The Queen’s speech last week contained some cracking headlines about spaceports, drones and driverless cars as part of the ‘Modern Transport Bill’. In terms of the cars, I believe the legislation is really just about providing clear regulation frameworks and insurance liability rather than any serious public investment. However no sooner was the speech completed when various ‘experts’ were being interviewed on TV accusing the government of wasting time on pie in the sky projects rather than focusing on the real transport issues. One such expert on Channel 5 news that evening even said that driverless car technology was ‘decades away’ and that at the moment the cars couldn’t even do basic things like drive in the rain!
What’s the truth? And why should we care?
Well driverless cars are most certainly not decades away. At the moment Google cars are being driven for millions of miles on public roads in California, Texas and Seattle, the last specifically to overcome the challenges that heavy rain and hilly conditions pose to the cars. Apparently the reflective glare from rain and spray on the roads makes it harder for the cars to determine exactly what its sensors are looking at, and so they’re not yet deemed safe enough. They’ve also not yet overcome the rather basic problem that when the car drives up or down a steep hill its sensors don’t quite compensate for the changes in elevation and so end up looking down at the floor or pointing up at the sky!
Despite this, there should be no doubt that these challenges will be overcome. One pragmatic reason to be optimistic is that Google alone is throwing billions of dollars at this problem, pursuing its stated mission to bring fully autonomous vehicles to market by 2020. And Google is not alone. Ford, Tesla, Apple, some of the massive Chinese state car makers and many others are all spending a fortune working on this technology, and that’s before we start factoring in the public funds which will become available as governments around the world compete to ensure they’re not left behind. It might be a few years into the next decade, and it might not be all cars on all roads, but driverless technology is coming soon(ish).
And here is why those seeking election should be paying attention. If you take a look at this excellent ‘TED’ talk which gives a view of how the technology works, it becomes clear that the car is basically a robot. It has to see and think and act for itself in a complex geometric world full of other people. This sort of artificial intelligence isn’t just useful for cars, it will be the basis of operating robots in all walks of life, including military drones, nursing carers, surgical assistants, construction rigs, mining, farming, manufacturing etc. Whilst applications in some of those fields may well be decades away, the technology being developed for driverless cars opens the door on a world of unimaginable economic opportunities. That is why these companies are throwing money at it. Consequently, an imaginative industrial policy encouraging some of this research onto our shores, giving these companies the clarity of regulation, intellectual property protection, testing facilities, the graduates with the hard science skills, the open door to investment and skilled immigration, and I dare say a few tax incentives, could pay big dividends.
In that context, the government is spot on with the Modern Transport Bill. Liberal Democrats should go much further, crafting an industrial policy that puts exactly this sort of future industry at its heart, and simultaneously associating our party with the truly progressive technology that the 21st Century will be built on.
* Tobie Abel is a software designer and PPC for Richmond Yorks. He joined the party in 2013
Follow @LibDemVoice on Twitter
Like us on Facebook
Subscribe to our feed
Sign-up for our daily email digest
Mystery at Witchend is Went the Day Well? and Seven White Gates is A Matter of Life and Death
The long role of the Jewish community in Liverpool
Labour's post-election (non-)honeymoon: The Week in Polls
September now
Farron photos
Joe Bourke
Peter Martin
Mick Taylor
Michael BG
Jennie
40 Comments
“Consequently, an imaginative industrial policy encouraging some of this research onto our shores, giving these companies the clarity of regulation, intellectual property protection, testing facilities, the graduates with the hard science skills, the open door to investment and skilled immigration, and I dare say a few tax incentives, could pay big dividends.”
I am sure the nurses, carers, taxi drivers, bus drivers, railway workers, construction workers, oil rig workers, factory workers, soldiers, miners, surgeons, farmers and so on will love being laid off, and then competing with more immigration, whilst foreign investors rake it in.
If such an approach is pursued, then you will need to have some form of Universal Basic Income, which should be sufficient to live reasonably off. And that means no tax breaks for foreign investors, as they will have to shoulder the cost of it.
@Richard – 250 years into the industrial revolution and you’re still worried that technology destroys jobs. It does, it creates new ones too, and soldiers and nurses and everyone else somehow always manage to adapt their considerable existing skills.
The link the the TED talk by the way is : https://www.ted.com/talks/chris_urmson_how_a_driverless_car_sees_the_road?language=en
Any kids who are reading this should pay attention and consider the good life they will have after getting a law degree needed for the driverless car class action lawsuits that are coming.
A million mile proof of concept publicity stunt is has no value in proving that a driverless car is safer than human operated vehicles. You guys are just fooling yourselves. Also, if any one feature of a driverless car makes it safer than human operated vehicles, then that feature can be installed into the real cars.
There will be no driverless cars for sale until after it is proven they are safer than human operated vehicles. In other words, Google wants to watch your activities and harvest metrics about your travel. Google wants to create a fictional moral equivalency that they are like seat belts and wiper blades so laws will be passed that allow them to be installed into every human operated vehicle. The driverless cars you see today are just marketing to make this happen.
Soon, the only way you will be able to escape Google monitoring your activities will be if you die in a horrible traffic accident.
I believe if Google wants to create a government mandate to spy one human drivers, they should look to do this in areas of the world where people have no conflict to let their private right become infringed.
Your reply to Richard strikes me as a bit glib, Tobie. Change happens, yes. We cannot always predict how, and human attempts to manage the economic transitions that we do see coming are often partial and imperfect (and at worst, counter productive).
But writing people off as economic collateral damage, saying ‘hey ho’ and moving on as if nothing has happened is not an adequate reponse to any of this.
We need to count the cost of a reform, even if we are convinced of its benefits.
With regard to driverless cars / buses / whatever – we need to remember that driverless cars in the US are being produced in a large nation that is not really very urban in terms of its overall usage of land, where its population is quite spread out, and where conventional public transport is under-invested, and railway passenger transport in particular is largely obsolete as a national carrier.
None of that applies to us. That doesn’t mean that we shouldn’t consider this change, but I think we should listen more to that Japanese and Korean experiments with this technology than to the US and Chinese, as Japan and South Korea are somewhat more like ourselves in terms of population, infrastructure demography and geography.
Arguably, the technology might benefit Northern Ireland and Scotland more in some ways, as there you meet more often with urban centres surrounded by sparsely populated hinterland which does not create ideal territory for simply defined commuter routes for trains or buses, and where train networks in particular have been stripped away rather aggressively in the past.
So if we think this technology (or some iteration of it) is viable we need a multi-strand approach that recognises any new technology needs to coexist with the continued existence and ongoing development of existing technologies and infrastructure, and does some serious early thinking on how that will happen, because it won’t happen by magic, and it is really not certain it will happen purely by market forces.
Another thing I feel is important from a liberal and democratic approach to well, really any infrastructure planning, is that certain regions of the country are not left behind whilst the richer areas are picked off by the promoters.
And never forget that whilst science and development industries certainly do benefit from going first and experimenting, often a country that waits, learns and goes second gets a better, tested workable infrastructure in the long run.
@Matt (Bristol) – The only way to those better paying jobs every party seeks is through disruptive technology, not sure why driverless technology would be viewed any differently than the internet. People adapt, the economy grows, nothing to be afraid of.
I’m hopefull your comments about the rural/urban differences and needs for infrastructure are largely irrelevant for this technology as its more like TomTom sat navigation. The car carries everything it needs within it, and operates just fine with existing cars with drivers.
@Neal Palmquist
I did think that driverless cars were a very long time off. I now think that they are very much sooner than that – it wouldn’t now surprise me if there were a lot around by 2025-2030. Google’s PR website https://www.google.com/selfdrivingcar/ for its driverless cars – says that there are being tested in urban areas such as Austin, Texas and Phoenix, Arizona.
There are obviously a multitude of legal problems – such as “faulty” software and possibly being hacked. But we have overcome these in other areas from a legal point of view. Google says their driverless cars have done a million miles with only 14 accidents and these have either been other drivers’ faults or when they have been driven by a person (not autonomously). If as Google says 94% of accidents are due to human error, there is the potential to save millions of lives worldwide.
Although Wikipedia https://en.wikipedia.org/wiki/Google_self-driving_car says that driverless cars are only at the moment doing preprogramed routes and can’t cope in heavy rain or snow.
On technology and jobs – technological change is difficult – particularly for those affected but arguably we are better off today than having most agriculture involving a lot of labour – first mechanisation and industrialisation and now the knowledge/IT revolution – by and large has given people better lives and greater incomes.
Consequently, an imaginative industrial policy encouraging some of this research onto our shores, … the graduates with the hard science skills, the open door to investment and skilled immigration, and I dare say a few tax incentives, could pay big dividends.
If Tobie you had done some further research you would know that under the Coalition, Google was encouraged to locate some of its R&D efforts in the UK. This permitted Google to make a before tax profit on it’s UK search activities and reinvest it here (in the UK) in it’s loss making R&D activities, along with other monies that it didn’t want to return to the USA and thus be subject to US taxation…
One of the reasons for the government being so supportive is that such R&D is naturally closely related to Universities, so by promoting our Universities to Google, the government was also promoting our universities to potential students, ie. study here if you want skills for the future. So the government has already started the chicken and egg cycle, because regardless of whether driverless cars ever become economically viable or not, there is going to be much technology spin off and acceleration. Additionally, as Richard Easter pointed out, there will be significant economic changes both in terms of the established order, but also in what new businesses such technology enables.
Some of the technology being developed will undoubtedly be very useful in making road vehicles more efficient (in navigation not mechanically) and safer. But the idea that all cars or even any cars on our ordinary roads will be “driverless” within ten or twenty years is pie in the sky. What it may well do is to change the nature of driving.
At the most basic, the person in charge of (“driving”) the car is still going to have to decide where they are going. And that is not something that is decided before they set off. It’s something that is subject to lots of decisions from time to time and often at frequent intervals.
Someone – a person – always has to be in charge.
Tony Greaves
@Roland – yes I know. The coalition also setup the Catapult centres which are also investing in this technology (which is shown in the picture with Vince Cable). Not sure what your point is.
The Heathrow ULTra system seems like a far more likely model for future rapid transit.
@Neal
“There will be no driverless cars for sale until after it is proven they are safer than human operated vehicles.”
I certainly hope so. The moral and legal arguments around all this are intriguing. In order for this to be viable, the government would presumably have to grant some kind of legal immunity for companies who produce these “robots” that go out on the roads and (inevitably) kill people. No corporate manslaughter charges for Google and Apple, however negligent their software may be. People whose loved ones are killed by these things will get no justice. And yet if they really could be made much safer than human drivers…
I dunno driverless cars sounds cool and more robots taking over human jobs sounds intriguing too.
But maybe at that point it’s time to admit that the notion of work in the traditional sense is basically an extension of feudalism or even a form of slavery. I often wonder why people seem to passively accept the idea that it is reasonable to expect great chunks of the population to give hour after hour of day after day of year in and year out of their one precious time on Earth over to the task of other people wealthy? I think I may even be turning into an anarcho hippie of some sort as I get older!
@ Tony Greaves
I think that the first uses are likely to be public transport/taxi like applications – a taxi but without the expense of drivers. And uses where the time of the person can be better used – e.g. working while travelling. And where the person can’t drive – disabled, too young, too old, no car.
“But the idea that all cars or even any cars on our ordinary roads will be “driverless” within ten or twenty years is pie in the sky.”
Well – maybe. But..
Number of cars in US in 1895 300, 1914 1.7 million – http://www.carhistory4u.com/the-last-100-years/car-production/the-increasing-number-of-cars
Probably thought to be 1. expensive, 2. unreliable, 3. Needs petrol etc – no petrol stations, What about trains, buses, trams, bicycles – all better initially.
Mobile phones in 1985 in UK:0, 20 years later: millions
What’s the need? What about phone boxes? Too big/unreliable, expensive.
world wide web use: 1989: 0, Today in UK everyone (virtually!)
You mean like a newspaper but less portable?
Driverless cars????
You mean like a car and I sit in a car – where’s the difference?
It’s difficult to predict!!! And there becomes a tipping point. Given a car at the same price with and without driverless ability, most would opt for the later. Teenage child out late – send the car to pick him up. There can be new applications to technology that are only dimly if at all foreseen. It could well be that driving oneself will be seen as antiquated in the future as a horse-drawn carriage is today. The ability to get somewhere (potentially) much more safely and do things on the way there (surf the internet!!) – who knows!! Personally I don’t see driverless cars being such a quantum leap as other technologies but they could be a sizeable percentage – once the technology works well and is cheap.
Trains/buses/trams obv. still exist and dominated for a long time but the dominate form of motorised transport is the car.
That should be “former” in my previous post – opt for with driverless ability
There was a story on the news today of a school bus that was driving onto a bridge and the driver saw that the bridge ahead was collapsing, so he reversed off and saved the lives of himself and the passengers. I wonder how a driverless car would have behaved?
Human driven cars kill hundreds of thousands of people worldwide every year. I find it hilarious that anyone would think that the bar for driverless cars doing better is particularly high. Basically, once they can reliably get you from a to b, they will be safer. The legal approach will need to be similar to the aircraft industry where rather than endless blame and law suits everyone learns from everyone else’s mistakes.
Driverless cars will be hugely liberating for those on low incomes who will be able to afford taxis, and to the elderly and disabled who are often trapped at home. They will also save most people from needing to own their own car so will solve parking problems in urban areas too. Traffic will flow better and congestion will decrease, because most traffic jams are caused by human drivers driving too close or changing lanes without warning.
Some jobs will become obsolete, but robots will make other jobs viable for the first time. Technology just means we spend less time doing manual chores and more time being creative.
We have planes that are more than capable of flying themselves but still insist on putting pilots (and a back-up co-pilot) just in case. These pilots are highly trained and motivated, so when the time comes that they need to take over in an emergency they are hopefully not asleep and don’t both have their faces buried in angry birds on their smartphones.
Self driving cars will need to be good enough that they can react to any unexpected situation without the driver taking over suddenly.
Of course the vast majority of accidents are driver error, and so self driving cars are likely to reduce accident statistics overall. However, I think the biggest issue will be liability. Car manufacturers are mostly massive multinationals with deep pockets. If I kill your child whilst driving, I may go to jail and you can try suing me, but I don’t have much. The car manufacturers however……
Expect massive lobbying of the Government to enshrine some sort of indemnity for the manufacturers in law.
I also predict in the future reckless kids playing “chicken” with self driving cars, by jumping out in front of them at the last moment and forcing them to brake or swerve. You heard it here first….
I really like Tim Ps naive and wonderfully optimistic “Technology just means we spend less time doing manual chores and more time being creative.” I am old enough to remember the 1980s when the mantra was “We need to train people to prepare them for the increased amount of leisure time all this new technology will give them.” Instead we found a never ending series of mobile phone calls to take up what had previously been a natural break in the work and a trail of e-mails telling us to work on our bosses latest half baked ideas. This was accompanied by ever reducing time budgets for work because we were all tasked with improving our own productivity. Typists were laid off because everyone could type for themselves on their PC, except they couldn’t and were slower than the typists (and all the admin associated with using them). Of course senior directors still had their secretary to do their typing.
It’s not the technology, Tim, it’s what those in power choose to do with it.
Any technology can and does go wrong. What is more, designers and engineers are quite poor at anticipating unusual accidents because they think in channels and the essence of an unusual high-tech accident is that it leaps the channels (see Charles Perrow, “Normal Accidents”).
That is not an argument for not developing the technology, and while we need to be thinking hard about the implications for society of many jobs becoming redundant, we can’t stop it being developed somewhere in the world, though we could legislate to control use.
Potentially driverless cars would be safer than the present ones because they could see a pedestrian or judge the speed of an approaching motorcyclist when human judgment might fail. But as Tony says, features that automatically brake when a hazard appears, for example, can be introduced into cars still possessing a driver. There are dangers here too until the computer’s judgment is better than a human’s (for example, when road conditions would make braking hard more dangerous than other reactions). It all needs a lot of testing. I suspect the best imperfect solution may be driverless cars with a time-limited override.
But many people (even environmentalists) take some pride and pleasure in driving well and the same goes for many other activities that could be automated. It could make people feel irrelevant unless other rewarding activities are developed.
@David Evans – I’m young enough to remember the 80’s and to have lived through the fruition of some of that mantra. Working from home, instant communication, flexible hours, and no-one has to do the typing (i’m using Google Voice Typing!)
Still, I do wonder where the promise of extra leisure time went. Just goes to show there is always more work for humans to do, no matter how technically advanced we get.
Although driverless cars will make the news, driverless trucks will probably be more important at first. Big trucks are EXPENSIVE so the cost of all the automation wouldn’t put buyers off, and lower costs would justify the investment – especially as there is a severe and growing shortage of drivers in the UK at present. Whether they’ll be able to do the whole journey, or take on a human driver for the delivery end, will probably depend on political issues as much as on the technology.
Germany is already experimenting with driverless lorry convoys, in which a number of lorries follow a human-driven lead vehicle.
The downside of this is that lorry driving currently provides quite well paid work (£25k to £40k pa) for a LOT of people, some of whom probably won’t be able to find other work – though as the average age of a LGV driver is 53 and with 13 per cent over 60, many may be ready to retire by the time driverless goods vehicles become the norm. (Source of data: BBC and other web sites.)
@Tobi – sorry, my point was the existing measures and incentives and supportive government words, seem to be sufficient to encourage technology R&D investment into the UK. Your sentence implied additional incentives were necessary, without providing any rationale for why.
As a software developer I am fully aware that the leading edge technology soon becomes the bleeding edge technology.
So let Google explore all the issues in wet, hilly Seattle and then release such open source data to the world so that the information will lead to better decisions wether or not to allow driverless cars.
Michael “I think that the first uses are likely to be public transport/taxi like applications – a taxi but without the expense of drivers.”
Well I suspect for many years (or even decades) that is likely to be their only major use, given firstly, the current cost of the in-car equipment and taking into account the likely benefits arising from advances in manufacturing and volume production and secondly the massive levels of on-going R&D and support needed to keep the vehicles working [1]. One thing is certain driverless car technology isn’t like mobile phones, so it is unlikely to go from a £100,000+ fully equiped van (excluding the cost of the van) to a £400 in the pocket device that gets replaced every couple of years.
Hence it would seem the only way to get this technology out at a “commercially viable rate” (ie. a rate that discourages competition) will be to operate it via hire/shared usage schemes – which is exactly where companies such as Uber and Lyft are positioning themselves…
[1] An example of the on-going support work, will be the addition of new developments – yes those 100,000’s of new homes some would like us to build – currently, it takes 3~4 years for a new development to appear on maps and a further 3~4 years for Satnav’s to be updated (assuming the user is paying for updates)…
I’m happy for the government to allow driverless vehicles to be tested on UK roads. I would certainly disagree with any government investment or tax breaks for driverless technology.
We have to consider that a lot of the hype amounts to wishful thinking. Driving in the UK is rarely a pleasure so we’re positive about self-driven cars which allow us, theoretically, to take a nap or read a book. A lot of people want something without considering how much has to be achieved.
Artificial Intelligence (AI) can be split into machine learning (statistical processes on top of a set of rules) and deep machine learning (understanding systems and dependencies as well as or better than an expert human). At the moment, AI uses machine learning for written language translation, voice control, problem solving etc. As we’ve all experienced, contemporary AI mangles things up. Voice control systems fail to understand dialects. Written language translators may not catch nuance, irony or neologism. The systems are good enough for limited purposes — and when they aren’t, we use a human. Deep machine learning is years away, and when it arrives it will initially require a data centre to make decisions.
Humans understand motoring, roads, urban and rural environments owing to years of observation. A 17 year old European who is learning to drive typically has a few years as a passenger or a social observer. As an observer, we can’t help noticing how different the world can be within a very short distance. Without realising, we absorb a lot of data. Sometimes we comprehend why things have changed; at others we have so much information (e.g. in a busy urban street) that we lose the background in order to focus on what is in front of us.
to be continued.
Continuation…
In the UK, we have built a road system primarily for cars and lorries over 100+ years. Well, we’ve reworked what the Romans built and cattle drovers created, then added to the medieval streets. We’ve also created driving cultures for our different environments — small town manners versus city manners. Built up environments have changed to accommodate cars.
France had a different road system from the UK when the modern automobile was conceived. It was the best place to drive cars so French manufacturers had a huge advantage. But in order to sell cars which worked in the UK or in European colonies, the cars needed a bit of reworking.
Google has made a huge fuss about its driverless car studies. This is a bit of a distraction; European companies such as Volvo, Audi and BMW are the ones that matter; they are most likely to build cars available in Europe to European consumer needs. They are far more aware of cultural and technical differences, and much more realistic.
Here’s a test for Google’s cars: Drive from Paris to Moscow to Beijing on minor roads (i.e. no motorway), the 1907 Peking to Paris race sort-of in reverse. My guess is that the trial would fall apart in Belarus, perhaps in Poland.
@Roland
Costs:
There is an article at http://www.fastcompany.com/3025722/will-you-ever-be-able-to-afford-a-self-driving-car and it says: “IHS Automotive forecasts that the price for the self-driving technology will add between $7,000 and $10,000 to a car’s sticker price in 2025, a figure that will drop to around $5,000 in 2030 and about $3,000 in 2035, the year when the report says most self-driving vehicles will be operated completely independent from a human occupant’s control….IHS predicts that annual sales between 2025 and 2035 will jump from 230,000 to 11.8 million. That’s about 9% of all the world’s auto sales in 2035. Seven million of those 11.8 million vehicles will rely on a mix of driver input and autonomous control, with the remaining 4.8 million vehicles relying entirely on computers to get around.”
And such a forecast could easily be off by 10+ years (basically either way). Technology suddenly reaches a critical mass of usable that is difficult to predict.. I would guess not one report at this stage of a technology has been accurate! But 20 years is not that long – it seems (to me at least!) a blink of an eye since mobile phones were expensive bricks only used by a few yuppies.
Moore’s law of microprocessor power doubling every 18 months – 2 years is still going and IF it continues that would make them 1000+ times more powerful than today in 20 years time.
There are massive ethical considerations – not just on this but AI in general and there is the debate that has been had here between humans being good at dealing with novel/unusual situations and sometimes inattentive in routine ones and probably computers the reverse.
There is a quite interesting report on road accidents at http://www.telegraph.co.uk/motoring/road-safety/8702111/How-do-accidents-happen.html
@Michael: “Well – maybe. But..
Number of cars in US in 1895 300, 1914 1.7 million”
Michael suggested a few explanations for rate of car adoption. He didn’t mention roads on which cars could be driven. If we consider the Edwardian period, USA cities had paved roads within themselves, but there was nowhere to go by car. Many early USA car adopters bought electric cars for town use. The roads out of the city were built for horse drawn carts at best. Even in the 1950s, driving out of town to meet the highway could be a challenge.
The Ford Model T was designed to work on such roads; it is a notable point in the differing evolution of American and European cars (the Model T was a big seller in the UK for a few years because roads here were so horrible).
@Michael: “X forecasts that the price for the self-driving technology will add between $7,000 and $10,000 to a car’s sticker price in 2025, a figure that will drop to around $5,000 in 2030 and about $3,000 in 2035…”
Why aren’t these people forecasting the weather? Perhaps they advised the banks? I note some scepticism in your voice — beyond AI perception, perhaps.
@Michael: “Moore’s law of microprocessor power doubling every 18 months…” Nope, Moore’s law is about how many transistor circuits can be crammed onto a chip (density), not about processing power.
@Roland – oh i see, sorry. Yes, the current government and the coalition before them have done a lot of positive things. The point of my article is that:
1. We need to keep doing more because the technology is key to our economic fortunes
2. The more Lib Dems associate themselves with these progressive technologies the more electorally attractive we become.
@Tobie Abel: “The more Lib Dems associate themselves with these progressive technologies the more electorally attractive we become.”
Technological change is just change. It does not carry moral or progressive weight with production efficiency. We’re not Ayn Rand or Joe Stalin followers.
The RAC says do not drive between 3 and 4 in the morning because your body clock puts you to sleep and the accident rate is double.
If you arrive at Accident and Emergency quickly the doctor might be in the same situation.
@Michael
I personally have driven over one million miles with only zero accidents, so I am not impressed.
There are people who do this testing for a living. I know some and will try to engage a discussion to learn more. For example, the tiniest feature to be changes in the software of an ATM machine requires an unthinkably large amount of highly regulated and documented testing for certification. But here’s the point, those tests labs can be set up in very large and expansive buildings in isolation where they do not affect any end users or customers. Driverless car road tests cannot. To imagine five nines for road test, you can’t possibly think one million miles with 14 accidents is sufficient. How do you know that humans are not already six or seven nines? Google knows this, trust me.
I think an effort to your point would be to offer that driverless cars can not be consumer grade products. They will be a category between military and hospital grade.
It is not a position I would stand firmly to say the immediate and indefinite ban of all driverless vehicles would never save so little as just one life. I also see more lives at stake and I consider their weight in this matter. Countless have died and continue to offer their lives for love of country and it’s values. With all respect to a great man who it would serve me to learn more, it is those other lost lives that are the source of the private rights of American citizens as they were not born from King George. This driverless car fever appears not serious at all and there is nothing more than marketing to attack private rights here.
@Tobie Abel: “The link the the TED talk by the way is : https://www.ted.com/talks/chris_urmson_how_a_driverless_car_sees_the_road?language=en”
It is a TED talk. Not everyone trusts TED.
When Chris Urmson proclaimed that Karl Benz, “a true story”, crashed his car “on his first drive” , I looked up Karl Benz on the internet. The web saves looking facts up in a book; but I think I have books to back up my version. Benz made many trials of his creations and one crashed into a wall, not driven by Karl Benz.
So the “true story” told to us by Chris Urmson isn’t true. The tale that Urmson narrated in the first two sentences of his TED talk didn’t happen. It was a funny anecdote to shift our perceptions. One has to question whether it is worth listening to any presentation if the opening arguments are so flawed.
@Matt (Bristol): “With regard to driverless cars / buses / whatever – we need to remember that driverless cars in the US are being produced in a large nation that is not really very urban in terms of its overall usage of land, where its population is quite spread out…”
Google are surprisingly open about their USA car testing. A lot of testing is conducted on suburban roads but the cars have a low max speed limit which prevents them from being tested in driverless mode on freeways, highways, whatever. Testing around Seattle where the roads may be icy or in places where sheep cross the road are the next effective tests. Google aren’t ready to test anything in the USA snow belt. What is that as a percentage of USA land area?
It seems that Google driverless cars work best in comfortable suburbia where Google-type people live; chaufferless trips to the pub for Google-types.
Having slagged off Google, I’ll have a go at European car manufacturers. Mostly, Europeans strive for partial-driverless control. It means that the driver can release control to the car to drive itself if the car interprets that it is all right. OK.
And the driver is expected to take control when the car says “take over”. When should the computer ask the driver to take over? How many seconds may pass between “take over” and thumping the back end of the vehicle ahead, having passed computer controlled braking to a human brake controller?
@Tobie Abel: “In that context, the government is spot on with the Modern Transport Bill. Liberal Democrats should go much further, crafting an industrial policy that puts exactly this sort of future industry at its heart, and simultaneously associating our party with the truly progressive technology that the 21st Century will be built on.”
Nope. Remember, Tobie, that government is supposed to be the servant of people. All of the stuff that you have written disempowers people and disregards what they think. It is all top down policy implementation.
@Phil Beesley
There is a paper at http://cacm.acm.org/magazines/2011/5/107702-the-future-of-microprocessors/fulltext that says that microprocessor PERFORMANCE has increased 1000-fold over the last 20 years – not solely directly from Moore’s law of having more processors per area but indirectly and from other improvements. So a slight elision of facts but the same result.
A million miles is not nearly enough testing and I am sure that Google would be the first to say that – using https://en.wikipedia.org/wiki/List_of_countries_by_traffic-related_death_rate and https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/473850/quarterly-estimates-april-to-june-2015.pdf – would suggest if I have got my maths right that there is approximately 1 person killed or seriously injured (KSI) per 100 million miles driven in the UK – the average incident involves more than 1 person – although I can’t immediately find any figures on that. Although per mile, built up roads are the most dangerous and probably the most complex to negotiate.
On forecasts – obviously we don’t today have any fully working driverless cars so people can only guess. But if I had to guess and if I had to say whether there will be driverless cars and a lot of them in 100 or 150 years time I would say yes – of course we may have moved on from cars then anyway (tele-transportation????) – whether that is 15, 20, 30, 50 or 100 years is difficult to say at the moment but if I was to guess I would put it at 15-30 personally given the current state of google (and other’s) cars and on potential hardware and software developments. And others that have looked more into all the details have said this is more than possible.
On certification, testing and maintenance: These are obviously difficult and complex issues. But they are issues that we engage with at the moment with a product that kills over 1 million people worldwide (but probably saves considerably more lives).
On the role of Government – firstly Government can be obstructive and do the equivalent of having to have someone walk in front of a car with a red flag. Or it can be supportive. And there are as been noted many legal issues to sort out. Clearly if driverless cars were to become anything near widespread there will be deaths and serious injuries and it is a question whether we as a society and nation will throw up our hands in horror or realise that if they are shown to be safer than human driving than that is not too bad.
I would strongly echo Tobie’s view on an imaginative industrial policy.
@Michael: “On certification, testing and maintenance: These are obviously difficult and complex issues.”
Very complex issues, ones that a human being finds challenging. If you break down in the middle of nowhere and one of the car brake mechanisms is faulty, you have a choice whether to proceed cautiously or wait to flag down another car.
Let’s add another factor: a child fell out of a tree, didn’t seem much harm at the time but now there are signs of septicaemia and the child is travelling in your car. Nine times out of ten, you or another person would drive a conventional car, with dodgy brakes, to seek assistance. Would a driverless car continue to work if one of the sensors failed? If there is a special over ride mechanism, how and when should it be used?
The last mile in many endeavours is the most difficult and longest one, it is often said. Completely driverless cars are to be tested in sleet and snow this winter on roads where Google has mapped the environment in advance. Google is still a long way from the last mile at a technological level, on top of which there are sociological and legal adjustments.
@Michael: “On the role of Government – firstly Government can be obstructive and do the equivalent of having to have someone walk in front of a car with a red flag.”
Indeed, and UK government (or rich people) used old carriage laws to prevent early motorists from using common roads or crossing bridges. The Red flag is just a symbol.
But the early French car industry resulted from the fact that they had roads in 190x on which rich people, buyers, could test fast cars. The UK didn’t have the roads.
—
Sometimes we get a bit confused about power interpretation. Who has power and who has knowledge?
At the end of the day, a rich person will tell you the answer.