UKC

When their income from speeding fines dries up.

I know it has existed for a few years but I don't profess to know very much about the technology behind driverless cars.

However one thought occurs to me. I assume there will be some sort of means by which driverless cars will interpret speed limits and that, consequently, all cars will "proceed" within the legal speed limit and that there will no longer be any prosecutions for speeding.

Over to you...

In reply to FesteringSore:

I wonder about whether the insurance companies will have to pay up if a pedestrian gets hit? Humans can detect slight movements that suggest someone might step out right in front of them, you can pre-empt from really slight body movements. I can not imagine a computer will manage that, so the owner of the car can presumably not be held liable for someone's death / injury in an accident. Does it become the fault of the manufacturer?

In reply to girlymonkey:

Yes, I think there are a lot of things that need to be considered. Will it be legal to be over the drink/drive limit in such a vehicle or will it become illegal to down five pints and then travel as a passenger?

In reply to FesteringSore:

Unfortunately I'd imagine that the "driver" will still be "in charge" of the car, hence liable.

In reply to FesteringSore:
Nothing compared to the dramatic shrink the in insurance industry, body shops going out of business, the end of taxi driving as a "profession", the end of speed awareness courses, no more driving instructors etc. Lower maintenance from more sympathetic breaking and acceleration, higher density "outsourced" parking where cars wait outside urban areas and park precisely without doors needing to open, fewer cars made due to increased viability of shared ownership schemes as they're no longer limited to areas of high housing density. No more unaffordable insurance pricing young people off the roads.

Beyond all of the above cutting the cost of motoring, the benefit for disabled and elderly drivers is immense.

Biggest societal change in a lifetime perhaps.

Also, once the cars are driving themselves safely, perhaps death by dangerous driving will come to be regarded and sentenced with the severity it deserves.

In reply to girlymonkey:


It will be interesting to see the statistics if and when driverless cars go mainstream.
Personally I think the computer will be able to register the odds on someone walking out but even if they can't will the fact that they will have multiple sensors that are always on and not distracted by phones, advertising, short shorts, kids being noisy in the back etc reduce the overall number of accidents.
Also I think response times will reduce the injury level of accidents that do still happen. When I was about 14 a kid ran out in front of my mates mum and bounced up the bonnet and broke the windscreen.
I remember traveling at least 10 mts or so before the shock cleared and she hit the breaks. The computer because it is looking in 360 degree's would have been on the breaks as soon as the kid emerged from between the parked cars and already slowing down.

In reply to girlymonkey:


Eventually, yes. They'll have insurance for it just like drivers do now. Only it'll be a load cheaper as the machines will end up better and more perceptive than most meat bag drivers.

In reply to FesteringSore:

The google driverless car wiki makes for interesting reading. Just like some people, the Google cars can not tell the difference from police and pedestrian's.

In reply to wintertree:


You've a lot more faith in the meat bags writing the software and testing the algorithms than I have.

In reply to FesteringSore:

They will never be put into use in any of our lifetimes.

In reply to zimpara:


What makes you say that?

In reply to wintertree:

Surely "Death by Dangerous Driving" will be eliminated as will Drink/Driving because there will be no driver.

In reply to Ridge:


Do you get on jetliners? They land themselves half the time these days. Unlike an aircraft a car can stop in a worst case scenario. Software would have to be really bad indeed to be as bad as most drivers.

I don't need faith in the software writers though, the evidence is mounting. 300,000 miles with the Google ones alone. It's very early days yet...

Also every driver has to train from scratch. Every self driving car can pool it's experiences. To me the only question is when, not if.

In reply to FesteringSore:
I can't wait for them, imagine you want a quick trip up north catch a route in nick on the Ben. Get in the car get in your sleeping bag, press go wake up when you get there, yes!
As for speeding fines, i'd imagine its small fry when compared to cleaning up the mess bad drivers leave in their wake.

In reply to FesteringSore:


Indeed, but it's a long way from self driving cars being in public use to manual driving being banned.

In reply to FesteringSore:

Can't be any worse than the standard of driving shown by most drivers

In reply to Zoro:


The mind boggles at the thought of hordes of cars going up and down the motorway system with their sleeping occupants totally oblivious to their surroundings with loud snoring accompaniment

In reply to FesteringSore:

Since we're all speculating, I'll join in.

I think it's all going to come quite quickly - the Google car has been going round 'Frisco for months.

I hope so - for all the reasons wintertree pointed out.

Eventually we will get to the point where it's cheaper to insure if you *don't* drive the vehicle yourself.

However, taking rental cars up worn out alpine roads and swerving the rocks and rivulets will always need a pro at the wheel...

In reply to wintertree:


20 years if they work? If every car is going at the regulation speed and the regulation distance what'll be the point of driving your own car anyway? The speed sensors will stop you or fine you. Manually driven cars will seem like horse drawn carriages.

We'll look back on the "car century" as weird just as we do red telephone boxes.

And Jeremy Clarkson will be out of a job….

In reply to Postmanpat:


Bring it on!

In reply to JJL:
Haha, no chance! Those are exactly the roads i wan't the car to drive up!
I crap myself every time!
I dont think it'll be that long either, you can buy a base model citroen with technology to stop you accidentally swerving motorway lanes, and anti-collision warning. I'd like the first of these cars, my wife is a Parisian........

In reply to Andy Hardy:


Currently the trend is to have the driver still in charge with an override in case something goes wrong but cant see that lasting since it just doesnt make sense. Since chances of the driver having maintained enough focus to take back control at 70mph 3 hours into a journey before the crash happens is pretty minimal i would have thought.

In reply to Ridge:


As a developer I dont.
However I have more faith the risk of the meatbags writing the software is lower than the meatbags driving.
At least you can apply slightly higher standards to the former and use some pretty well tested methodologies to reduce the risk (as wintertree says aircraft have some pretty good standards).
That said given the incompetence the car companies are showing currently with security technology it doesnt bode well.

In reply to Postmanpat:


I just don't know, it's such a hard one to call. It probably will take a generation. Given the cost of insurance to young drivers, self driving cars will be a no brainer to those born this year. The tech will easily trickle down to low end cars in time, and why would you fork out for lessons and £2k/year on insurance just to drive yourself in ever worse traffic?

I wonder if it will be as easy to automate emergency response vehicles? A mountain rescue car that needs to go at speed on the road and robustly and carefully on the fells might be quite a challenge without a mass market use case.

In reply to wintertree:

There is one fairly large market use case for a not dissimilar scenario. Indeed its the one which seriously kicked off the self driving car research.

In reply to FesteringSore:
One driverless HGV running night and day from warehoues to supermarkets would save £125,000+ per year in salaries (5 full time drivers) .
That sounds viable now for deliveries or maybe taxis and busses but not private cars - you're not going to pay £150,0000 so that your car can drive you home from the pub.

In reply to wintertree:


>
Well Obviously I don't know either! But I do think that the J Clarkson petrol heads are a minority. There are a lot of people who find driving a stressful tiring chore and , I suspect, a vast mass who could take it or leave it. If the choice is between driving in a manner highly regulated and restricted by technology, or sitting in a car playing on the computer or reading a book and getting to their destination at exactly the same time then they'd choose the latter.

If the economics-insurance etc favours the latter then if becomes a "no brainer".

In reply to dissonance:


Perhaps. I suspect that'll be a much harder one to get right than well defined road driving, and they'll stick a few extra naughts on the end of the price tag... I also think the challenges there make some of the fast running legged vehicles from General Dynamics an equally viable contender.

In reply to Dax H:

"Personally I think the computer will be able to register the odds on someone walking out"

There was an article on Wired not long back where they 'test-drove' a Google driverless car. One things that sticks out is that the car actually had a bit of a whitey when it was at a crossroads and there was a pedestrian that looked like they might what to cross the road. The car saw them, and the other car waiting to turn at the crossroads, but couldn't/wouldn't go because, although the pedestrian was waiting for the car, the car was playing it safe on the odds that the pedestrian might step out.

Sounds a little convoluted, but supports what you're saying here! And that's all dependent on my memory replaying the article correctly...

In reply to FesteringSore:

My other hesitation is price. At the moment I can buy a car for £400, and do most of the repairs myself, and it costs me £150 per year in insurance. There is no way a driverless car is ever going to be able to be fixed by the owner, so this just escalates the cost, and it will never last long enough to be cheap. My £400 car is 16 years old, no electronics last that long. I don't even want modern manual cars because there are too many electronics to go wrong. Our other car is much newer, and I don't like the amount of gadgetry in it. The driver's window currently doesn't work, and I am not paying to get it fixed! I guess once it gets warmer I will have to delve into the electronics, but mechanical stuff is just much simpler to do yourself!

In reply to girlymonkey:


Yee of little faith. My last diesel car is still going strong with its new owner, just passed 16 years old and coming up to 200k miles. Stuffed full of electronics. Some of the peripheral sensors fail and get replaced. The ECU has just sat there in the nasty, hot, noisy engine bay and has plodded on with zero problems. I suspect it will outlast most of the moving parts.

Electric cars herald a dramatic reduction in the number of electronic control units, sensor and actuators required to make a car work.

The self driving car does not add any requirements to the actuators within an electric car, it simply uses the regenerative and hydraulic breaking that is already electrically sequenced and controlled, and it simply uses the electric power steering that is already present. There is no need for physical linkages to a throttle or gear box or clutch. They are going to need a new suite of electronic sensors, one imagines these being relatively standardised modules that can be swapped with little more difficulty than a lot of current sensors etc. If they can get away with just SONAR and cameras you're talking devices so cheap they're built into car bumpers and disposable cameras. If they end up needing LIDAR or RADAR it's going to take more work to bring the cost down.

One of the main causes of electronics and sensor failures in cars is the harsh environment they have to operate in, resulting from the combustion engine. That's all going to go the way of the dodo.

No doubt it will take time for these things to trickle down to the low cost end of the 2nd hand market, but I don't see why they won't. I can go down to a breakers yard and get all sorts of electronic control modules and so on. An electronic module in my car failed last week. I had it out in ten minutes and a compatible one off e-bay was in the hole 3 days later for £12. There are problems with the injectors on some high performance common rail diesels being coded to specific engines, but there are also reasons and these go away with electric cars.

My thoughts are almost opposite to yours, but just as bad for the cheap 2nd hand market. Looking at the bare essentials of the Tesla Model-S in the London showroom you realise just how much less maintenance there is going to be with electric cars - no water system, no oil system, no fuel system, likely to be lifetime breaks (most breaking is by the motor/generator), no clutch, no drive belts, no cam belt, the list goes on. If cars don't need as much maintenance half the driving force of the insane depreciation surrounding car ownership goes away as they become longer lasting assets. Not such good news for buying old clunkers.

Edit: But yes, it's going to be a long time before they've trickled down to the sub £1000 end of the market. Then again, what did your car cost when new?

In reply to girlymonkey:


Mechanically I can't imagine driverless cars will be enormously different to the current crop, so I don't see any reason maintenance costs will be substantially higher. For the software, just plug in a computer, type in a few passwords and you're on the latest version. Easy as that...


The ECU in your car suggests otherwise.

In reply to FesteringSore:

Driverless cars will never happen. Nobody would ever buy one. It would actually be like being in one of those cars with a fish symbol on the boot, and a Panama hat on the parcel shelf. <shudder>
Unless the thing could be programmed to joyride mode, i.e.. wheelspins and handbrake turns and outrunning the Babylon

In reply to FesteringSore:

This will happen pretty quick. Certainly cars could be introduced tomorrow that would not speed as I own a car that tells me very politely when I'm over the speed limit. Also, have you not been a car with automatic parking, and sensors that tell you about bikes, pedestrians et al. ? That's pretty stock on new cars

In reply to wintertree:


Or perhaps higher maintenance/repair costs as driverless cars get older, and the vastly more complicated sensor/computer systems start to fail.

In reply to Phil79:


Perhaps, perhaps not. The future sensors on self-driving electric cars are going to be in a pretty tame environment, where as many of the current sensors are subject to strong temperature cycling and high vibration whilst being bathed in highly corrosive fluids or gasses.

I don't think the sensors are going to be "vastly" more complicated than those already shipping on many high end cars. Likewise the computers needed are going to be positively basic compared to the state of the art 20 years from now. There is plenty of industry experience at making these items robust, modular and easily replaceable.

What would you rather swap, a timing belt or an electronics module?

In reply to birdie num num:


It'll happen because it'll be the only type of car you can buy. Governments will see to that

In reply to FesteringSore:

I wonder if there will ever be an F1 race with driverless cars?

In reply to wintertree:


Personally? A timing belt! But that's only because I'd have some idea what I'm doing, I do see your point though.

Still, I'm yet to be convinced that we are at the threshold of some massive step change in driving technology/motoring habits. I would think general up take of totally driverless cars will be slow, although I can see auto braking/collision avoidance type systems becoming more prevalent, in much the same way ABS and traction control have.

Electric cars still seem hobbled by battery technology TBH, I haven't really been keeping tabs on development, but has any major advance been made, or likely to be made on this front? Useful for urban environments/short trips, but not as a general all purpose vehicle.

In reply to wintertree:

As a motorbike rider, can't come soon enough. No more "SMIDSY". I wonder how the cars would react if I came up alongside them when filtering? Suppose it will all be programmed in

In reply to imkevinmc:


Why would they do that? Have they ever stipulated anything similar before?

I suppose Gov/EU now require new cars to meet certain safety and emission standards, but I cant see a reason they would enforce this, by whatever method.

In reply to Phil79:


Yes and no. The popular view holds that battery technology has not improved in decades, but the energy density of batteries has actually been increasing relatively predictably, with a doubling every decade or so. There's no indication that this is going to stop in the next 15 years - there's plenty of research filtering its way down in to production. It's not major advances so much as persistent plugging away.

The range of electric vehicles is going up, and charging times are going down. Major car makers are now coming out with plans for mid-level cars in the 150 mile range bracket to hit market in the next few years. That would not work for everyone - but it's enough for a sizeable fraction of 1-car households to switch, and for 2-car households to switch their second car.

I think there's a big synergy between electric cars and self driving cars, in that car parking can be outsourced from where you live/work to compact out of town "charging centres" - every day I get to work, the car could drop me off and take itself out of town to charge during the day, then pick me up at home time. Boom - I never have to worry about going to a garage again. Although the generating and delivery capacity is only really there in the national grid to charge a large number of cars over night.

In reply to Phil79:


For the UK:

160,000 people injured.
21,000 people seriously injured.
1,800 people killed.

Whilst an improvement on the past, these numbers are shockingly high - the death rate is 3x the murder rate, yet which receives more news coverage? It's only tolerated by society because there is no viable alternative that maintains our lifestyle.

I don't think human driven cars will be outlawed for a long time, so much as priced off the road by increased test/regular retest and medical check requirements and by the insurance market.

Most interestingly, if car accidents and fatalities are all but eliminated things are going to look very bad on the motorbike side of the fence.

In reply to Bjartur i Sumarhus:


I'd echo that as a cyclist - although given the sheer quantity of sensory data that'll be recorded and shared by these cars all self-controlled 2 wheel users are going to need to be on their best behaviour...

In reply to wintertree:

Is that 150 mile range factoring in heating and/or cooling of the cabin for passenger comfort?

In reply to FesteringSore:
If everyone is in driverless cars, then assuming they can also communicate amongst themselves (or possibly even if not) speed limits on restricted roads (ones limited to auto-driving cars only e.g. motorways) can be vastly increased. People can't be trusted to drive 100mph on motorways, but computers can. You can also pack cars closer together because although you might not be able to stop in the space between you and the car in front, your computer can guarantee that your car will be able to stop before the earliest point that the car in front will be able to stop...

It would suck for me though - I get travelsick if I'm not driving.

In reply to wintertree:
"Most interestingly, if car accidents and fatalities are all but eliminated things are going to look very bad on the motorbike side of the fence."

I commute on a motorbike because it is far quicker and cheaper than either public transport or a car. I don't really see how this will change with driverless cars (certainly for my individual circumstances of working in Central London and living in rural countryside) I would imagine motorbike riding will become safer as well.

How do you think it could effect the motorbike world?

In reply to Bjartur i Sumarhus:


Who knows... They're starting to be economical with the heating though, for example you can get a Leaf with heated seats and a heated steering wheel.

Still, it's early days, and I think a lot of UKC posters have a rather enlarged view of what the average use of a car is. Working from https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/2... :

For example, the overage commuting mileage of a private car is ~2800 miles per year, or 6 miles each way, per day. 6 miles is a very, very long way from 150 miles. You could start a week fully charged, burn as much battery on heating the car as running it, and still have 13% battery left at the end of the week, without ever charging it. Non-commute mileage is harder to portion out into individual trips from this report, but if there was one private trip per week people would average 96 miles round trip. Only 2% of trips in the UK are over 50 miles. Clearly far fewer are going to be over 150 miles.

They're not going to work for everyone, but they are going to work for a lot of people.

In reply to Bjartur i Sumarhus:


Eventually... Collisions between motorbikes and cars should be nearly eliminated, regardless of who is at fault. Single vehicle car accidents should be nearly eliminated. Single vehicle motorbike accidents, involving pedestrians/cyclists or not, are going to stand out like a saw thumb. It seems about 25% of motorbike accidents are single vehicle accidents.

In reply to Phil79:

EU is constantly imposing new regulation. All new cars will soon have an embedded mobile SIM. It's designed to issue an alert in the case of an accident. But why stop there. It will be possible then to remotely disable cars.

And so on, and so on

In reply to FesteringSore:

Calling them "driverless cars" is a bit of a misnomer but cars with some sort of auto-pilot mode can't be far off. I could see motorways with one lane reserved for such cars, all travelling at the same speed, communicating with (and warning) each other. The only reason you'd need to change lanes would be to leave the motorway, so steering would be automatic.

Can't see it working away from motorways due to the numerous other factors. I suppose if cyclists and pedestrians all wore a hi-viz reflective jacket then the car would be able to avoid them. No hi-viz and it's your own fault if you get killed. So not much of a change there.

In reply to wintertree:



Yeah I agree they are shocking numbers, any reduction in road death has got to be welcome.

In reply to Chris the Tall:

Driverless cars can see people/bikes/anything a laser bounces off better/more reliably than people can from early reports. This was an interesting report on the google one: http://theoatmeal.com/blog/google_self_driving_car

In reply to Chris the Tall:


The computers can already see the pedestrians :P

In reply to Chris the Tall:


What's with all the "Oh the car won't be able to do this..." that people on here seem to think? Yes it will, it mostly already can, and better than you. These are all solvable problems with todays technology and a large enough team of researchers and programmers.

In reply to FesteringSore:

I'm expecting the first (and maybe only) incarnation of driverless cars will be that the driver still sits in the drivers seat with their limbs hovering over the controls ready to take action in case the computer f*cks up. Any alternative to this will take liability away from the driver and I can't see that happening unless the technology proves itself to be flawlessly reliable in it's initial incarnation.

Personally if given the option to let the car drive itself whilst I sit there watching the road, ready to take over at any point, I would rather just drive it myself.

In reply to girlymonkey:


So many people are thinking about the same situation as we have now, but the car just drives itself. Instead we're talking about a fundamental shift in the entire 100 year old concept of buying, owning, driving and maintaining a vehicle. I think this is 'a good thing'

Why would you own a car that could drive itself? Since it would spend on average 22 hours a day sat parked, surely it would be better off driving elsewhere, taking other people around. So instead you would PAYG, and get a car to pick you up when you need it and drop you off. We would need far few parking spaces as most cars would be in circulation, so our city streets would be vastly improved.

We in the west will take a while to get used to this, but I would expect in a lot of developing markets they might even leapfrog our model and go straight to collective use of self-driving cars rather than widespread individual ownership. It makes so much more sense on so many levels.

We're also looking at the traditional car manufacturers to evolve this concept but the real innovation will come from the Googles and Apples of this world who have a culture of embracing new markets and technologies, rather than the GMs and Fords, who have basically spent the last 100 years tweaking the same concept and making no real fundamental or innovative shifts in that time.

In reply to Jimbo C:

>I'm expecting the first (and maybe only) incarnation of driverless cars will be that the driver still sits in the drivers seat with their limbs hovering over the controls ready to take action in case the computer f*cks up.

No driver is going to remain alert and engaged if not actually driving, not for more than a few minutes. You might as well be driving....


Why does it have to be flawless? Human drivers sure as sh1t are not. It just has to kill and maim fewer people. There are plenty of life-critical systems in the world where liability rests on something other than a human, and this has not proved to be an insurmountable barrier to either legislative approval or obtaining insurance.

In reply to yorkshireman:


I ventured this opinion in a previous discussion. I was repeatedly told by many people how this would never work. Well, it'd work for me and it'd work for you. I can't wait for a future world where convenient travel by motorcar no longer involves insane capital outlay and loss of value.

Bring it on.

In reply to wintertree:

Will I be allowed to throw rubbish in the back and drop crumbs and get in with muddy shoes on and not bother about hitting the door when getting out in narrow spaces?

In reply to dsh:

Maybe. I suspect the problems are bigger than you suggest though as soon as cars leave well maintained roads in daylight in good conditions.

In reply to Jimbo C:


This is already happening. See Google

In reply to MG:


Fine - you treat your car as you want and pay for the privilege because you can't be bothered to take basic steps to look after it.

Me, I'll enjoy much lower cost motoring.

Mind you, parking scratches and scrapes caused by incompetent parkers and people who can't look before opening a door should be eliminated, because the car will drop you off at the door or waiting bay and then go and park itself with a modicum of competence. As for muddy shoes, I often have these so I have an easy to clean interior and not fancy leather/fabrics. I'd be quite happy to go shared ownership on something like that.

In reply to yorkshireman:
There are so many possibilities. In your model, the difference would largely be to urban spaces, with only about 10-20% as many vehicles as before, but all of them moving , instead of 80-90% of them parked at any one time. No huge car parks needed, no individual garages needed, no cars lining every street from one end to the other etc etc.

In reply to wintertree:


I wasn't entirely serious! But I think there will be a cleaning thing to get sorted and the likes of Avis and Google are probably loving the whole idea.

In reply to MG:
Based on the million(ish) accident free miles for Google alone, the driverless cars have fewer problems interacting with humans than human drivers.


Yes. I think the Google cars can't cope with roads obscurred by snow or heavy rain and they can be blocked by crumpled up paper which it can't distinguish from a rock.

In reply to MG:


It'll give all the unemployed taxi-drivers something to do!

In reply to elsewhere:

I like the crumpled paper! I wonder about things like single track roads too. How will they distinguish soft and hard verges - or will they suddenly reverse a mile to the last passing place! I expect rather a lot of hilarious "my car drove into a manure heap" type stories in the first few years. I have a suspicion Google are cheating in various ways currentyl and intially the focus will be on "driver aids" rather than no drivers.

In reply to wintertree:

Maybe some concerns here are because some posters simply don't realise these things are already being widely tested pretty successfully (so far) in real road environments. The pedestrian at the traffic light scenario was fun but for every one of those I'll give you at least one incompetent driver doing something worse staying stationary for no good reason let alone all the hundreds of stupid dangerous things that happen at speed.

http://en.wikipedia.org/wiki/Google_driverless_car

In reply to dsh:


The problem isn't the technology, it's the acceptance. Will people accept the (perceived) risks and pay the costs for benefits. I can't see much benefit in having an automated car in an urban environment, and as a cyclist I fear the risks.

However I can see huge benefits for motorway driving.

In reply to Offwidth:


Or not (from your article below). This will happen but I reckon 20 years before truly driver-less cars are widespread.


As of August 28, 2014 the latest prototype cannot "handle heavy rain and snow-covered roads".[30] Functionally it can go at sluggish speeds when crossing an unmarked 4-way stop due to the algorithms of the computer taking extra precaution. There are also other limitations on discerning objects such as trash and debris that can unnecessarily veer the vehicle. Additionally Chris Urmson of Google has said that the lidar technology cannot spot potholes or humans, such as a police officer, signaling the car to stop.[31]

The vehicles are unable to recognize temporary traffic signals. They have not proven themselves in snow or rain. They are also unable to navigate through parking lots. Vehicles are unable to differentiate between pedestrian and policeman or between crumpled up paper and a rock. Google projects having these issues fixed by 2024.[32]

In reply to wintertree:


My current car (18 months old Volvo V40) can currently parallel park itself. I press a button and it scans the cars on the side of the road and beeps and tells me when there's a big enough space ahead. I let go of the steering wheel and it reverses in automatically steering hands free and using the radar sensors to keep away from other vehicles and the kerb. It will also slam the brakes on automatically in city traffic (and unfortunately if I approach a toll booth too quickly) if it looks like I'm going to crash into the back of somebody.

I've seen articles where the imaging technology in driverless cars can detect body language in (riding) cyclists and anticipate when they're about to turn or pull out even if they don't signal.

With adaptive cruise control and lane deviation technology we're creeping towards the self-driving future and we'll get to universal acceptance one day. The problem with driving is people forget how ridiculously dangerous it is, and would never doing other things that are objectively less dangerous but seem so (climbing springs to mind - but this is the wrong forum for that I guess!).

Quite simply if any other technology killed 1,000,000 people globally every year, as vehicles do, we would be clamouring for it to be banned.

In reply to wintertree:

My car is a mobile kit store! I don't want it to dissapear during my working day with the stuff I'm not taking on the hill!!

In reply to girlymonkey:


I think the google one can do this. I recall a story of their car at a junction and a woman on a side walk looking like she wants to cross having a standoff as to who moves first.

You could argue that the car can compute more things and concentrate on more things, so it will preempt this happening in the first place by spotting the pedestrian in a dodgy spot and slow down before the pedestrian even knows he wants to step out.

As for the careless person who comes from behind a parked truck and walks out without the car having a chance to see them, then I don't suppose there is much to be done about that. But the on board cameras (and there will be cameras) will show its the pedestrians fault and no insurance claim can be made.

In reply to Postmanpat:

I think you're right. Driverless cars will be with us pretty soon I think but the reality is that human driven vehicles will have to be banned outright. Much better that way for all concerned since every driverless car can/will be aware of the location, speed and direction of every other driverless car in the vicinity.

In reply to wintertree:


Nonsense. Some can, very few do.
jk

In reply to FesteringSore:

IF we ever get them, my guess is that they will quickly become as reviled as caravans, as no doubt they will strictly adhere to the speed limit and cause traffic build-ups behind them. As how many people actually drive at 30MPH in most 30 zones? Even the local police cars do 40 around here.
OK, so Google has a test vehicle which probably has a team of technicians looking after it and has cost them upwards of hundreds of thousands of pounds. Maybe they will become a reality in 10-20 years but don’t expect them driving around your street anytime soon, so plenty of time for the gov to argue about the laws and the insurance companies to find ways to cover themselves when they mess up.

In reply to Mark Edwards:


Err, me. An urban area is the most dangerous place to use excessive speed. I've been known to 'have fun' on empty country roads, but will always reign it in for 30 areas.

In reply to Mark Edwards:

I think if all or at least the majority of vehicles were obeying the speed limits and/or responding automatically to the traffic flow then jams will be greatly reduced. Driverless cars could be programmed to follow strict rules so that they never follow too closely to the car in front and respond to the traffic around them. This would then allow none driverless vehicles to filter in to and out of the gaps etc. I'm all for it, the sooner I can ditch the driving the better as far as I am concerned we are 10 years too slow with the technology!!

To be honest in my opinion the sooner vehicles are built with technology that automatically regulates their speeds in different zone the better. I.e. automatically adjusts the maximum speed for the zone that you are in so no more doing 30+ in a 20mph zone or excessively breaking the speed limit in a 30mph or even on a motorway etc etc.

In reply to FesteringSore:

Driverless Cars - What will the government do?

That rather depends on how rudderless the government is...

In reply to MG:


The idea of shared ownership is already functioning with Zipcar and the likes. I'd imagine if cars are built not to be owned, the interior wouldn't need to be as flashy and could allow for easier modular replacement at low cost to keep it all looking top notch.

In reply to jkarran:
Very few do simply because of the conservatism of the industry, and no doubt the strong pilots union trying to keep the crew relevant. Also, keeping in mind that most technology in aircraft, even the latest such as the 787 and A350, is at least 10 years old, or more commonly 30 years old, simply because commercial aircraft aren't landing themselves doesn't mean the technology doesn't exist. UAVs give a far better picture of where the technology is at, as do low-cost carriers that do everything possible to remove the pilot involvement.

Apart from very few accidents, pilots appear to a bigger risk to passengers than automation is.

In reply to David Martin:


I was going to note - the times I've been on a flight and the pilot announced that the plane just landed itself were in conditions of appalling visibility.... I suspect this is going to end up a lot like the issue with self-driving trains on the underground...

In reply to girlymonkey:


I'm sure you can hire a car for the duration of a trip.

Me, I'm excited about self-driving cars for days out walking. The last time we did a 30 mile walk it involved going out of the way at the end to a bus stop, waiting for 50 minutes and getting a slow bus back to somewhere else we could get another bus from to get back to the car. I'd love to have one that can drop me off and then pick me up elsewhere hours later, with no worries about the lass bus.

As for storing stuff in the boot - imagine that the boot is a modular cargo unit that can be removed and compact stored by a robot. Now people like me in little terraced houses without space for everything can have different boots, one for climbing/walking, one for commuting and one for long trips. They can all be stored compactly in some facility, and no matter what car model I summon, I get my stuff. It'd be almost like having a TARDIS.

(Although I do know people who appear to live out of their cars who this would never suit...)

In reply to Mark Edwards:


The current speed limits are there because humans have to drive cars, and as the million-plus fatalities worldwide every year demonstrate, we're not actually very good at it.

What is more likely to happen in the early days is a two-tier system - with say for instance the outside lane of motorways given over to 'trains' of driverless cars being allowed to do (say for instance) 100mph at 10cm apart, because they're all WIFI connected and there's no delay in braking or speed adjustments, and they will save fuel by drafting each other. Even just doing this at 70mph, the overall journey speed would be quicker than the current system of speeding, braking, ripple effect etc.

All the human-driven cars will have to stick to the other lanes and contend with the stop/start braking and concertina effects that impact all motorway driving at the moment. Surely

I fear life on the roads is going to get very boring ;0(

In reply to FesteringSore: Driving, transport will already be a completely different experience. I was going to say it's already boring, but being bored can give you time to think - the worst case scenario is that it will be a new chunk of time that can be spent working.

Really - how long do you think till there is automatic speed restriction - new cars have a GPS, and know the speed limits. Your throttle is an electronic solenoid. You have cruise control - I imagine the cose already exists, but isn't implemented to limit maximum speed. I guess you'd need to consider overtaking situations but otherwise.
One outcome would be that as most new cars won't speed, speeding is a conscious decision by the driver, and speeding fines can be commensurately increased. You could also mine the GPS data and retrospectively fine

In reply to David Martin:


But for privately owned cars the opposite may be true. Since they all go at the same speed the way to differentiate will be the quality and the gizmos inside: large screen TVs, beds, bars etc etc.

In reply to David Martin:


I didn't say nor think for one moment the technology doesn't exist, it clearly does and has done so for the better part of 50 years. I derided the assertion that half of all landings are fully automated, they're not, nowhere near and for a variety of good reasons beyond conservatism and unionization.

Anyway, we're drifting off topic, aircraft automation is a rather different problem to automating personal road vehicles, they operate in radically different environments.


For the present mix of manual and automated control I might give you that while bearing in mind the pilot is often at the controls long after the automation has switched from flying the aircraft to flashing orange lights and sounding warnings. Highly automated cockpits aren't a bad thing but then nor are well trained and current pilots, they both have their place.

jk

In reply to girlymonkey:


Check some of the videos by Tesla and Google. It seems computers are far better than humans at factoring in human error than humans are...

D

I fear driverless cars will only be very slow and hesitant, causing awful congestion. There is no substitute for intelligence, and they will not have any. Hopefully I'm wrong.

In reply to Dauphin:



Yes, although I think they'll find that the driving conditions on Axe Edge in the fog with no GPS coverage, no white lines at the edge of the road and sheep randomly lying on the warmest bits of tarmac are a little more challenging than the standardised 4G environment in California.

In reply to David Riley:

As opposed to now?
It depends on the approach you take. Stand alone driverless cars may or may not be better than people. However if you networked them all together you would end up with less congestion eg as others have mentioned motorways could flow faster and more packed in. Well until the point the bloke driving his classic focus manages to bugger it up.

In reply to Postmanpat:


You could still use the same modular approach but just slot in the flash interior.
Its supposed to be a trend in New York for example to have vans customised to be luxury transport. More room than a Bentley etc and also more discreet.

In reply to Postmanpat:

Well at present an awful lot of people are now purchasing cars on fuel economy, not speed...

Neil

In reply to dissonance:

Much worse than now.
Speeding up could only happen by separating the road areas like a railway.
Then congestion would increase due to loss of road space.

In reply to David Riley:


its depends how quickly you stop the meatsacks driving.

In reply to Neil Williams:



Same result. Presumably they'll all use much the same fuel/electricity so the interior becomes the differentiator. A bit like horse drawn carriages!

In reply to FesteringSore:
Ideal for long straight walks, drive to the start, walk to the end , and the 'driverless' car is there waiting for you, no need for two cars


Edit, Just spoted Wintertree was there already.

Can't wait to be picked up after a long walk and not have to concentrate driving on the likes of the Loch Lomond Road on the way back (in the dark,) instead I will be sleeping like a baby (maybe)

In reply to Postmanpat:

I doubt it - there will still be different sizes and classes of vehicle, I would imagine.

Neil

In reply to FesteringSore:

I can't remember exactly where I read it... but initially at least, the 'driver' will be as responsible in a driverless car as they are at the moment. And the driver will have to sit, fully alert, at a full set of controls/pedals. I think anyway. So you won't just be able to go to the pub and get it to pick you up after 5 pints.

In reply to yorkshireman:


The falling number of road traffic accidents in this country seem to indicate that we can get better at it. And if any gov had the balls to make people retake there driving test every 5-10 years, with the test getting ever harder they would fall even more. We might not be very good at it but we are far more capable of that sort of task than machines.


How wonderfully sci-fi and completely unrealistic. Do you really think that as soon as we have a few of them wandering around that we should cut the capacity of the motorway network by a third?
Until they prove their reliability they will probably be little more than golf carts and restricted to crawling around city centers, blocking up the bus lanes not racing down the fast lane. 100MPH Cm’s apart? I can’t see any manufacturer certifying their vehicles for that sort of use, as mechanical/electrical failures are still going to happen and it would only be a matter of time until people start being killed, although the greatest challenge would be getting from the entry lane to the fast lane and back again, which could generate some amusing you tube clips.

In reply to Mark Edwards:

It might be that cars being closer together reduces the lethality of collisions. If I'm following a car at the same speed and it brakes suddenly, my impact force will be less if the car was only a few cm away than it would be if it was 20m away (and therefore on the brakes longer and at a much different relative velocity by the time impact occurs).

In reply to FesteringSore:

http://www.thedailymash.co.uk/news/science-technology/obsolete-drivers-will...

In reply to David Martin:

If it's a perfectly straight shunt you are likely to be right - but if it isn't and one car (or both) goes spinning off and hits other things, that's when people really get hurt or killed.

Neil

In reply to David Martin:

Now you have two entangled cars with compromised structures, out of control, still at speed. It's not hitting the other car that kills you in that scenario, its what happens after.

A lot of what's said about driverless cars is for the foreseeable future pure (and often silly) sci-fi, they're going to have to coexist with other road users for many years to come and they're going to have to obey the same rules without any special infrastructure to support them. Likely we'll see a growth in what we already have: Auto-park, auto-brake, auto-speed limiting, smarter cruise control that can maintain gaps, 'see' through fog/darkness, 'look' far ahead at speed and stay centralised in lane. After that perhaps with lane changing/navigation abilities and collision avoidance, likely restricted to use on motorways and dual carriageway where traffic flows are segregated and always requiring the human driver to be notionally in control (fat chance) for legal purposes. Once they reach a critical mass the road rules may evolve to make better use of their capabilities, drag reducing trains and the like but I really doubt we'll be seeing commercially available driverless cars cruising around empty, running errands for their owners or cars ferrying snoozing drunks home in the next 20 years.

jk

In reply to wintertree:



I am one such person, in fact I often live IN the car, a criteria for buying a car is whether my husband can lie down in the back to sleep.

I wonder if their big use would be a park and ride system for cities. Whereas personal ownership seems less likely. I think normal cars are likely to remain alongside

In reply to girlymonkey:
I don't think they have much of a use for that. P&R relies on getting people together in a car park then getting them in - it isn't just about reducing city centre parking - indeed, with underground and multi-storey car parks that (unlike road capacity and pollution) is an easy problem to solve.

P&R needs buses, trams, trains and such - mass transport. Sure, you could have a self-driving bus, though. And you already have self-driving trains.

Neil

In reply to Mark Edwards:


Number of accidents or number of fatalities? The large drop in fatalities in the last 20 years in the developed world has been largely attributed to the rise of mobile phones and so less delay for first responders, and increases in car safety technology. I don't think driver skill has dramatically improved.


No, and you know that's not what i'm saying. But 20 years ago talking to somebody's moving face on a touch sensitive screen while walking down the street was the stuff of sci-fi, widesspread wifi, GPS and 3D printers were stuff people couldn't really get their heads around until it really got into the hands of mass consumers.




All the technology exists - its essentially wifi and adaptive cruise control. The technology isn't the problem here, its legislation, infrastructure and consumer acceptance. Just because it freaks you out doesn't mean it won't become commonplace. Just look at people's reaction to traveling on 'fast' moving trains in the early 1800s - we laugh now. One day we will think it weird that we ever let humans do something as potentially dangerous (and quite frankly, menial) as driving.

In reply to Neil Williams:

I guess my thinking is that some people don't really like the whole communal transport idea, so currently take their own car in and have to find parking. If these were available as a park and ride type system (suplimentary to the current public transport options that are there) then it would cut down on city centre parking for those won't use public transport.

They have a lot of developing to do before they become a reliable long distance option - I don't fancy having to try and charge one anywhere between Perth and Aviemore!! (normal fuel is non existent on this stretch after 10pm!). So initially anyway, it seems they would be more of a city car.

To the people saying the car would pick them up at the end of a walk - how do you suppose you would summon the car? Pre-program, so walks cannot change their plans? Text message, so need mobile signal? I like the idea,hitching back can be a pain, not sure about the practicalities!

In reply to yorkshireman:


Ha ha, the only one of these things you have mentioned that I use is wifi - I think driverless cars are at least a lifetime away from me!!

In reply to Trevers:


So you get the worst of both worlds?! You have to be alert with nothing to do which will keep you alert?! Great, technology is wonderful stuff eh?!

In reply to FesteringSore:

What will driverless cars do when four meet at a four-way roundabout? One has to go first, so how will that be decided? I can see the manufacturers engineering too much safety margin into the things so that, faced with a snarl-up, they will just sit passively and not use a bit of nous to get through.

In reply to Rigid Raider:

I imagine the one that gets there a microsecond sooner will go first.

In reply to girlymonkey:

You seem rather resistant to the idea of change!
If I asked you in 1910 if you thought there would be a national distribution network for petrol you would probably have asked me why. Electricity has the advantage that we already distribute it quite effectively around the country we don't currently power many cars with it but there aren't may technical challenges to setting up charging facilities. The biggest challenge is the battery technology to give electric cars a sensible range.

As for driverless cars I agree that in principal I agree we can technically do it but think we are probably 20 - 30 years off acceptance. What I do think we will see is an increase in better and better driver aids that attempt to eliminate the human error factor in driving. You can already buy cars that can parallel park, has adaptive cruse control, lane exit warning, junction shunt prevention, blind spot detection etc. I think the next big leap will be between car communication where the car ahead can tell the car behind that it is applying it's brakes. This will just lead to creeping automation where less and less human intervention is needed.

In reply to Rigid Raider:


What do you think?

The biggest problem with traffic is that everybody is out for themselves. Cooperative driving (not tailgating, letting people out, smooth filtering etc) gets everyone there quicker, but as humans we don't do it as we only care about our own little bubble.

If four cars meet at a junction, they would communicate with each other and determine which order to go in based on algorithms which take in various environmental factors. One of them wouldn't be in a rush, another wouldn't be talking on his mobile and not paying attention, while the third wouldn't be so aggressive that he just 'had' to go first. People think the driverless car concept is just replacing the driver with a robot - its not - its replacing the entire collective consciousness of everyone on the roads with a networked system that can allow traffic to move more freely and cooperatively.

In reply to Andy DB:

Change is fine if it actually makes life simpler, but sometimes it just doesn't! Why have a car that drives it's self if you have to sit alert at the controls? That is not making life simpler. If a car is only going to go a short distance before needing charged and then you have to wait ages for it to charge, then it's worse than what we have. I'm sure this second problem will be fixed sometime.
I currently have a really old skoda that I can fix myself, everything is mechanical, so it's cheap to run. My other car is newer has some fancy computery stuff for heating controls etc which are the biggest distraction in the world to my driving, I hate them, and loads of things are electronic and so have to be fixed by a garage - it's extortionate! (Although it barely has any computery stuff compared to newer cars!). If it costs me more, then it's not making life simpler!
Mobile phones make many things in life simpler, but smart phones break easily and don't work with proper gloves on, so they don't make my life simpler so I don't have one.
I'm all for technology if it actually enhances my life, but not just because we can.

In reply to girlymonkey:


You are confusing technology with people writing laws about technology and pandering to stupidity.
Its something which wont last beyond running a few trials where you test whether someone can react in time after doing nothing for an hour.
Although some manual controls may remain more along the lines of a joystick for certain maneuvers eg parking offroad or some random campsite.

In reply to girlymonkey:


As you say though its a problem you have now. So you may need to retain hitching for emergencies but majority of the time you would be able to use it.
Possibly set it up with a range of options to fall back through.
That and another advantage is there is no worries about driving back late knackered.

In reply to MG:
"Or not (from your article below). This will happen but I reckon 20 years before truly driver-less cars are widespread."

I agree with the not points but think they will be solved soon and hence the cars will be common sooner (in the same way electric cars are common but very much in a minority now). Some posters above seemed unaware they are legal in places and being used in real road tests.

On the snow, pot holes, debris, police, temp lights problem points: all are fixable in a few years given the impressive progress so far and in any case many drivers of normal cars fail to respond properly to these daily. I remember the childlike anticipation a pal had when snow was forecast such that his new snow tires might get a test run: my cynical response to wait-and-see proved right...he couldn't get out of his street as it was blocked by incompetants.

In reply to Rigid Raider:


They'll be able to tell a lot more clearly than humans what the actual order of arrival was down to milliseconds. If that doesn't decide it, I'll pick a technique out of my ass and go with each one picking a random time to wait before setting off, and then stopping if another car moves first. Standard ethernet technique from the old days of shared pipes, and any abortive motion needed to signal this would probably not even be noticed by the humans.

Or they could just talk to each other and do it virtually in microseconds.

Better still, especially for 4-ways stops in the USA - they'll just schedule their arrivals to miss each other.

In reply to Andy DB:


There's an exponential adoption mechanism forced by this though? The more automation one has, the less engaged a "driver" will be, and the less likely they will be to react in time in any situation that needs their involvement.

I still don't feel I can call the disappearance of habitual human driving to within ±15 years though.

In reply to Offwidth:


You may be right in well-mapped urban areas and main roads but they will still need the back-up of immediate manual override for when the GPS fails.

I also think that you are still over-estimating the ability of an automated system to react appropriately to the genuinely unpredictable. Think how few options you have to input data into your PC and how often the system fails to cope even with that and hangs.. or, indeed, crashes.

In reply to imkevinmc:


I sincerely hope you are right.

In reply to Mark Edwards:


No, but we should slash the amount of valuable land being wasted on car parks.

In reply to silhouette:

What's the difference between a driverless car and a driverless government?

In reply to silhouette:
Build them underground. Require every building to have ample underground parking. It takes no space at all.

Neil

In reply to wintertree:

Possibly so, but realistically because of maintenance, infrastructure etc and the need to make a profit from it (no way will the Government give up all that revenue) I expect the price might be somewhere around the level of train fares, possibly with things like book in advance discounts. So not all *that* cheap, but quite possibly more convenient.

Is it perhaps more likely (as transporting a big tin box up the motorway with one person in it is inefficient) that this kind of thing might just replace taxis, at a slightly lower price than taxi travel, for the local bit? (If the same technology could be applied to trains, which as driving a train is a far simpler thing and there already are driverless trains on city metro networks like the DLR, there would no longer be a problem of "can't run trains at 3am or on Sunday as it costs too much", I guess)...

Neil

In reply to Dave Garnett:
Possibly just a big red stop button as the passenger may not have a driving licence and for freight there is no passenger.

Most accidents are entirely predictable (eg a driver, child or other road user makes a mistake) and the automated cars seem to be making a better job of dealing with those everyday mistakes than us humans.

It amuses me how many people in this thread are saying either:

a) Computers can't/can't currently handle situation 1/2/3 (e.g. four-way interchanges, pedestrians, cyclists), or
b) The cars can't handle situation x/y/z because the infrastructure isn't ready/the cars will need to communicate with each other/it will only work on special lanes etc.

when in fact the Google cars are driving around already without any special infrastructure.

In reply to Neil Williams:


This model of on demand car hire is already proved and working though - and at rates far less than the train. For example Zipcar will do you a VW Polo or similar for £5 per hour + fuel costs (after 60 miles free fuel) + £50 per year. The problem with schemes like this is that they rely on a critical density of users to ensure that they are finically viable and to ensure tat there are enough cars near people. This really mandates dense urban areas - where if you switch to a ZipCar the problem of the cost of parking also goes down significantly.

Self driving cars would extend the business model out to less populous areas and would remove some of the problems balancing numbers around areas. Eventually, the capital and maintenance cost of the car will be much more than the associated costs of the self-driving component.


I live in hope.

In reply to wintertree:
I'd imagine, though, that if this became a day to day operation the cost model would be different. I would expect that at present most users are relatively occasional - not, for example, going to work in one each day (for which you'd have to guarantee availability of one). So a lot of people (think gyms) will be paying £50 a year but making little use of the cars.

And with that comes one of the railway's biggest costs - rolling stock that is in use between 6 and 9am, and 4 and 7pm, each day, and spends the rest of the day sat in the sidings. These cars would be no different unless you institute sharing, which starts tending towards public transport anyway.

Better, is it not, to have people *not* commuting?

Neil

In reply to andrewmcleod:


But the areas in which they operate have been specially mapped to a level of detail not available in most areas here. I have reasonably high spec sat nav in my car but it still manages to panic that it is 'off road' fairly often and there areas where the satellite signal is lost.

I guess I do have larger reservations about having basic functions like communications and travel increasingly dependent on external technology over which we have no control. It is a rather urban Californian attitude that simply can't believe there are still third world areas like Europe where there is no mobile phone signal.

I'm not saying that driverless cars won't work most of the time in many places. It's just that I'd guess that many us spend more time than the average person in exactly the sorts of places they won't work. And working most of the time isn't really good enough in this case.

In reply to Dave Garnett:



Absolutely not. Firstly comparing a consumer PC is a bit of straw man.

Secondly, so what if GPS goes down? The car isn't instantly blind. Unlike the numpties you read about who blindly follow their TomToms into fields and rivers, the cars that are in testing now use LIDAR and other imaging technogies to build up a 3D model of the world around them. They can see more than humans ever possibly can, and more importantly they can pay attention to everything, and navigate simply through it. They don't just go point a-b-c.

OK it might get lost without GPS as anyone would but it would happily drive around, perhaps asking for input on direction from the passengers, using stored memorised routes, reading road signs, or syncing with roadside transponders or other networked vehicles.

I'm not blindly putting my faith in technology - this needs to be fully proven - but we're a long way there and so far things look very very promising based on how soon this has all developed.

In reply to Neil Williams:


Then again, that cost is there for commuting *right now* - outside my office window are 65 cars that will burn through capital for 7.5 hours and not move. Car sharing can only help with that?

The commute "rush hour" (3 hours) is spread out over ~8x the duration of the average commutes so there is scope for self-driving cars to share the capital cost out over several commute, and then provide much of the non-commute mileage during the day.


Absolutely. How many jobs really worth doing can be done in isolation at home though? Perhaps that's a provocative statement, but I have a feeling that the jobs that could be made commute free largely have been, already, and that there's not much to gain there.

As the step change in roads that comes with half term week reminds me, the biggest gain would be in running sufficient school busses and making parents accept them over driving. Better yet, reverse the centralisation of schools.

Although plenty of miles have nothing to do with commuting, and that won't have the same time multiplexing problems as commuting.

The new models of ownership don't have to work for everyone, all the time.

If I could have a ZipCar style car for occasional use, I could then live with a simpler/cheaper/smaller 1-seater for my commute such as the perhaps-yet-to-be Lit Motors 2-wheel CMG car. The total cost of ownership/use to me from that should be lower than running one car that's sized based on the occasional/extremal needs then using it for a daily one person commute (actually I don't use it for that, it sits there burning capital whilst I use my legs or lift share...)

In reply to yorkshireman:


Quite. The tomahawk cruise missile has been navigating quite happily of terrain matching imaging systems (the images happen to come form RADAR) for 30+ years now. A system is perfectly capable of navigating by a combination of dead reckoning, map reading and imaging, with the benefit over humans that it can incorporate terabytes of georeferenced imagery as well as performing image analysis.

The "GPS going down" merits a bit more thought - we've got NAVSTAR (USA) and GLONASS (Soviet/Russian) both giving global coverage right now (my iPhone receives both), and Galileo (EU) being rolled out world wide. There are other currently localised systems such as BeiDou (China).

If all these independent systems all go down over an area, you've got a serious problem on your hands such as coming under direct military attack or a solar flare serious enough to cause major terrestrial consequences.

In reply to FesteringSore:

Really into cars. Don't own one because the cost/ ease / benefit of owning one in a large city is outweighed by the excellent public transport system vs money pit. I'm old enough not to buy into the independent, rugged symbol of my masculinity bollocks that's peddled heavilily in the advertising of the things. Self driving cars, boring, never be as be as bright as the driver, insurance companies and government taking it up the shitter from lack of kick backs from their protection rackets. Brilliant, what's not to like. If I can jump in the back and wake up in Wester Ross ten hours later I'll be first in the queue.

D

In reply to wintertree:

I suspect that the average TomTom is designed with trade-offs between capability and cost. Given the second requirement, its no doubt dumbed down as a system and therefore expected to have the odd seizure. Afterall, its not intended to "drive" the car - just aid the driver.

Would seem entirely likely the full-size GPS on a driverless car will have capabilities well beyond that of current cheap-o GPS.

Seems a lot of the naysayers are also overlooking how recent road mapping GPS is. The flaws inherent in such a system under two decades old are fast being ironed out.

In reply to FesteringSore:
We are a very long way from mass production driverless cars that will have no means for the driver to intervene.

Until that time, all the systems fall under the umbrella "Driver Assistance Systems". They do just that, assist the driver, they don't replace him/her.

Cars already have knowledge of the speed limits, and features exist that make the car coast down pre-emptively in anticipation of the new limit, but the driver still has an accelerator and a brake, and is still 'in control', s/he is still required to be aware of his/her surroundings and situation.

I think we are decades away from production cars that the driver cannot control manually (if it will come at all), we are probably 5-10 years away from production cars that can automatically take you from A to B without driver intervention, although the driver will always be responsible for any mishaps.

There are already systems that will handle the driving on the motorway (Google 'Highway Pilot'). It's all very cool, and wonderful to drive.

Source - Driver assistance systems design engineer.

In reply to JimboWizbo:

Personally I'd say the most likely first implementation is an automatic car lane on a motorway, maybe two lanes once it becomes popular. That would be relatively easy to implement and could work in conjunction with managed motorways.

Neil