UKC

The future has already arrived.....

In reply to The Ice Doctor:

Its going to be worse than following a Rover with a Panama hat on the parcel shelf and a fish sticker on the back.

In reply to birdie num num:


Haha.

In reply to deacondeacon:
I'd love to hear the cynics and realist's pull apart the positive attributes communicated by the production team on the dream (that is being sold), that may well follow....

Hilarious to think that humans may be working towards making themselves redundant. We won't have any purpose.lol

The programmes punchline- We are freeing up time to work even more! Good grief! Blissful.

In reply to The Ice Doctor:


Works for me, as a service engineer I can spend 4 hours a day moving between sites, it would be very handy to be able to crack on with email's, reports and quotes etc.
The down side is I get travel sick if I'm not the one driving.

In reply to Dax H:



Won't the car be sick for you?

In reply to The Ice Doctor:

Be great to see driverless cars obeying speed limits and in most cases staying well below them. Slowing down or stopping when a cycle or pedestrian is ahead then changing lanes to overtake whilst other driverless cars stop to allow the manueavere

In reply to The Ice Doctor:


Work has only ever been a necessity for our survival, not our purpose.

Humans and their ancestors have been working towards eliminating the necessity to work ever since the first use of tools over 2,500,000 years ago.

Why do you find it "hilarious" that we continue to allow more to be achieved with less effort, freeing up human time for both more enjoyment of life and to accomplish even more profound tasks with that life?

Do you really think it's a net benefit to the species for ten million man hours a day to be spent piloting metal boxes around roads in the UK alone?

In reply to Wintertree : (whispers) I detect sarcasm

Aren't they testing driverless in the UK now? This is coming so fast . I think it will be great. Program in that far away destination and go to sleep

In reply to The Ice Doctor:
Reduced cost for freight so more lorries as they become economic for more businesses.
Taxis replaced by driverless uber.
White van deliveries - pick up lockers on a driverless van.

You can afford the technology if it saves the cost of paying a driver

All sorts of additional car journeys become possible for children, non-drivers, disabled, city centre workers who don't have parking or sleep as you travel overnight.

For irregular domestic use you can'take justify cost of technology so maybe car hire or car club so the vehicle is used more intensively.

Alternatively it may be an expensive extra for top end cars that rapidly gets cheaper.

The technology will never be perfect but it is already bether than humans.

In reply to wintertree:

Hilarious, in that the promise of more time may never actually be deliverable, will people really enjoy themselves more? besides, its tongue in cheek sarcasm, not transmitted through the written word.

In reply to The Ice Doctor:

Could spell the end of the personal car. That can only be a good thing.

In reply to Lion Bakes:

Could spell the end of the personal car. That can only be a good thing.

Not to the people that actually enjoy driving and where driving somewhere is all part of the experience.

In reply to FactorXXX:



Or for climbers whose car is a scummy hole where they store kit and doss.

In reply to FactorXXX:



Well enjoy that whilst you can as it's coming to a necessary end

In reply to Lion Bakes:


I simply cannot see people accepting it. I expect there will have to be some sort of system where non personal driverless cars are used for routine journeys such as commutes and lorry tansport etc. but personal cars being used for less mainstream stuff like climbers going somewhere to doss in laybys and so on.

In reply to Robert Durran: that's special case pleading isn't it ? . Really, I don't think there's much accepting to be done - they will simply grow in number as people rapidly accept that there daily commute isn't really the advertised vision of driving. Think of the growth of online shopping as an analogy.

In reply to Dax H:



The Downside is that you or some colleagues will be made redundant due to the productivity increase.

With increased automation we need a plan on how to adjust work load expectations. Reducing typical hours, increasing basic pay to balance the economy. No point putting in robots, making half the work force unemployed and increasing the social security bill. Better to adjust societies expectations - our current 37 hours / 5 days mentality is arbitrary.

In reply to The Ice Doctor:



To be honest I'd love to have less purpose!

In reply to The Ice Doctor:


Oh, no! If we lose our purpose we'll have to invent God again!

In reply to birdie num num:


I had a real "the car in front is a Toyota" journey home from Swindon yesterday. Every time the traffic ground down to 30mph you could guarantee that the car in front was, indeed, a Toyota. So I think your reference to Rovers is a bit dated.

In reply to The Ice Doctor:

I read somewhere that cyclists are the biggest stumbling block to driverless cars in towns, as they are too unpredictable, so the cars just default to stopping.
Bring it on I say!

In reply to Philip:

We've been here before though with factory automation and, as a society, we had a choice to work less or buy more shite in order to support the same amount of jobs. I wish I'd been around to be consulted.

In reply to The Ice Doctor:

Will driverless cars be able to observe good single track road etiquette. Even the Loch Hope road which doesn't even have passing places and you have to assess the bogginess of the verge?

In reply to Dax H:



If I am concentrating ( reading etc) I too get car sick, however, I found out by accident on a long journey in a black cab one time that if I was facing backwards to the direction of travel that the sickness was minimised, and bearable.

Maybe a driverless car with the seat facing backwards will work for you

In reply to Andy Hardy:

Segregation on the Dutch model will probably be the way to handle that.

In reply to wbo:


I can see driverless cars being technically feasible in the near future, but how you integrate them into existing road conditions is the issue.

It won't take long for pedestrians, cyclists and drivers of non-driverless cars to realise they can simply intimidate a driverless car into automatically stopping/swerving/getting out of the f***ing way to avoid a collision...

In reply to Ridge:


Although those cars will be logging 360 video, radar and sonar of every act of bad/dangerous driving. Why wouldn't the firms running fleets or cars not want to use that data to reduce their insurance liability and to improve the user experience of their products? The question unanswered for me is if the data can be certified to a sufficient level for it to be advisable as prosecutorial evidence.

Self driving cars are going to lead to a dramatic change in what is and is not acceptable from all road users, both through data logging and because it will no longer be necessary for driving to be treated as such a necessary right.

In reply to The Ice Doctor:

This is fun

http://moralmachine.mit.edu

In reply to Jim C:



The entire design of cars will be radically changed. Why even have a front/back? Why not have a coffee table in the middle with all seats facing inwards?

THere's a lot of amateur whatiffery as usual in these threads and the bottom line is we don't know how this will play out. When mobile phones first appeared they were sneered at and people never really thought about smartphones (Just look at any sci-fi film pre-1998 and even the most futuristic visions* failed to predict universal, wireless, mobile network connections and the impact it would have on our lives) .

And yes I'm wondering how the 4am drive I had to make out of our village in late Spring for 5km to the main road, in an overnight blizzard that left 30cm of snow before the plough had come through - I couldn't see the edge of the road but I knew where it was - maybe an autonomous vehicle would have given up, or have been much better than me.

But these are edge cases. 99.9% of all driving is predictable, on well maintained busy roads. The opportunities for improving our life drastically are huge. Less pollution, more free time, cheaper journeys - even less space needed for roads in the first place as there will be fewer vehicles (being reused rather than owned) and will rarely need to be parked (again can be reused, or can be parked elsewhere if necessary), and a massive boost to our economy and well being.

Forget Brexit - what can the NHS do with the billions it will save from not having to mop up as many road accidents in the future?

*mostly

In reply to wintertree:




For me it opens up all sorts of questions about our purpose and direction as a society, and I don't think we've even begun to grasp what it could mean.

Like it or not many or even most people are largely defined by their job (perhaps less so on this forum than in general), and what work they do, regardless of how productive or ultimately useful it is. If much of that disappears in a short span of time with little to replace it, it will massively impact society.

It strikes me it could be like the collapse of industry in the UK (i.e. 1980 coal mining etc) but on wider and more permanent basis. Couple this with growing inequality and upwards wealth migration that is already well progressed, then it seems to me you have a potential recipe for disaster.

If we're really taking about a massive and fundamental shift to a work less society (as often seems to be forecast), then society needs time to adapt - i.e. universal incomes, life long education, better health care, etc.

In reply to Robert Durran:


Easy peasy. The cars communicate, everyone goes through with optimal timing.

The radar they use for avoiding people sheep and so on - sees thro' snow, fog, heavy rain , snow on the ground and will differentiate tarmac from mud. You can sit in the car with a pot of tea and be transported to the grubby quarry of your sport climbing vchoice

In reply to The Ice Doctor:


Eh, no. I've just watched the programme and I am, I think, reassured that I'm very unlikely to be forced to go fully level 5 driverless in my lifetime. I'll still be able to potter along and pull onto the verge for a piss or to take a photo when I feel like it. The future is safely in the future. I can't see much more than taxis, delivery vehicles and limited personal journeys being driverless in the foreseeable future.

In reply to wbo:

Easy peasy. The cars communicate, everyone goes through with optimal timing.

I'm envisaging a scenario, where two cars are stuck for hours in the same place arguing over who has right of way...

In reply to FactorXXX:



youtube.com/watch?v=DtmLJnTRL64&

In reply to Ridge:


In other words they can go about their daily business without worrying some numpty at the wheel who is speeding, fiddling with a phone or radio and isn't paying attention. Driverless cars will be paying attention all the time and staying below speed limits at all times. In fact I don't know why they are not fitting the speed limit obeying software in all new cars anyway.

In reply to Robert Durran:

You may actually not have a choice. That's the scary bit!

In reply to Lion Bakes:


What's sauce for the goose is sauce for the gander.

I look forward to seeing riderless bikes

In reply to wintertree:



What a grim and soul less, big brother style vision

In reply to timjones:


It's hard for me to feel negative when it means 2,000 people a year less being killed on the UK roads and many times more not being injured.

Over a million lives a year saved globally.

In reply to The Ice Doctor:

Fun!
https://www.theatlantic.com/technology/archive/2017/06/how-do-buddhist-monk...

In reply to wintertree:
It seems a little strange that a climber is risk averse enough to surrender the joy of driving?

In reply to wintertree:

It's hard for me to feel negative when it means 2,000 people a year less being killed on the UK roads and many times more not being injured.

Will driverless cars mean zero accidents then?

In reply to The Ice Doctor:


Well, there will be uses for autonomous vehicles but they don't address issues of pollution, congestion, personal fitness, and, they potentially remove the pleasure of 'piloting' transport from A to B.

The future in cities is bikes - eBikes and cargo bikes. Autonomous vehicles will have a supplementary role, but cars are basically the wrong choice for the urban environment.

In reply to Shani:


They may be wrong if your journey starts and finishes in the same city but not all journeys are so simple.

In reply to Shani:


The Horizon programme seemed to imply that they would be electric.

In reply to timjones:


Interesting. I wonder whether climbers are actually often relatively risk averse in their non-climbing lives. Come to think of it, climbing is actually significantly about risk management rather than risk taking so one might expect them to be at least risk aware in other aspects of life.

In reply to timjones:


...in which case the journey is extra-urban. Not urban.

In reply to Robert Durran:


Hopefully they'll be electric - just leaving the issues of congestion and an increasingly obese population to resolve.

In reply to Robert Durran:

I'd say that risk aware is very different to risk averse.

For me both climoving and driving are pleasurable freedoms, I don't want to lose either of them. I'm aware of the risks in both activities and believe in minimising that risk but you can definitely go to far and lose the spirit of an activity.

In reply to Robert Durran:

Interesting. I wonder whether climbers are actually often relatively risk averse in their non-climbing lives. Come to think of it, climbing is actually significantly about risk management rather than risk taking so one might expect them to be at least risk aware in other aspects of life.

I think the risk management undertaken by climbers is relative and unique to the activity that they are doing. The obvious risk management would be not to climb at all and that's why most sensible people don't climb.
If you do apply risk management to climbing, then some will only Top Rope and some will do a Honnold and solo El Cap.
There's a long history of climbers doing mad cap things on motorbikes and in cars, and it seems a lot of the bolder ones are the maddest of the lot...

In reply to Phil79:



While to some extent we all benefit from new technologies they generally start off as a way for someone (an individual) to increase their means of production. It's all back to Marx really - those who own and control the means of production and all that. Clearly, in one sense, there's already far less need for 'work' but most of us do it because we need an income and we have no control over the means of production.

The shift to universal incomes will need governments to be far more interventionist than they currently are in the range of neoliberal democracies that Westerners inhabit. And really, of course, in the globalising world we currently have, it's hard to see the planet developing the supranational bodies that are actually necessary.

Rich get richer, poor stay relatively poor, even if they don't have to drive their own car around. Back on topic, the implications some have referred to above in terms of more / fewer journeys are interesting. Maybe it will be possible to make travel more energy efficient but it seems unlikely - most car drivers can't stand the public bit of public transport, or is that just me?!

In reply to Shani:


A cull?

In reply to FactorXXX:


I didn't say that, did I? Not even close. Zero accidents sounds statistically highly unlikely.

I do think they have the capability to cut the number of accidents by an order of magnitude as well as reducing the severity of those that still happen.

In reply to timjones:


Where did I say that? The fact I enjoy risk doesn't mean I want over a *million* people killed by accidents largely due to bad decisions a year.

Especially when driving is far less optional for many people than climbing, and many of those killed are killed by other people's bad judgement with regards risk. I'm including pedestrians and cyclists in that category.

In reply to wintertree:

Then why not work on improving peoples judgenet rather than removing their freedom to drive?

In reply to Shani:

So would you still allow extra urban car journeys into cities?

In reply to timjones:


I'm not in favour of banning cars in cities, period.

In reply to wintertree:

I didn't say that, did I? Not even close. Zero accidents sounds statistically highly unlikely.

The current fatality rate on the roads is under 2000 and you said driverless cars would mean '2,000 people a year less being killed on the UK roads'.

In reply to timjones:


I never advocated removing freedom to drive. I never said I think that will happen. I said bad driving is going to be less acceptable when there is more data logging and driving is not so important.

So I am saying people who chose not to improve their judgement or who are incapable of improving it will be less likely to get away with it.

So I believe we are in agreement.

In reply to FactorXXX:



My most humble appologies for not stating that I was rounding to within approximately ten percent. If I understand your logic correctly you should have been asking me if I thought driverless cars would bring about 200 people back to life a year, not cause zero accidents. Not that I ever claimed all accidents would be eliminated either.

In reply to FactorXXX:



Not just fatalities though is it? There were around 186,000 people hit by people driving cars in 2015. That's far far too high.

In reply to wintertree:

My most humble appologies for not stating that I was rounding to within approximately ten percent.

It still means you stated that there will be zero deaths.

In reply to Robert Durran/Tim Jones - ITS got nothing to do with being risk averse -it's a lot to do with being averse to boring driving when I concentrate on something more interesting

In reply to Lion Bakes:

Not just fatalities though is it? There were around 186,000 people hit by people driving cars in 2015. That's far far too high.

Driverless cars may or may not reduce collisions, but it won't mean zero deaths as implied by wintertree.

In reply to FactorXXX:



Please show me where I stated this. I did not. If you wish to construe my statement in a way that reads like this despite my clarification then I can do little but suggest you are acting without honour or integrity. As well as bad maths.

Oh dear. I just simply don't know how to live with myself given this massive misunderstanding that does not change the crux of my argument one iota. I guffed up using a round number and not saying "approx 2000".

If you genuinely want to claim that is how you interpreted my message as saying precisely zero deaths then you appear to be very bad at maths because any recent year's death rate minus 2000 does not equal precisely zero.

You also appear to conflate deaths with accidents. I have apologised for my guff but don't let that stop you posting daft stuff back at me.

In reply to wintertree:

Oh dear. I just simply don't know how to live with myself given this massive misunderstanding that does not change the crux of my argument one iota. I guffed up using a round number and not saying "approx 2000".

Driver car deaths is approximately 2000.
Driverless cars will reduce deaths by approximately 2000.
Therefore, driverless car deaths is approximately zero.

If you're saying that driverless car deaths will be approximately zero, then I can't argue with your maths. Is that what you are saying?

In reply to FactorXXX:





Somewhat. Replace "is approximatly zero" with
"Will be orders of magnitude lower than now" then yes. Close to zero I hope one day.

Also I thought my use of an approximate number was bloomin obvious from my post at 20:05...

In reply to wintertree:

Somewhat. Replace "is approximatly zero" with
"Will be orders of magnitude lower than now" then yes. Close to zero I hope one day.

Why didn't you say that in your initial statement? It's a worthy goal and it would have saved a lot of confusion brought about by you saying the amount of lives saved essentially equalled those lost.


Also I thought this was bloomin obvious from my post at 20:05

Not really, you just carried on insisting that your flawed logic and maths was correct, despite evidence to the contrary...

In reply to FactorXXX:


Why didn't you state your "logic" (or broken maths as I see it) in interpreting my post to mean zero accidents, which I take to be a preposterous extrapolation from my statement regardless of weather I gave a precise or approximate (but not labelled as such) number.


The confusion was brought about by you jumping to a conclusion that can not be supported by any sensible interpretation of what I said.



No I did not. I am honestly at a loss as to how you've got to this from my post. I continue to give you the benefit of my doubt here.

I appologised for not stating that my number "2000" was approximate.

I see that as blooming obvious however to anyone as real numbers change year after year and are almost never round to 3 digits...

I can normally understand other posters positions even if I disagree to my core. In this case I can not see how a reasonable person would jump to the assumption you claim to have with regards your first reply to me - not that you stated your assumption then. I sure as hell cannot see a reasonable person sticking to that in the face of my initial apology over the omission of approximate.

In reply to wintertree:

I appologised for not stating that my number "2000" was approximate.
I see that as blooming obvious however to anyone as real numbers change year after year and are almost never round to 3 digits...
I can normally understand other posters positions even if I disagree to my core. In this case I can not see how a reasonable person would jump to the assumption you claim to have with regards your first reply to me - not that you stated your assumption then. I sure as hell cannot see a reasonable person sticking to that in the face of my initial apology over the omission of approximate.


This is your initial post with what you believe is the all important word approximately added: -

It's hard for me to feel negative when it means approximately 2,000 people a year less being killed on the UK roads and many times more not being injured.

We also know that approximately 2000 people a year die on the roads every year.
You can bluster all you want, but the point remains, your initial post indicates that you believe driverless cars will result in approximately 2000 lives being saved and using FactorXXX logic and a calculator I get the following: 'approximately 2000 - approximately 2000 = approximately 0'.
Maybe your logic and calculator are different to mine?

By the way, I'm an insomniac and have next week off - I can carry on doing this for days!

In reply to FactorXXX:


This is a reasonable interpretation of my post. It is entirely different to your previous assertion that I "stated" there would be zero accidents. Ignoring your massive mistake of conflating "deaths" with "accidents", you entirely muddled "implied" and "stated".

I thought it was clear from my initial apology at 20:05 and from my posts since that I understood the way you arrived at "zero deaths", and that I explained that was not what I meant.


I think that has been obvious since 20:05, which is why its so strange you continued using your logic to insist what I "stated" in the absence of a statement.

In reply to FactorXXX:

Give it a rest, you are being utterly tedious.

In reply to wintertree:

This is a reasonable interpretation of my post. It is entirely different to your previous assertion that I "stated" there would be zero accidents. Ignoring your massive mistake of conflating "deaths" with "accidents", you entirely mistook for "implied" with "stated".

I readily admit that I mistakenly said 'Will driverless cars mean zero accidents then' in my initial response to you.
However, if I had said 'Will driverless cars mean zero deaths then' instead, would that have made any difference to your many posts trying to excuse that flawed logic? Somehow I doubt it, as I don't believe this has anything to do with the validity of the actual statement, but has everything to do with you not wanting to admit that you made a fundamental error in that opening statement.
To end this discussion, why not just state the number of lives you believe will be saved by driverless cars?
Not a "Will be orders of magnitude lower than now" type statement either! I want numbers! Lovely logical numbers, ones that we can discuss about how realistic they are and how they may or may not be achieved, etc.

Bloody hell, is that the time! Beer to be drunk!!!

In reply to Blue Straggler:

Give it a rest, you are being utterly tedious.

Why?
Has Blue Struggler got bored of having a go at The Lemming and Timmd?

In reply to wintertree:

You do realise that your just feeding the troll?

We all know there will be deaths even if we hit 100% of all vehicles being driverless.
There will be the odd software and hardware problems there will still be the kid that runs out I front of a car that physically can't stop in time but taking the distracted human out of the equation can only be a good thing.

All that said I'm a biker and you will need to pry my bike from my cold dead hands to get me to part with it.

In reply to Dax H:

You do realise that your just feeding the troll?

Maybe there's an element of truth in that, but my initial reply was a bit of a tongue in cheek response to a statement that didn't add up numerically.
Anyway, it takes two to tango and I'm pretty sure that wintertree is as guilty as me for not 'giving in to the opposition'...
I apologise if it's felt that I have derailed the thread, it wasn't my intent, it just sort of got out of control when the two of us couldn't come to any sort of agreement.

In reply to Dax H:


It's starting to dawn on me. I'm having a very slow day though.


My sentiments exactly.


I might even dust off my full class A licence if/when autonomous safety features become widespread...

Motorbikes are a really interesting one in the autonomous future; it's much harder to imaging a bike proper being fully automated. Some gyro bike like the Lit Motors C-1 (vapourware that it is) perhaps, but I don't think that will hit the spot for most riders.

In reply to wintertree:

It's starting to dawn on me. I'm having a very slow day though.

Truce?
I'd be genuinely interested in what you would think are realistic accident numbers with driverless cars?
I don't think they'll be zero though... Well, not yet anyway!

In reply to FactorXXX:



Struggler. Clever and original.

And no, that's never going to happen.
I don't usually care about your posts but on this thread you've been such a tit. Acting like some of the teenage trolls that appear from time to time. Bye

In reply to Blue Straggler:

I don't usually care about your posts but on this thread you've been such a tit. Acting like some of the teenage trolls that appear from time to time. Bye

As I've said in above posts, I have apologised to the thread in general and offered wintertree a 'truce' to get the thread back on track.
I did say in one of those posts, that it takes two to tango, which it does and I think we could well have won Strictly with our dance routine tonight... (Strictly Come Dancing is a television show which features celebrities and their proper dancer partners competing against similar couples. They get marked by a panel of 'experts' and the winner gets to win something. I think it's ghastly, dread to think what you would make of it!).
Anyway, I'll shuffle back to my correct position in the hierarchy of the UKC forum pecking order, doff my figurative cap and stay in the background for a while.
You can go back to intimidating the like of Timmd and The Lemming. Don't worry about that from my behalf, the smeggers deserve it!!!

In reply to FactorXXX:

According to gov.uk in 2016 there were
1,732 deaths
the number of people seriously injured 22,137.
there were a total of 186,209 casualties of all severities.

I would be willing to put money down that these numbers could be cut by a factor of 90% by taking the human out of the equation.
Between my van and my bike I average 40k a year with a fairly even split between motorway, urban and country roads and the levels of muppetry I see every day is beyond belief.
I'm far from perfect myself, just last week I almost changed lanes right in to a car that was overtaking me, Fortunately I caught a glimpse of him in my blind spot mirror but I had started my manoeuvre before I saw him.

In reply to FactorXXX:




You did although it came after a lot of titting around and it's a bit unfair to do the "winter tree was just as bad" as he only started after a few calm posts trying to back you down.
But for my part I WAS in a bit of a grump yesterday having found a snapped micro SD card meaning that every photo I'd taken in Rome and Venice recently plus possibly some photos from a concert in Dublin, was irretrievably lost

So I am a tit as well


Also was disappointed that nobody on this thread responded to my two "morality" links which really are interesting

In reply to Dax H:
Something like 94% is down to human error. That's not just driver error though.

Nearly 10% still down to alcohol.

A pretty comprehensive article here on what the numbers are.

http://www.telegraph.co.uk/motoring/road-safety/8702111/How-do-accidents-ha...

In reply to wintertree:


And you're happy to accept so much data logging?

A significant proportion of the population get uppity about Google knowing their browsing history!

I would tend to have reservations about data logging on vehicles when it is unlikely to exist on other road users, it would be necessary to be very careful when the available data only covers half of the story.


Maybe the real question is who would own the data and under what circumstances would they have to disclose it.


I'm sure that we agree on some aspects but I have some concerns about your apparently unconditional enthusiasm for a greater degree of observation of us all.

In reply to Shani:


I guess that leaves the question of hoow you incentivise those whose journeys are entirely urban without unfairly penalising those whose journeys are extra urban?

In reply to The Ice Doctor:
Automated cars could be good for rural pubs & piss ups generally.

In reply to timjones:
We need to provision cycling - segregated cycle lanes. Cars will gave to give road space over to these other forms of transport - which offer safe passage - from infant to pensioner, reduce pollution and congestion, improve health, improve throughput, all on infrastructure that is cheaper to build and cheaper to maintain.

The Dutch have done it since the 50s. They're doing it very successfully in London now.

The question of 'how' has long since been addressed. Cars will still get access. There will still be public transport.

In reply to timjones:




How uppity do they really get though? I see a lot of intermittent 20%-informed muttering about "none of their business" but generally nobody taking "action" to protest or even just personally mitigate this...

In reply to The Ice Doctor:

I have a blog on this one:

https://in-tech.blogspot.co.uk/2017/04/are-autonomous-vehicles-really-safer...

In reply to timjones:


I am saying what I think will happen. Why do you persist in turning this round without evidence Ito questioning what I want or what I am happy with?

Vast amounts of data are already logged on the roads, far more than most people realise. It's carefully regulated.


Many new cars already log data in the run up to an accident. Anyone can buy a dash cam already so the other half of the story can get logged if they want.



Where did I say anything unconditional? You've been consistently over interpreting what I write. It's pretty clear to me that this is going to happen. I do hope the law continues to evolve at pace with the technology to protect personal privacy.

In reply to timjones:
Several millions of miles have been logged by driverless cars. They learn exponentially.

If I go out and drive one mile then I have learned one mile of road and have gained 60-120s of driving experience. Someone else then has to go out and drive that mile and will learn another completely different lesson.

If a driverless car drives that mile that logged data is uploaded to Google and all the cars now have that experience and learn whatever lessons were gained in that 60-120s, without even having to drive that mile.

Couple that with a machine that never gets distracted and very quickly you can see how advanced these machines will become very quickly.

In reply to wintertree:

For the first time ever I actually see ice doctors point here where he says we're working towards making ourselves redundant. Whilst innovation is an amazing thing and ultimately humans have always worked towards higher productivity where comes the point when there simply won't be jobs for people because of just how productive the world becomes?

In reply to wintertree:

You did sort of say that. You said when it means 2000 fewer deaths which is about the UK death rate so that would require pretty much zero serious accidents - assuming you were talking about the UK.

In reply to The Ice Doctor:

I really don't know enough about the technology or how it will all develop but I am very sceptical. Where I live most of the vehicles are farm trucks/Land Rovers that spend half their time crossing fields and farm tracks. Will these be banned from the road as I can't see how they will mix it with tourists in driverless cars. The local single-track passing etiquette includes protocols such as reversing for a truck with a livestock trailer and knowing where it is safe to pull over onto a verge and avoiding mega potholes and sheep. I can see automated cars coping with some of that but I'm yet to be convinced generally.

In reply to elliott92:


I was thinking exactly that today.
Been to the Helmsley steam fair watching early stationary engines do the work of multiple men.
I suspect the same arguments were had back then about making ourselves redundant.

In reply to blackmountainbiker:

The vehicles will communicate with each other so each time they reach a passing place they will know whether there is another vehicle approaching and will know whether to wait there or continue.

Exactly like trains do on single track sections. Someone/something knows what is happening and controls them/ signals them remotely.

In reply to DancingOnRock:

But not with a non driverless vehicle I assume. So would the default be that the driverless vehicle would reverse?

In reply to blackmountainbiker:


Oh. My. God. It was bad enough when one poster conflated "deaths" with "accidents" but now you're trying to muddle it even more by taking their conflation and changing it to "serious accidents". To be clear the "that" you refer to in your first line was someone claiming I said there would be "zero accidents". I never said anything of the sort. Period.

I made it clear PDQ that what I think is that we are going to see orders of magnitude decrease - so much closer to zero than now. Sure some people outside of cars will still make irreversibly bad decisions that a moving car can't avoid, and some cars will still have catastrophic failures and all sorts of oddball things will still happen.

But one day it won't be about 2000, or even about 200. Hopefully less than 20. Seems like a worthy goal.

In reply to wintertree:
Wow, you need to seriously chill out! What you said was 'It's hard for me to feel negative when it means 2,000 people a year less being killed on the UK roads and many times more not being injured.'

My point was, and it was very gently made, that for all fatal road accidents to be prevented in the UK (deaths are fatal and very serious accidents) it would (surely) mean we'd need there to be almost zero accidents. Therefore the other poster, don't know who it was, was reasonable to infer that meant zero accidents. Now go and have some Horlicks.

In reply to wintertree:
Oh, I see this argument has rumbled on for a bit, sorry to resurrect it. Let it be hey? I had got bored reading older posts but will be more diligent in the future. For the record, I'd like to see more driverless car technology used in cars that are driver controlled but not total automation as I can't see it working where I live. And I would love to see zero deaths which for the record, I think that increased automation could make a real possibility.

In reply to wintertree:

Oh. My. God. It was bad enough when one poster conflated "deaths" with "accidents"

Mmmm, was willing to let this lie, but you are being a bit disingenuous here...

In reply to john arran:


Unfortunately we seem to need to keep doing that over and over.

In reply to blackmountainbiker:


Why on earth would reducing about 2000 yearly fatalities to close to zero "(surely) mean" reducing about 140,000 years accidents to almost zero? I don't understand that logic at all, so I disagree with your point entirely. At the least it requires reducing the severity of the worst 1.5% of accidents.

I have on this thread developed an allergy to being told what I said, by people conflating two very different things. See the total piss taking by another poster if you wonder why I "overreacted" to you making the same attempt to cast what I said as relating to all accidents.

In reply to blackmountainbiker:
I suspect most people will start to use some kind of satnav to warn them that cars are approaching ahead or at least get some sort of basic traffic light warning system.

The technology is there and like smartphones everyone will adopt it eventually. There will always be 5-10% of people who will resist but they'll find it so hard to get anywhere they'll either adopt it or just get left behind.

In reply to blackmountainbiker:


This isn't going to happen, some accidents are unavoidable e.g. child steps out in front of a bus.

The question is what accident rate can you hit while still getting through the same volume of traffic on the roads as before? Clearly humans could reduce the accident rate by driving slower but you still need to get from A to B driverless or not.

In reply to springfall2008:


Computers can parallel process from multiple visual sensors, without getting distracted. Humans at best have two optical inputs.

In reply to Shani:

Yes, but that doesn't make the accident avoidable. Your driverless car can not slam on the brakes every time a child walks by on the pavement and a bus can't stop from 30 mph in less than it's braking distance. So if a pedestrian steps in front of a vehicle a collision maybe physically impossible to avoid.

What the driverless car has is faster reaction times than a human driver and could in theory be safer than a human driver (not yet proven mind you) but that doesn't make it 100% safe!

In reply to elsewhere:

I haven't seen any evidence that this is the case, lots of media hype but that aside what real world tests have actually been conducted? One million miles in a few cars in California with a human on standby doesn't really count.....

In reply to Shani:


If they were already that advanced then why haven't we seen the automotive equivalent of Deep Blue in F1? Should fit right in with all the other robotic drivers on the grid these days.

In reply to Stichtplate:


Sadly I don't think they are allowed in F1 (yet), but there is a driverless car racing series and they do have regular crashes!

In reply to springfall2008:


Didn't know that. I'll look it up.
Surely case proven, still better with a human?

In reply to springfall2008: You've decided to ignore the testing in the rest of the world then, including London?

In reply to springfall2008:


They showed that on the Horizon programme. it did look pretty rubbish!

In reply to wbo:


Maybe things have changed, all I have seen is very limited trials of a few vehicles with a human driver as a backup. Things are going to be quite different when they unleash 1000's of these vehicles onto the roads by themselves.

Take for example in my local area, a right turn onto a busy A road, you mostly have to go half way and wait for someone to let you out to actually make the turn.... I'd love to see a bunch of google cars trying to do that?!

In reply to Shani:


I think we can track about 20 objects at the same time.

Riding in a large peloton is a particularly bad idea.

In reply to Shani:


To quote a famous Scottish Engineer; "Ye cannae change the laws of physics Captain".

Once driverless cars become established, and more importantly all road vehicles become driverless, then accident rates will definately fall by a significant amount.

However computers can't alter the friction coefficient of tyres on the road, or magically create a safe evasive manoever when none exists. If the hypothetical kid runs out in front of the car and the car cannot physically stop, (no matter how fast the computer initiates an emergency stop) and the alternative is a head on with a truck or mowing down a bus queue, then there will be a nasty collision.

I'm also unsure how good the predictive abilities of these systems are. Yes they can probably sense a vehicle at a T junction moving before an alert driver can, but can they read subtle clues about the other car and driver that make an observant driver think "He's distracted / a dickhead"?

Once all cars are driverless then that issue largely goes away for cars and larger vehicles (but not with pedestrians, cyclists or probably motorcyclists). I can imagine the several years it takes to transition to fully automated vehicles might end up with a few unintended consequences.

In reply to springfall2008:


A bunch of google cars could probably communicate, but put a few apple cars in the mix which won't allow android cars to go first...

In reply to Ridge:

You're thinking too far downstream. In the example above, the computer will have detected and be continuously monitoring the child with an unblinking electronic eye, as well as every other person and vehicle around it; anticipating the dangers and calculating risks. We just cannot compete.

In reply to Ridge:


An interesting moral dilemma for the programmers then.

In reply to Robert Durran:


I got one Like, one Dislike, and no written reply to this which I posted much earlier in the thread.

I think you'll enjoy it.

http://moralmachine.mit.edu

In reply to Robert Durran:

No. The car will protect the occupants over the lives of other people.

.

In reply to DancingOnRock:


So mow down the child every time rather than collide with another vehicle?

That sounds to me like a dubious solution to a very real moral dilemma.

In reply to DancingOnRock:


Or more pertinently,the car will be programmed to protect the paying customers over the random strangers.

In reply to DancingOnRock:


"No" to what? A nasty collision? I assume that Robert meant there will be a collision with either the kid, the truck or the queue.

The link I provided creates various scenarios that presumably are the tip of the iceberg regarding what programmers will need to consider. It factors in legal stuff such as should the car plough into pedestrians crossing on their red light. It's interesting.

In reply to Stichtplate:


Not necessarily. It might be programmed to do the least ILLEGAL thing. Here, the kid has run out into the road whereas the truck is legally driving on the correct side of the road and the people in the bus queue are assumed to be law abiding citizens. Arguably the kid is the one who "deserves" to be hurt. Which is perhaps the same outcome as what you suggest, but for different reasons. The lawsuits that could be thrown at driverless car companies who programme them only to protect their passengers, would be huge.


And this is where it gets very complicated.

Click on the moral machine link I provided. Scenarios are randomly generated. Sometimes you get "hit the pedestrians legally crossing the road, killing 3 men and 2 women, or slam the car into a wall, killing 3 men and 2 women in the car".

I think it should take the latter course of action. I may be wrong on this.

In reply to Shani:


Of course it will be monitoring them, but it's not a profit and can't see the future until it happens.

Has anyone tried throwing themselves under a google car yet?

In reply to DancingOnRock:


As identified above the "other people" are likely to be pedestrians, cyclists or perhaps motorbikes. They don't pose much risk to the occupants of the car.

My best guess is that the car will stop as rapidly as possible. Far safer than trying to execute a last moment swerve. If it can predict ahead to a collision even by a hundred milliseconds it can also pre-deploy pedestrian safety airbags. It will certainly be breaking 250 ms to 500 ms before a human driver.

This sort of unexpected pedestrian or cyclist collision is usually (not always!) restricted to areas with lower speed limits, so the outcomes should be quite optimistic.

I've wondered before about cyclists on what we might call "future digitally networked roads". A cyclist could choose to use a location sensing/GLONAS app on their phone that broadcasts their position in real time to the traffic management systems. This way all cars would know about them in advance, and the cyclist could opt to receive audio or heads-up (glasses) warning of inbound vehicles.

If vehicles end up in a live, real-time mesh network then one car may see a pedestrian hazard and warn other cars that are at-risk but can't see it. All sorts of other information could feed in to the traffic management system - motor vehicles, cyclist and pedestrian phone apps, CCTV, "smart pavements", incident-responsive UAVs, all sorts.

The next 5-20 years have a lot of hurdles to overcome, but the longer term is fascinating.

Imaging growing up in a 2040s UK where it's basically impossible to get yourself hurt by a motor vehicle, and then going on holiday to somewhere with traffic like it is now in Bangalore. You probably wouldn't last long...

In reply to Blue Straggler:

I had a look at the link , but in the real world you can't code for human nature. The nexus between morality , legality and human nature is where the push towards privately owned driverless vehicles will fall down.
How many random strangers would you trade for the lives of your children?

In reply to Stichtplate:



Human kill people all the time on the roads and it hasn't stoped humans driving. It's down to some combination of training, attitude and sheer bad luck. Society accepts the approx 2000 deaths a eat unquestioningly and jail terms are often very light where the driver wasn't drunk/speeding etc.

Why would a much lower number be harder to accept?

In reply to wintertree:



Most people don't make the moral choices, they pick self interest instead. Look at the common reasons people give for buying the increasingly popular SUVs.

- they give great all round visibility.
>even though they obstruct the view of everyone else.

-they afford greater protection to me and my passengers.
>even though they are significantly more dangerous for anyone in a collision with them.

In reply to wintertree:

Do these exist outside of testing labs?

The last thing I heard about them, around 7 years ago, was that extensive testing showed that conceptually they were almost a non-starter as they tended to catapult the test dummies, damaging them AT LEAST as much. Admittedly this was "bloke in the pub" info. But I have never heard any talk about them since....

In reply to wintertree:


I think somehow that machine error is less acceptable than human error. A bit like if a climbing wall switched entirely to autobelays and one failed.

In reply to Blue Straggler:

The link I provided creates various scenarios that presumably are the tip of the iceberg regarding what programmers will need to consider. It factors in legal stuff such as should the car plough into pedestrians crossing on their red light. It's interesting.

I did the test and I think there is a fundamental flaw in the algorithm.
In any scenario involving the Concrete Block, then that was the option I chose - it was the car at fault, why kill other people?
For some reason, the algorithm didn't take that into account and seemed to simply add up the difference in number of females, homeless people, etc. killed and came to the conclusion that I was murdering misogynist, who doesn't like old people and thinks that the homeless are better off in a wooden box as opposed to their normal cardboard one.

P.S. If I was driving, there'd be no way I'd drive into a concrete block...

In reply to Robert Durran:


It may well be too complicated to code. How many of us would choose to kill a child rather than an old person in a split second moment? That older guy might be a cancer surgeon.... We know how awkward these contrived situations can become.

More likely, once a danger threshold is reached, the car would brake. The measure of danger would trigger the braking rate.

In reply to Shani:


But surely there are situations where it would have to swerve (and brake) and risk hitting something other than the child. A bush? A small tree? A parked car? A moving car?

In reply to FactorXXX:
To be honest I mostly don't look at the results and certainly don't treat it as a test. I think the value comes from looking at the scenarios and just trying to think about what's best.

I think there is a way for users to upload their own, in fact I somehow ended up on that page the other day. Some interesting scenario titles e.g. "MILFS or dogs" :-D

In reply to Shani:




Moral Machine site scenarios all assume total brake failure on a car travelling at speed.

In reply to The Ice Doctor: *


Following a white line around a race track it might have.
Driving around Manchester on a Saturday afternoon, forget 2040, maybe 2100.

Your Brave New World is decades away.
As I said before, look back to 1994, nowts changed.
I can look back to when I first came to Manchester 1978 and I can cruise around and whole swathes of Mancunia are exactly the same as they were 39 years ago, apart from shell suited chavs wandering around with smartphones.

I'm definitley going to try and remember what it is now and what it will be in 2040, if I live that long.

In reply to Blue Straggler:


Talk about cart before horse. Brains of Deep Thought, brakes of Austin Allegro!

In reply to Robert Durran:


Arguably true, but the discussion is about humans dying as a result of human error, in which machines are involved. Its very much "but what if a human steps out immediately in front of the car".

There will be "machine error" deaths just as there are now we cars (a small fraction of deaths) and every other kind of transport machinery.


Given the context of a human jumping out in front of an automated car, it's more like a climber not tying on to an autobelay and lobbing off. Still an imperfect analogy but it happens and auto belays are still a thing.

In reply to Stichtplate:



I said "at speed"!

We briefly had an Allegro as our family's second car, so I am allowed to comment .

In reply to Robert Durran:

Morals have nothing to do with it. It's a machine.

It's not my suggestion, it's how they're being programmed now.

In reply to Blue Straggler:

You're lucky , we had a second hand one as our family's main transport for a couple of years until terminal rust did for it . My dad loved it, the lunatic!

In reply to DancingOnRock:





Do you know for sure? Even given that I, as no expert at all, have managed to posit different possibilities for how they might be programmed? It sounds like remarkably unsophisticated, if you're really saying that for sure all driverless cars will simply be programmed to protect the interior occupants regardless of all other factors.

In reply to Stichtplate:


Hahahhahah, I nearly killed us all careering into a petrol station once behind the quadril joy of an AA!

In reply to Stichtplate:

We sold it to a couple of lads. They painted the bonnet black with a flower on it (the car was a dull orange). Some months later we got ourselves into the back of an uncharacteristic long tailback heading into our home town on a A road. 30mph instead of the usual 50-60. At the front of this tailback, with all black smoke coming out of the engine, was our old Allegro. A fitting end....
We'd had a Hillman Avenger as "second car" before that.

In reply to springfall2008:
It certainly won't happen whilst there are a mix of driverless and human controlled cars. Also, what would a driverless car do in icy/snowy conditions? Would it just not drive at all?

In reply to DancingOnRock:

In that case, if that is true (and I very much doubt it's fixed for all time), the moral decision has already been made. Of course that's got to do with morals!

In reply to Lusk:


When challenged on his love for the Allegro, my father claimed that he found the unpredictability of it's cornering 'exciting'!!!

In reply to Stichtplate:

Did he have a Vanden Plas, or whatever it was called, with the Rolls Royce grill?

In reply to GrahamD:

I'll grant you the Rover bit. But the Panama hat, and fish symbol remains a guarantee of a doddery, frustrating, clutch burning urban trudge that will have you wishing that you'd gone on the motorbike

In reply to Lusk:


It was indeed. Supposed to be the classier model I think? I mainly remember it for the pvc leather seats and an overpowering smell of wet dog and condor pipe tobacco.

In reply to Stichtplate:

The man's got class! Me mate's dad owned a garage back in 70s and he was a proud owner of said vehicle.
Did it have a vinyl roof as well?

Best car I've ever had was a 2 litre Capri, shit brown with black vinyl roof, nothing has come any where near close since.

In reply to Stichtplate:

is nobody going to mention the steering wheel ?

In reply to Lusk:




Pretty sure it did . I always coveted the Capri 3.0 with the speed bump on the bonnet, loved that car. I'd take one of those over a self driving Lamborghini with Hal behind the dash any day.

In reply to lithos:


HaHaHa! Square and massive!

In reply to Blue Straggler:


Yes.

http://www.autoexpress.co.uk/mercedes/97345/mercedes-autonomous-cars-will-p...

In reply to The Ice Doctor:

Wow, I'm late to this party, but for me, the main benefit (which I haven't seen discussed yet) is small, one person, probably uber-like vehicles, taking up half the space on the road.
As most journeys appear to be 1 person, 1 car, makes sense to save fuel and have a pod...
No one owns these, but books them for the commute, taxi journey etc. they're never parked on the street, they could be allowed on narrow roads to ease congestion, car at home is more conventional. I think this will be where driverless and green go hand-in-hand.

In reply to DancingOnRock:

Thanks for replying.

In reply to Oliver Houston:




A vehicle which uses less space (less than half the space of a car), and saves fuel, and eases congestion? We already have this green technology. It's called a bike.

In reply to Oliver Houston:

I think the problems would arise where everyone wants to use them at the same time.

More likely would be minibuses with more intelligence than a current bus which runs on a fixed timetable.

You book a 'bus' time slot and it stops close to the end of your road for you. Giving you an accurate time for pickup.

The busses could be networked together extremely well.

This could probably already work with Uber cabs if someone got their act together and programmed it.

In reply to Shani:

I'd love it if I could cycle safely everywhere, but sadly, too much time/space/money is spent making the roads more convenient for cars, people won't cycle if they feel like they're going to die. (also my wife's commute is a hilly 25ish miles, not really practical by bike).

If city centre's were "narrow car" only, more space could be dedicated to cycle lanes, cycling becomes safe, more people cycle, I see it as win-win.

In reply to DancingOnRock:

Yes, that would work also, but for people with inconvenient/odd commutes, a small vehicle would work as well.

It would mean that one vehicle could do the commute for many people, going from one to the other, reducing the need for city-centre car parking.

In reply to DancingOnRock:

It's annoying we don't do this already by installing a tracker/transmitter (a basic smartphone in a blackbox) on every bus so anybody with a smartphone can look at a website (scan qr code at bus stop) to see an accurate time for when the bus will arrive.

In reply to elsewhere:

You could couple this with a 'booking' system. Let the bus company know 30mins in advance of your journey and instead of a bus running around a route empty all day, it could sit at a depot until enough people requested it to run.

It would probably need a bit more thought, research and testing but would be a good use of technology.

In reply to elsewhere:

Certainly in London the bus stops have elecrionic displays with accurate arrival times.

In reply to Oliver Houston:

You'd probably be best having a mixture.

Could even add a bike trailer to hold 20-30 bikes.

In reply to Oliver Houston:



Yep.

As a general rule, if six year old cyclist cannot safely use it, then it is not adequate cycling infrastructure. There is a lots of room on the roads, but the space is used inefficiently. We need to use one-way systems to facilitate cars. The 'other lane' can then be used to facilitate segragated, two-way cycle lanes. With appropriate bollarding, emeregency vehicles could access the cycle lane.

The Dutch faced a lot of opposition about there being insufficient room to build great cycling infrastrucutre, but with perserverance, they did it.

Add in 'narrow cars' with electric engines and cities will be a lot healthier.

In reply to DancingOnRock:


You are being tongue-in-cheek, right?

In reply to Blue Straggler:

https://en.m.wikipedia.org/wiki/IBus_(London)

In reply to wintertree:

Don't take it so personally, if you raise points in a discussion others will challenge them or raise counter points.

That is how debate works.

In reply to timjones:


You are the one who has responded to multiple pairs of what I think *will happen* and translating that into what you think *I want* with negative judgements. You're making it personal.

You might review this. I am not taking the debate personally, I am objecting to you attempting to over interpret my personal views etc.

In reply to DancingOnRock:

Have you never in recent years been waiting for a bus in London at a stop armed with a readout of expected times, and seen the time get shorter and shorter and then the bus "disappears" just like they did in the initial phase of this technology in the late 1990s?

In reply to wintertree:

I've merely asked questions based on points raised by other people.

I'm sorry to burst your bubble but it hasn't even registered that you had apparently posted a disproportionate number of the posts that I replied to