"that buyers of driverless cars"
So where can I buy one then?
These people are in cloud cuckoo land.
I predict in 20 years time you will still not be able to buy an autonomous vehicle for general road use.
A UK minister has reassured the motor industry that buyers of driverless cars will not have to employ a gentleman with a red flag and top hat as he outlined a UK testbed for autonomous vehicles yesterday. Speaking at the SMTT Connected conference in London, industry secretary Greg Clark outlined a “cluster of excellence” …
"I predict in 20 years time you will still not be able to buy an autonomous vehicle for general road use."
Good. Can you imagine how many idiots who can't be arsed to learn to drive decide to buy a driverless car to take them to Tesco?
Hellooooooooooooooooo gridlock.
I suspect most people who don't have a car do so because they either can't afford one (so they wouldn't be able to afford a driverless car either) or they can't be bothered driving where they live (e.g. London, where you can get by without a car fairly easily). Add in the fact that driverless cars should ease gridlock by driving better (less having to hammer on brakes, smoother drives, no slowing down to rubber-neck, etc), I don't think it's as bad as you might be worried about.
Of course, the roads are getting busier year on year, so in 20 years time, who knows what mess they'll be in?
Of course, the roads are getting busier year on year, so in 20 years time, who knows what mess they'll be in?
But all the migrants will have left by then, so the traffic will be flowing freely once more like on a 1950s motorway film. At least that's what my mate Nigel down the pub says.
"And have to use their SatNav to get to Tesco even though it is 400yds down the road and can be clearly seen before they start."
Last year I had the dubious pleasure of travelling with what I loosely call a colleague[1] who drove entirely by SatNav, deliberately ignoring road signs which were quite clear.
He totally missed some roadworks diversion signs and was completely baffled when the road turned into a building site.
[1] He was from HR and a definite candidate for Ark B.
>but 20 years is a very long time.
Yeah, it's more than 20 years since we had a proper transport policy.
And my current 1999 Peugeot 306 diesel does exactly the same as most of today's cars: same pollution, same auto windscreen wipers, same stork radio controls, air cooling system, rear window heater and 750 miles out of a tank of fuel. It hasn't got cruise control, satnav (I use phone instead) or mp3 (upgraded years ago) or a dual mass flywheel but it's done me for the best part of 20 years.
Reliability isn't something you'll be associating with safety critical autonomous car systems any time soon - evey sensor failure will stop you using it.
Also there is concern that autopilots on ships and aircraft result in LESS ability [for the human] to deal with the unexpected.
There is talk of changing the design of aircraft autopilots so that that the human is more involved and thus able to correctly assess what to do (c.f. Air France over the Atlantic, the human intervention was the opposite of what was needed).
@Mage
There is talk of changing the design of aircraft autopilots so that that the human is more involved and thus able to correctly assess what to do (c.f. Air France over the Atlantic, the human intervention was the opposite of what was needed).
Indeed, it's already happened. The A350 apparently 'makes' the pilots do more flying themselves, though this is as much about mandating more pilot hand flying time in the operations manual (which airlines have to follow to be licensed) as any technical changes to the autopilot itself.
The same approach won't work with cars; people's driving isn't logged, monitored and regulated like a pilot's flying, and introducing such oversight isn't going to be an option. So I think with driving it's an all or nothing situation. Either we do all the driving ourselves (adaptive cruise control is allowed), or the car automation is perfect and does it all the time.
The difficulty for the self drive industry is as follows. A self driving system that is nearly perfect is more dangerous (in the long run) than one that is rubbish.
In the UK, roads are designed with driver psychology taken into account. We don't build straight roads anymore to stop drivers getting too bored.
> "or the car automation is perfect"
> "A self driving system that is nearly perfect is more dangerous (in the long run) than one that is rubbish."
I don't agree. Car automation just has to be better than humans for it to be worthwhile.
Currently 10 people a day die on the roads of the UK. If 100% car automation is only twice as safe and so kills 5 people a day, we'll be saving 1,800 lives a year.
There's no such thing as "perfect" and to put any technology on a pedestal of expected perfection is to doom it to failure from the outset.
How about a better, faster, cheaper public transport system (most of us not in London have to endure dismal public transport ).
Less cars on the road / less car journeys = more lives saved with no magic technology fairydust required.
This post may be partially related to me just checking the price of some train tickets for UK travel and going "How ******* much?!?!, you must be ******* joking"
More and bigger public transport won't work. You can't get a bus down my lane and it's too far to walk to the nearest bus stop even without shopping so I just get in the car.
A fleet of publicly owned driverless cars may work if there are enough for them to be treated as cheap taxis and they may be better for the environment if there are so many that there is always a free one nearby (that's why taxis are so bad - they spend half their journeys with just the driver i.e. zero passengers).
@ Natalie Gritpants
How far away is bus stop for it to be regarded as out of walking distance?
Just out of interest as depending on what distance people regard as too far could make a difference to what % of people would be regarded as too far from public transport.
Ironically, on the walking topic, it makes more sense for me to walk home from my nearest train station than take a bus home.
Buses and trains both only run hourly, bus stop a couple of minutes walk from station, however bus leaves approx 5 minutes before train arrives in a monumentally stupid piece of timetabling.
How far?
In London it seems that they stop every 100m. They were not always that close. In one stretch of the old 253 Route (Warren St to Stamfrd Hill and beyond) where there was two stops there are now four. Two extra ones have been added in between the others. The same goes elswhere. IT takes even longer to get anywhere in London if you go by Bus than it did 50 years ago and not all of that is down to traffic as back then there were no bus lanes.
How far away is bus stop for it to be regarded as out of walking distance?
Any distance at all, IMHO. If I wanted to share a dirty enclosed space with the fat, the smelly, the weird, the anti-social, the deranged, and assorted ratboy vermin from social class F then I'd happily hop on the slow, smelly, uncomfortable bus. And funny thing is, there's only ever two types of bus I see - the almost entirely empty ones for most of the day, and then the hideously overcrowded ones at peak hours.
And that's the real problem with any form of public transport - not poor planning, but poor asset utilisation because traffic flows are not homogenous, and the fact that people don't always want to go to the same places at the same time. My morning commute is 45 minutes by car. Fastest public transport is 2 hours 4 minutes using bus and train, and 2 hours 45 using just bus, and that's using relatively well interconnected services around Birmingham.
That's me told, I'm social class F then! You missed out the obligatory inbetweeners "bus w****r" though ;-)
I like buses e.g. otherwise I would have a very long and erratic walk back from the (4 miles away) pub - would not fancy driving in that beer enhanced (FSVO enhanced!) state
"How far away is bus stop for it to be regarded as out of walking distance?"
That depends on age, physical fitness and how heavy (and how many) bags of shopping to be transported.
Coat. It looks like rain and the bus stop is nearly half a mile away. Uphill. In the snow. Both ways. And I know I'm lucky to have that.
If you want to properly use technology to save lives and ease congestion in the cities, why not spend the money on a really good broadband network and get rid of the idiotic notion that communication must be face-to-face? Some jobs will always require a physical body in the room, but I'd suspect that at least half could just telecommute. The numbers will level off the more rural you get, of course, but that countryside isn't the part where gridlock is the problem.
The other difference is that a really widespread and robust broadband infrastructure could be used for other things as well. Driverless cars, less so.
I'm not going to even begin to point out the obvious things, like allow the cars to travel at reasonable speeds and widening roads. Those are quite clearly madness to a politician.
> bus leaves approx 5 minutes before train arrives in a monumentally stupid piece of timetabling
Alternatively, the bus arrives early enough to drop off people who want to catch the train, in a rare piece of timetabling synchronisation (Assuming 3 minutes is enough time to buy a ticket)
"You can't get a bus down my lane".
If there's full autonomy you won't have large buses. It'd likely be built around passenger vans.
In places with cheap drivers you have jitneys and dolmuses. If drivers are expensive larger buses are more economical. With (affordable) autonomy the model is again based on cheap drivers. That allows for more, smaller vehicles giving more flexible, faster public transportation, and can allow lower density areas to be served at reasonable cost.
A fleet of electric autonomous vehicles for hire would be a public transport system. Just nitpicking.
It's said that most rural buses only have a handful of passengers most of the time but need more capacity when they get into towns. My own experience bears this out. So electric taxis could be more efficient than buses in terms of capacity and energy consumption. I say could be. Prediction is always risky, especially about the future.
How about a better, faster, cheaper public transport system
Well, of course, autonomous rail systems have been around for quite a while now. There are even some in London, but they all have staff on board because apparently passengers don't feel safe in the hands of a robot, despite the fact that the signalling systems are designed to be failsafe and communicate unambigously from machine to machine.
I still don't understand why passengers are supposed to have more confidence is a small autonomous vehicle that is trying to manoeuvre amongst lots of other small vehicles, most of which are not autonomous, essentially trying to guess what those other vehicles might be doing and attempting to infer signalling information from lights and signs designed for human drivers.
If you're intending to move towards ubiquitous autonomous transport, then you should really skip the step of mixing self-driving cars with those carrying meatsack pilots.
Yeah, strange that London, that hotbed of UK capitalism, was the one place allowed to keep a regulated public transport system while the rest of us have to put up with a non-regulated mess. Another example of how privatisation only works if you regulate the resulting market to avoid the (natural) tendency to form monopolies, just like we do in other markets.
London, that hotbed of UK capitalism, was the one place allowed to keep a regulated public transport system
As a regular user of London's public transport, I'd like to voice the opinion that it is shit. Totally and utterly shit. Painfully, slow, overcrowded, uncomfortable, and at times unsafe. It is only the considerable effort devoted by the public sector to screwing over car-based transport that make public transport not an option of choice, but an evil necessity.
"How about a better, faster, cheaper public transport system (most of us not in London have to endure dismal public transport )."
How about reversing the aggregation of work places into bigger and bigger urban centres? The people wouldn't have to commute stupid distances to work.
"government will direct its support in its efforts to establish the UK as a leader in autonomous vehicles"
Seriously? Where are we at the moment? Google, Apple, Tesla are all years ahead. Have we got even a single manufacturer working on this - and I mean under UK ownership?
This is another frog puffing out its chest to make itself appear bigger and grander than it is. But it's still a frog, and the UK is still an insignificant dot on the map.
Meaningless political soundbite with little actual intention to provide serious funding or support.
Google, Apple, Tesla are all years ahead. Have we got even a single manufacturer working on this - and I mean under UK ownership?
Fair point. But I'll bet you that none of them has thought about producing an autonomous car that drives on the left.
And before you shoot me down, I'll remind you that Google Glass was available only for the right eye.
the UK is still an insignificant dot on the map
Here we go again. The UK is the world's largest economies, one of the most significant and respected trading nations on the planet. Our motor industry turns over about £70bn a year, mass makers like Toyota, Honda, Nissan, Ford, BMW et al have large facitilies here, we're a world leader in premium cars with manufacturing and design for Aston Martin, Jaguar Land Rover, Rolls Royce, Bentley, there's 6 design centres, 13 R&D centres, we're the base of several world leading vehicle engineering specialists.
Have we got even a single manufacturer working on this - and I mean under UK ownership?
Domicile of the owning entity is irrelevant. We tried "owning" our motor industry by nationalising it. That went well, didn't it? Now look at JLR flourishing as part of the Tata Group. Why is Nissan's European Design Centre in London?
The biggest problem autonomous vehicles will have is non-autonomous vehicles.
No matter how complex the electronics, a person signalling right and turning left, or going straight on, will always mess it up!
Simply: autonomous cars will easily manage if they're on a road with similar vehicles.
Put one car driven by a person in the mix, and anything will happen.
And if every vehicle is driverless, there will be no reason to own a vehicle. Just call one from the idle-auto-car carpark.
"No matter how complex the electronics, a person signalling right and turning left, or going straight on, will always mess it up!"
They mess it up for the other meatsacks too - and apart from robots having faster reaction times than squishies, that's the kind of thing that onboard video will capture and make abundantly clear who's at fault.
Up to the advent of ubiquitous dashcams, this kind of dangerous driver usually got away with it when they crashed by blaming the other car. With the advent of robodrivers, it may be possible for repeat offenders to start receiving "invitations" to resit their driving tests.
Waste of taxpayers money and risk to public.
The real car makers have realistic "testbeds".
Note that Google, Apple, Uber, Intel (who just bought Telsa's provider) etc are not real car makers. Tesla is a niche market subsidised by the taxpayer, yet only affordable by the rich. His "autopilot" is far away from being a real autonomous car.
"Human error is a more dependable source of accidents and fatalities than well tested, well demonstrated, and well regulated technologies,” Clarke said.
The perfect cue for someone to have arranged for a driver-less car to come crashing through the wall for shits and giggles.
"I look forward to seeing self-driving cars try to negotiate the Handy Cross roundabout (M40 junction 4) during rush hour."
Surely Hanger Lane Gyratory would be an even better test area? From Wikipedia: "In December 2007 it was named Britain's scariest junction."
I haven't told my wife this, as I use navigating it as a test of whether I've still got all my marbles.
Scariest Junction?
Surely it has to be that one in Swindon where it seems that anything goes?
IMHO and as a Motocycle rider, Handy Cross is more scary than Hanger Lane. The reason is the GP start to try to beat the lights going red in front of you. (M40South and turning right towards Maindenhead/Marlow)
a “cluster of excellence” stretching from London to, er, Birmingham into which the government will direct its support in its efforts to establish the UK as a leader in autonomous vehicles.
Isn't the government also supporting a high speed rail link there, thus encouraging people to bypass the testbed area by taking the train?
I've been wondering when widespread testing would start in the UK.
It's one thing testing on American roads, nice and wide, mostly built using a grid system. Testing in the UK where a good number of our roads follow old Roman foot paths is something altogether different.
I for one can't wait to see how they deal with passing places and national speed limit b roads.
Not to mention the mini-roundabouts - and definitely don't mention the double mini-roundabouts.
At the top of my lane there's one of those y junctions onto a b-road - technically you are supposed to enter and exit on the left. Practically - you use right side to go right or you can't see the oncoming traffic and need to do a three point turn onto the national speed limit b-road. We have got plenty of roads with blind turnouts where you need to sight the traffic disappearing into a dip to calculate the turnout. We've also got loads of single tracks with very few (and variable quality) passing places where if you get it wrong you need to go and find a friendly farmer.
They are, however, welcome to the M40. Used to be a regular on that road and it always seemed to me that most of the vehicles would have been better off without whatever maniac was driving them.
Clark, in common with other speakers at the conference yesterday., emphasised the safety and societal benefits of driverless technology. Road fatalities in the UK are now 3,600 per year, compared to 6,400 in 1975. Given that 95 per cent of crashes involve human error, taking the human out of the equation will result in a massive drop in road deaths, the reasoning goes.
There are quite a few more drivers on the more congested roads now than there were in 1975 and road building has not kept up with demand, but even against all those odds, the figure is 50% less than back then. The vastly improved safety of modern cars also has a lot to do with things as we've got better at making them.
The bit that the stats glosses over is that "95% of crashes involve human error" - please remind me again who wrote the software for the automated death machines ?
I'm also prepared to be that the new "funky bubble" cars that keep popping up are a lot less safe in a crash than a modern hatchback or SUV.
I've been in charge of my body since a couple of years after I was born and if something is going to happen to me, then I want to be as much in control as possible. I don't want to become a statistic against bug number 3412-432, so don't expect to ever sell me one of these pieces of junk.
"The bit that the stats glosses over is that "95% of crashes involve human error" - please remind me again who wrote the software for the automated death machines ?"
1: Humans tend to make a serious driving error every few minutes.
2: Robots tend not to fixate on the rear of the girl on the footpath and miss the brakelights in front of them, etc.
3: Humans need individual training - and the results of that training are wildly variable
4: When a change needs to be rolled out, all those humans need individual retraining.
Road crashes tend to involve at least 2 glaringly serious errors combined.
Even the vastly imperfect Tesla autopilot has achieved a 40% reduction in crash rates over what would be statistically expected from 100% human piloted cars - and that's according to the NHTA, which investigated the things after Joshua Brown's death(*) in the expectation of finding they were dangerous/should be banned.
(*) There are still far too many unanswered questions about that crash, including "Where's Joshua's dashcam? He never travelled without it"
The government could invite motor manufacturers to provide autonomous (test) vehicles to replace the fleet for of chauffeured vehicles for Ministers.
The government saves money, and if the ministers are still in one piece after a given time, the motor industry has a group of influential supporters in government for the technology.
Highway Code
Rule 268. Do not overtake on the left or move to a lane on your left to overtake. In congested conditions, where adjacent lanes of traffic are moving at similar speeds, traffic in left-hand lanes may sometimes be moving faster than traffic to the right.
The magic words are "Do not" which means the rule is advisory rather than "Must/Must not" which is used to state that the rule is law. It is not illegal to pass on the inside lane, just take care.
" It is not illegal to pass on the inside lane, just take care."
Correct.
Beware though: If you pass on the left and the passee moves left on you, you're likely to find yourself facing careless driving charges - even if they're doing 35mph in the right lane and listening to Charlie and the Cockroaches at full volume, so not hearing the horns behind them.
Last trip I took up the M1, I counted 80 instances of laneblocking between the M25 and M6. It's pretty clear that the law change isn't helping much. I also saw a large number of potentially dangerous passing moves to get past the idiots causing the slowdowns so cracking down on the former has potential to cause significant reduction in the latter. (The problem at the moment is that only patrolling police can issue a ticket, only on the spot and only if they witness it occurring for some time beforehand.)