowadays, it's not an easy job to do so it's definitely need automation and precise control. Not to mention half-empty, over-sized routemasters around, they are complete joke. Big infrastructure needed in cities to support auto-pilot or self driving systems to replace these commercial vehicles, then we get to enjoy our cars more often In relation to integrating some of these automation to our personal cars, I do not see any issue again, most modern cars already have auto-cruise, embedding more smarter systems would make boring part of the journeys more efficient for us.

The issue is "design" of the car. Whether this is Tesla, Apple, Bmw, Porsche or VW...The car has to be designed by considering the "driver" and its attachment to the activity "driving" which essentially ties loads of components together, from engine design to, main body, styling or interior.