weltweit
Well-Known Member
What about airbags, are they also mandatory?
autonomous cars - the future of motoring is driverless
- Thread starter roryer
- Start date
weltweit
Well-Known Member
I may be kicking against the tide here but I think autonomous cars are a bit longer term, even Google's cars which have seen plenty of on road trials. There probably will be some cars on the road in the next 10-30 years that have levels of autonomy but I don't see them taking over any time soon. Quite apart from the technical complexity and liability issues, I don't see present day drivers demanding them.
Sure cruise control is useful on the motorway, and automatic braking might prevent some collisions, self parking might be useful for some people, but who out there is clamouring for a car that can drive itself .. the last study I read said that people thought it could be good for others, but they themselves preferred to be in charge.
mauvais
on reddit or something
I think you're looking at it through the tunnel vision of the contemporary.
I have no doubt that people will want autonomous cars, not least because it will be the end of personal car ownership. Almost noone likes owning and being personally responsible for a car, even people who like cars.
With very limited exception, noone likes monotonous driving, noone likes not being able to drink or sleep or check their phone, noone likes parking, noone likes paying in a variety of ways for metal to sit there doing nothing, noone likes inanimate cars clogging up everywhere, noone likes chauffeuring people or objects around or paying humans to chauffeur them. Autonomous cars dispense with all that.
mauvais
on reddit or something
In the US (where seatbelts are not necessarily expected to be worn), not in the EU, although almost all cars have them. At a guess, because airbags are not without risk - a harness and a rollcage is a better idea, although rare for obvious reasons - and possibly because legislators are keener to implement stuff that protects third parties.
Indeed. So then Volvo have a huge risk exposure. And one with up to a 70 year lifespan. how do we collectively deal with that? Volvo certainly aren't in a position to hold on to it unconditionally.
No, I'm talking post investment. As an industry, it doesnt even get close to paying for itself. The reason it keeps going is that some players do make money... and the rest have hope.
mauvais
on reddit or something
I'd not heard that about post-investment before, interesting.
As for Volvo, the '70 year' problem is again why we will move away from personal ownership to a service model. That puts the service provider back in charge of asset life. We're not even that far from it now - how many cars are PCP leases?
As for Volvo, the '70 year' problem is again why we will move away from personal ownership to a service model. That puts the service provider back in charge of asset life. We're not even that far from it now - how many cars are PCP leases?
Yeah, I'm totally with you on a leasing/service approach being the best long term option.
Even Volvo are nowhere near capitalised enough to meet insurance standards though. Insurers have to hold enough spare cash to be 99.5% sure they will still be solvent (including enough to pay off the premium of transferring old business) in a year's time. Motor manufacturers are held to no such standard. And the shareholders of motor manufacturers are not investing to take on long term liability business. Some kind of risk transfer product will need to be created but it isn't obvious right now what that should look like.
Maurice Picarda
Actually, might as well flounce.
I'm not convinced about this bright future where we all hail Jonnycabs instead of owning cars. It's a very metropolitan fantasy (and futureologists tend to be metropolitan fantasists, often unencumbered with children). Even in suburbia I can see the wait being irksome; in rural areas it would take far too long to get a driverless car to turn up. And parents customise their cars with child seats which are age-specific and often integrate with other kid-wrangling kit, with dedicated car fittings. This couldn't easily be replaced by a glorified taxicab model.
There's nothing wrong with paying people to chauffeur you as long as there's a clear agreement upfront that they aren't allowed to put Magic FM on.
mauvais
on reddit or something
Possibly, but even if you spend most of the time living in a remote Scottish castle and the thing stays sat on your drive most of the time, you still don't need to own it. You'll just pay a lot for the dedicated usage. Plus even then you might not be the only user.
Most people don't need that, and it'll be easy to properly commoditise car journeys.
Fez909
toilet expert
The DLR is driverless.
Also, this lot: List of automated urban metro subway systems - Wikipedia, the free encyclopedia
mauvais
on reddit or something
I don't think this is your point, but driverless trains are largely feasible because they essentially operate in a vacuum - a mostly sterile, segregated, regulated environment, where not that much can go wrong, and that which can can't necessarily be usefully addressed by its meatsack operator anyway.
Aeroplanes are next. The sky is largely a segregated and regulated environment too. The technical capability mostly exists, I think. Then the Venn diagrams of man-vs-machine accident cause and accident prevention are slightly more interesting, but not an impossible barrier.
Cars on the other hand are a mess. Motorways are reasonably good parallels with the above - segregated, mostly sterile networks with high commonality between users - but off them, and it all starts going wrong. Bicycles, pedestrians, cows, boulders, floods, drunks, idiots, etcetera. That's a very difficult set of problems to solve with a palatable outcome, especially given our appetite for mistakes and attitude towards blame. Comfortable enough to blame a human cockup, but...
I foresee the creation of a range of low-skilled jobs to supplement the self-driving cars. Car Assistants who can help you get into and out of the car, carry out small repairs and make certain decisions (such as whether to park illegally for a brief period) that machines would struggle to do.
weltweit
Well-Known Member
That is possible, but I was briefly involved in automotive electronics and have a distrust of complex manmade safety systems.
My belief is that if autonomous cars are the future then they will emerge from a step by step incremental addition of features over some generations of cars. As has been the case with safety features to date like airbags, ABS, ASR, anti-trap windows, cruise control, automatic parking etc ..
I don't know if you are thinking from the point of view of a city dweller where car ownership is already a great pain vis a vi parking etc but for rural dwellers I think the situation may remain different, we need a car for everyday living and calling a shared Uber style auto vehicle is likely to remain problematic in the countryside.
If auto cars are like the Audi in iRobot in which you can drive in auto or manual mode and they have come about because of incremental development and can be trusted then when they come I can see myself using one, owning one even, but the extra systems required for the car to drive itself are likely to result in a more expensive purchase price so I think they will also trickle down, like the innovations I mentioned above.
mauvais
on reddit or something
We can't go on down this path of half-arsed, set piece implementation, like motorway-only autopilot or 'keep your hands on the wheel' operation. It's not a workable long term proposition in consumer or safety terms. At some point the jump is going to be made.
Car ownership in general is a pain, no matter where you live. Huge upfront costs on a depreciating asset, legislative burdens, engineering headaches, poor use of capital on an hourly basis for most people, etc.
As I pointed out, the end of the private ownership model doesn't mean you can't always have a car on standby.
The R&D cost will no doubt rise at first, but the manufacturing costs will fall. System designs and licensing are expensive, but the hardware of sensors and computers are cheap. Some cars now probably possess most of the kit required for the job. I say manufacturing costs will fall because there's also a high cost associated with current user controls like steering columns, throttle systems (which also have to be developed) etc that become redundant in fully autonomous vehicles.
weltweit
Well-Known Member
This response makes me question your whole view, cars are complex machines and the safety implications of even small changes are often immense. Car manufacturers are conservative, risk averse and they and their suppliers test each new component ad nauseam in every condition the vehicle is likely to face and some more at great expense. In addition to testing components they also test systems and complete vehicles.
I don't know exactly what you mean by "At some point the jump is going to be made." but if you are suggesting carmakers are going to jump from largely manual vehicles to full autonomy this is just so unlikely it is hardly worth thinking about.
mauvais
on reddit or something
It's intolerable to have this mixed-mode operation where the car will conditionally do enough serious parts of driving autonomously but the occupant will have to remain in some way responsible for others, so for example a car that drives itself on the motorway but has to be driven around town. It's just about acceptable where there's a clear delineation, but only just, and when you try and blur the lines about who's in charge, it becomes impossible to sustain. People will be killed as a result, and the legal and consumer outcomes will be far messier than the car having been completely responsible. A human who is in charge only occasionally is not in charge at all.
That's the safety argument; the sales argument is similar. A car product or car service that offers the benefits of full autonomy will be attractive, but a half-arsed one that is neither here nor there is a confusing proposition that mostly says the manufacturer doesn't have faith in any of it.
All this means that you cannot get to full autonomy by gradual adoption. So at some point soon, manufacturers will either have to suck it up and make properly autonomous cars, or not.
weltweit
Well-Known Member
Gradual adoption is what is actually happening, and I would argue unless Google significantly disrupts the market my bet is it will continue.
We now trust upmarket cars to maintain their speed on a motorway with cruise control.
We now sort of trust some upmarket cars to park themselves. (some)
We now trust some upmarket cars to warn us on a motorway that we are drifting out of a lane.
We trust many cars that they will prevent us locking our brakes in an emergency.
Varying degrees of autonomy are being tested including an Automated HGV tested by Mercedes, but full autonomy is a giant leap, presently into the dark.
We now trust upmarket cars to maintain their speed on a motorway with cruise control.
We now sort of trust some upmarket cars to park themselves. (some)
We now trust some upmarket cars to warn us on a motorway that we are drifting out of a lane.
We trust many cars that they will prevent us locking our brakes in an emergency.
Varying degrees of autonomy are being tested including an Automated HGV tested by Mercedes, but full autonomy is a giant leap, presently into the dark.
mauvais
on reddit or something
Doubtful, I think. They claim this about Tesla and certainly it's slimmed down from a vehicle with an internal combustion engine, but certainly not to that kind of level.
mauvais
on reddit or something
These are mostly set pieces with clear delineation. Do you fancy your chances of suing the manufacturer after you relied on a non-operative lane drift warning and had an accident? Noone explicitly delegates to ABS either, going around supposing it will save them from something, any more than they do the crumple zone.
Adaptive following cruise control is the closest to the blurred threshold of responsibility, and has some of these problems itself, but it's only one element of driving. I don't believe the trend can continue with humans playing a minor role, for reasons already articulated. Again, you are either meaningfully in charge, or you are not.
weltweit
Well-Known Member
beesonthewhatnow
going deaf for a living
I spent most of this week mixing the sound at a large conference for a well known car maker.
From what they were saying anyone who thinks EV's will become mainstream in the next few years is going to be disappointed, and as for autonomous cars? Nope. We are talking 10 years at least for even moderate level of EV's. The fully autonomous stuff is ever further away than that, if for no other reasons than some colossal legal arguments that are going to happen.
From what they were saying anyone who thinks EV's will become mainstream in the next few years is going to be disappointed, and as for autonomous cars? Nope. We are talking 10 years at least for even moderate level of EV's. The fully autonomous stuff is ever further away than that, if for no other reasons than some colossal legal arguments that are going to happen.
beesonthewhatnow
going deaf for a living
Oh, the other interesting thing was talk about Apple and project Titan. They genuinely have no clue what they are up to and as such are rather worried about it. There was a lot of talk along the lines of "look what they did to the music industry..." 
Artaxerxes
Look out, he's got a gnu!
Dogsauce
I’ve got Skol but I’m not a Scholar
Every suburban wanker will buy autonomous cars for their kids to do the school run in, so they don't have to. It'll be like those student wankers with brand new minis they were bought for their 18th birthday, only worse. Twats who buy nobbish cars to project power will insist on getting the same for their 12-year old to roll up at the school gates in.
mauvais
on reddit or something
Apple are never going to actually build a car - you can chisel that into my gravestone - because if/when they sell one it'll be built by someone who already makes cars today, e.g. Fiat-Chrysler. Mass producing cars is hard. Ditto Tesla to some extent and the translation from niche to mass production if they don't go bust before then. So any such product might be interesting but not world shatteringly revolutionary.Oh, the other interesting thing was talk about Apple and project Titan. They genuinely have no clue what they are up to and as such are rather worried about it. There was a lot of talk along the lines of "look what they did to the music industry..."
I agree that EVs will take a long time, if only because of the supporting infrastructure. It's not technically difficult or exotic but someone would have to actually build it, and significantly ahead of demand. Based on the current rate of progress it's probably at least a decade away. Autonomous cars might well arrive first, in fact, because they solve the problem to some extent - they can take themselves to the infrastructure.
hash tag
Books, not bombs
Crispy
The following psytrance is baṉned: All
why not?
