Predictions are hard, especially about the future. And various futurists have managed to give the art of technological prediction a bad name. So it’s highly likely that most of the following post will turn out to be completely wrong. But here goes.
In 2004, 2005, and 2007, technology nerds like myself were enthralled by reports of the DARPA Grand Challenge. A variety of university and corporate teams competed to build cars capable of navigating rough terrain – without a driver. The first competition, in 2004, ended in failure; the most successful vehicle only navigated a fraction of the 240 km course before getting hung up on a rock. But, by 2005, five teams had built vehicles that successfully completed the tough off-road course.
The third challenge, tackled a far more difficult environment – the urban jungle. Rather than the solitary challenge of off-road driving, and static obstacles the vehicles had to navigate an urban environment (actually part of a disused US Air Force base), complete with things that move – pedestrians and other traffic. Despite these added challenges, five teams were again able to build vehicles capable of navigating the course.
Four years on, the technology behind autonomous vehicles continues to improve. Google’s self-driving car is perhaps the most advanced publicly demonstrated project to date, combining laser rangefinding, conventional radars, GPS, video cameras (for traffic light recognition) and a great deal of software. The end result? A system that has completed 1600 kilometre journeys without any human intervention, and hundreds of thousands of kilometres more with only the most occasional human assistance. It’s pretty damned impressive – watch the network news version, or the uncensored, tyre-screeching version.
It’s not just tech upstarts and universities working on this. Right now, you can buy cars that automatically park themselves, follow the speed of the car in front in freeway traffic, or warn you if you drift out of your lane. And, now, Toyota is demonstrating an autonomous Prius at the Tokyo motor show, that will show off “…how the Prius can avoid obstacles, be summoned from a parking garage and park itself”.
From a technological standpoint, the self-driving car appears to be pretty much done. While the Google car’s sensor is pretty ungainly and very expensive, there doesn’t seem any fundamental reason why the sensors couldn’t be made very cheaply once in true mass production. The computing power? Moore’s Law may be sputtering, but it hasn’t quite given up the ghost yet. So I can’t see cost being a barrier.
Nor do I think that fears of horror stories in the media, or endless litigation, will be an issue in the long term. Sure, it’ll be a little while before true driverless operation is legal. But drivers will be relegated to increasingly passive observers, intervening with decreasing frequency as the technology improves. My guess is that crash statistics will soon reveal that automatically-driven cars are far safer than human drivers. As such, it won’t be too long before the legal pressure will be in the other direction.
So…what does this all mean?
At the most obvious level, it means that many – over time, perhaps even most – of the 1300-odd Australians who die every year in car accidents will no longer do so, with a commensurate decline in the even larger number of people with severe injuries and long-term disabilities. Furthermore, it’s likely to lead to a radical restructuring of the road transport industry, which employs 200,000 people. But I think there will be even more fundamental changes in the way we use cars for private transport. If every car in the world comes with its own free chauffeur, it stands to reason that there will be ways to take advantage of it.
Teenagers, the elderly, and many with physical disabilities that prevent them from driving are obvious beneficiaries (notwithstanding the multi-purpose taxi scheme and other taxi subsidies like it). For teenagers – particularly those without convenient access to public transport in regional and rural Australia – it means a kind of freedom that they haven’t had since the horse was the primary means of transport. For the elderly, it will mean that they can continue to get around safely – a key to independent living for longer.
Parking will clearly be revolutionized. It wouldn’t surprise me if long-term parking was completely eliminated from our CBDs, major shopping strips, and the like. If your car can drop you to the front door and then go park itself, the need to park it very close to your destination is eliminated. But then…rather than sitting there, wouldn’t it be more efficient if somebody else could use your car in the meantime? Or, to put it another way, the only reason a taxi is costlier than running a private car is the need to pay for the driver. The economics of driverless taxis are going to be compelling compared to private cars, at least in urban areas. And they will avoid many of the downsides of existing short-term hire and carshare programs – you don’t have to walk to wherever the share car is parked, if the local car is busy, another one can drive itself to your door. And if you are gone for the day, you only need the car for the day if you’ve got stuff in the boot. Not to mention that a lot of garage space can be reclaimed for better use. While relaxing car parking requirements has been a hot-button issue for green-oriented urban planning wonks in the inner suburbs, this kind of thing might make it possible much further out.
So that’s the good news. What about equity? What about the environment?
My view is that it’s likely to be a net positive, in equity terms. I’ve already mentioned the advantages for those physically unable to drive. And, clearly, in absolute terms, it’s likely to make transport cheaper. Equally importantly, at the moment the least well-off can’t affordably get a small amount of road transport. Aside from the fixed costs of registration and insurance, the least well-off are less able to obtain, and pay more for credit if they wish to buy even a cheap vehicle. When pay-as-you-go transport service becomes the default model, even the least well-off can buy small amounts of road transport, without the extra costs of taxis or short-term vehicle rental. No, it’s not likely to be a huge deal one way or the other, but the effects are likely to be in the right direction.
So, what about the environment? For what it’s worth, I think that the effects are on balance likely to be positive, as well.
On the downside, if car travel is more convenient, people may be prepared to do more of it. And the additional shuttling around made possible by autonomous vehicles might be significant (particularly for autonomous freight). But I think that’s likely to be outweighed by several more important factors.
Firstly, if we have fewer cars used more intensively, the not-insubstantial environmental effects of car manufacturing are reduced. It depends a little bit on the location, but constructing your average car results in the release of about 3.5 tonnes of CO2. Much of this from steelmaking – and low-emissions steelmaking is considerably less advanced than clean energy.
Secondly, if we have fewer cars traveling further, fuel-saving (and thus emissions-reducing) technologies become relatively more attractive as energy costs represent a higher proportion of the cost of transport.
Thirdly, if people are hiring the car that they need, when they need it, they will be able to choose the most appropriate vehicle for the job, and only use fuel-intensive vehicles when required. Even range-limited electric autonomous vehicles can be used for urban running, with petrol vehicles reserved for the relatively rare longer trips.
The same also extends to other modes of transport. Once you’ve paid for a car, the incremental cost of driving can be relatively low. But if you’re paying for part of the car every time you drive, alternatives such as public transport or cycling might be a better deal – existing users of carshare schemes drive less than those that own cars. We might still be subsidizing the road network – but, then, would there be such resistance to road pricing if we were paying a fee-for-service for road transport anyway.
The driverless car won’t solve the world’s problems. And there will be generations of holdouts fighting for the right to punt their Tesla Roadsters and Audi E-Trons at speed on the Mansfield-Whitfield road for a while yet. But, on balance, taking driving out of the hands of humans seems that it will be a big win for society, with much further reaching consequences than is presently appreciated. And it’s coming, sooner than you think.
Elsewhere: Melbourne Urbanist weighs in.
As much as I love the idea of an imminent future with driverless cars, I don’t think it will happen at any point in foreseeable future. For at least two decades we have had all of the necessary technology for driverless trains, yet there are none.
Good one Robert. One downside of intensive use is that it may make it harder to go with pure EV because the distances traveled per day will much higher than for the private car. There are some serious advantages though that you haven’t mentioned:
Cars can be much much lighter and consume much less energy since the cars will get their safety from avoiding accidents instead of current systems based on building tanks so that we can survive accidents.
There is potential for reducing congestion substantially by having most cars narrow track vehicles designed for moving one or two passengers only. There are also substantial congestion advantages in being able to “school” cars so they move in a similar fashion to schools of fish (virtual buses) These schools could adjust their speeds so that they can go through intersections without stopping. (By communicating with other schools and the traffic light system.)
I liked this vision of the 2030 car It had some interesting ideas.
I’d say certainly the military will continue to improve automated vehicles. And in the Bladerunner genre, the military, police and presumably rich will ultimately seek flying cars.
Back to reality though, there’s the hydrogen fuel cell cars where the electric car doesn’t make the distance. But I’m not sure if the hire-car will catch on much… might be wrong – I imagine the automated “Johnnycab” will be cheaper.
And no doubt there’ll be a big dip in driving as oil goes above $250 a barrel, but I imagine the growth in the alternatives will be high the given our innate addiction to the private car.
Completely OT, but this reminds me of my favourite roadtrip moment in recent memory.
I mean, are we really looking for a campground near Mt Typo, or is this same mapmaker’s idea of a joke?
#1 John Carney,
Docklands Light Rail in London is driverless, and has been for quite a while. More than a decade I think.
John D: I’m not convinced about narrow-track cars, because you need a critical mass of them to get any advantages.
But you’re on to something with “schooling”. It also has energy efficiency advantages at freeway speeds.
John Carney: I could be horribly wrong, but driverless conventional trains solve a non-problem. Compared to trains, drivers are relatively cheap.
Quokka: Cool. I didn’t know that
Robert: Good point. However I still think any attempt to introduce autonomous cars will meet with massive, albeit irrational, resistance (Quokka’s comment notwithstanding). I hope I’m being overly cynical
Hmmm … I’ve found Mt Typo. Given its location, I may know the people who named it … the Corps did a lot of mapping in central Victoria.
I have not read the full post but i think that you missed this test
http://computervisioncentral.com/content/vislab-autonomous-vehicles-complete-journey-europe-china
You also have to take in
http://inhabitat.com/pipistrel-introduces-the-worlds-most-powerful-electric-airplane/
The G4 is four seater fully electric powered with a range of 480 klms. This was thrown together in a hurry to compete in the nasa/boeing/google sponsored green flight challenge, which it won taking away $1.35 million in prize money. There already exist fully autonomous control systems in use for aircraft.
If we look 50 years forward to where the combined impacts of global warming and peak oil have severely impeded the ability of nations to maintain the full extent of their road systems due to weather damage and economic stress (not to mention the reduction in availabilty of oil tar), then the ability to “hop” between more condensed communities will be a significant advantage.
Joh Carney,
California have already begun to examine the criteria under which fully autonomous vehicles will be registered. Not necessarily that far off.
I’m with JohnD on the narrow track vehicles.
I’m ready to buy VW’s 1000 klm per litre two seater as a commuter vehicle should it become available. Running this vehicle on bio diesel will sooth my conscience for some time forward.
The other vehicles from the VW stable, the Milano, the Bulli, and the most recent 2015 eT
http://www.goauto.com.au/mellor/mellor.nsf/story2/5699DE25E689A177CA25794F00236116
Read the, so far released, specs on this vehicle, particularly the autonomous features. This is really cool. VW are pulling no punches in what the future will hold.
Oops…. 100klm per litre VW 2 seater. A 200 litre drum of bio diesel will last the full year.
Robert,
I may be wrong, but my understanding is that most of the cost of taxis is in the “plates” (ie scarcity value of the licences) rather than the drivers.
In other words, don’t put it past the ability of the State governments to stuff up the economics of the driverless car too.
If recent experience on the F2 is anything to go by then most vehicles already are driverless. Driverless cars? How about carless drivers?
EV’s will be series hybrid with a small highly efficient free piston engine that “trickle” charges the battery at about the average energy use rate.
This engine will always run at full load and at the most efficient part of its efficiency curve. When the battery is fully charged it will shut down.
Fuel will be ammonia, or biodiesel; lots of options here.
Vehicle can be used as an energy store for the home.
In the more apocalyptic vision of the future the entire vehicle becomes a self supporting self driving residence that transports one from the trailer park to the factory and then drops the kids and the missus off at the workhouse.
Huggy
I & U – I think you’re partially right. The state governments restrict the number of taxi plates to make it economical for taxi drivers and/or to provide extra services such as accessible taxis. We exchange higher taxi prices for regulation, price certainty and having taxis around during non peak periods where its a lot less profitable.
Wiki has quite an extensive list of driverless train systems around the world: http://en.wikipedia.org/wiki/List_of_driverless_trains
I do agree with previous comments that this doesn’t have a big significance for the feasibility or desirability of driverless cars.
Chris,
I don’t see how restricting the number of taxi licences achieves any of the objectives you refer to. Even if it did, one could instead place explicit obligations in the licence conditions (eg that accessible and off-peak taxis are made available) and issue licences on request.
Conversely, if restrictions on taxi numbers are necessary to achieve policy objectives, the same would be true of driverless taxis and the economic benefits that Robert is anticipating might not be realisable.
Rio Tinto are getting on with driverless trains in the pilbara.
I will be a bit atavistic for a while I think. Not due to my nature, just my circumstance. I drive a reasonably long distance along country roads (sometimes poor and highly variable quality) most days. The economic contract that makes car share work would have to pretty bloody sophisticated to deal with my situation. And the computers and sensors I suspect aren’t going to be optimised for my road conditions – dirt roads, debris on the roads, potholes that have to be avoided, rather than other traffic and well-formed intersections. The electric car will have to have at least 250 km reliable range before I’m even going to think about it, or quick swap batteries available everywhere.
Maybe I’ll get an electric motorbike instead. Until they get banned.
I do reckon in 50years time, people will tell kids that vehicles where manually controlled, and those kids will recoil in horror “no way, the number of accidents, the injuries, why”
When the cars can negotiate between themselves, or be controlled at a switch say – then cars will become analogous to packets on a network and will be so much more efficient to get about.
Can’t happen soon enough
There’s some talk of placing recharge points along major roads Wilful. Just by passing over (or perhaps past) the point your battery would get a minor boost. One can imagine driving on tollways and main connecting roads as being a little like those games on your PC where you get extra health for passing certain points. Imagine sitting in a queue of traffic at the lights while your battery got a charge.
Wilful, fair enough. And I can’t imagine cow cockies using a carshare service any time soon!
Fran, we have major potholes that you cannot drive through and have to swerrve significantly to miss, that don’t get repaired for more than six months.
I don’t think the cost of installing and connecting some whiz-bang recharging system on the roads anywhere near us is going to happen any time soon.
We don’t even get ADSL 1 on our telephone exchange.
And I am not back of Bourke here, I’m only 120 km from Melbourne.
That may well be so Wilful but if they were on your multi-lane roads, perhaps the repair rate would be different?
Bilb @ 12: The winners of fuel economy competitions could manage your annual driving with a 6 litre bottle of fuel.
Wilful @19: Autonomous or near autonomous vehicles have a lot of attractions for country driving. Think infra red roo detectors for a start. Then systems that keep the car on the road when the driver nods off. Systems with solar charged radios that tell the car what the last car through found plus GPS systems that slow the car down when it approaches a level crossing…. What would be on your wish list?
Autonomuos vehicles on your kind of road, Wilful, will have the advantage of remembering the road condition, receiving updated information, gps accuracy, coupled with all of the surface observation systems that have been developed. The Europe to China trek by VisLabs encountered everything that you are describing and more. The entire program team was available and many software updates were made enroute, I recal.
If you watch some of the YouTube videos of manouverability control…and repeatability… you will be eased in your concerns, I believe. These vehicles can do the James Bond reverse park spot on every time.
Where the concerns would be are with the new overlay of technical complexity which will be a boon for the NRMA tow home service.
The first wave of introduction for these vehicles will be for the service industries, and the luxury end of the market, which I think Robert covered. But just thinking about the impact of that having been a lift mechanic in another life, it is truly awsome.
At the other end though I am planning my future to be the minimum in road transport (VW 2 seater) and PAV’s (personal air vehicles) operating from micro air access terminals.
http://io9.com/5714918/nasa-envisions-neighborhood-micro+airports-to-let-travelers-bypass-pesky-streets
don’t get me wrong guys, I’m all in favour of it. I’d prefer to telecommute more, and to read a book or do some work while being transported.
Sometimes I like to drive, but that’s mostly when I’m being a little bit hoony and risky and probably shouldn’t be enjoying myself!
I’m just slightly more skeptical that the car will know when to slow down because the old wombat lives around this corner and I’ve seen him crossing this point of the road (on a blind corner, the dea old fellow) several times. Or it was windy last night so there’s going to be some tree limbs over the road.
Nothing that wont be resolved eventually, I’m sure. And when I’m on the Monash Fwy, I put the cruise control on and would be well ready for autonomous driving.
The idea of driverless cars raise massive possibilities for the extension of commuting distances well beyond the urban fringe of major cities. What if you can jump in the car at 5am, sleep for a few more hours, get dressed and have breakfast on the road? Will we see driverless cars with airplane style reclining bed/seats?
Fran @ 21
If you allow a generous 1 km for the ” Just by passing over (or perhaps past) the point your battery would get a minor boost.” and a vehicle speed of 60 km/h,your (presumably) inductive charging loop and allowing that a “minor boost” would be 1 kWh, the energy transfer poerer level would be 60 kW. Put 10 vehicles on the 1 km and we get 600 kW.
In your “or perhaps past” scenario the energy transfer power level would have to be in excess of 600 MW*** !! Maybe we could install a nuclear reactor on every 100 km of so of freeway??.
*** inductive energy transfer at this level would most likely destroy the vehicle and kill the driver – but that is another story.
People should remember that when you put petrol into your tank you are transferring energy at a power level of about 6MW. Hard to beat.
Huggy
I wasn’t considering anything like that kind of range Huuggy. When I began reading about these things — and it was still very much at the prototype stage, people were talking about distances well under under 1 metre. Also, the amount one would deliver would be dependent on how long you rested there.
Accordingly, an EV resting at a traffic light might get enough for get the vehicle from zero to road speed and 500 metres down the road. On a major road (like the M2 where the road speed was 90kmh) and the boost points were every 500 metres, each vehicle might only get enough to maintain forward motion between the points but a 20km journey might leave the battery no more depleted than it was at the start.
Given that most people will do fewer than 60 km per day, and extension by this means to a range of perhaps 100km per day with scope to do inter-urban travel (I’m thinking of the occasional trip to the Central Coast/Newcastle, Blue Mountains and so forth) would be handy. If the system had the integrity to take such a vehicle all the way to Brisbane or Melbourne (perhaps with only one or two stops for proper battery swaps) then almost anyone could use such a vehicle.
Papermind: unconvinced, myself – at least to the extremities you’ve put the argument.
Aside from the unavoidable accelerations you’d be subjected to interfering with your sleep, you’d still be separated from your family for unacceptably long periods.
Sounds great, would free up the police to investigate and solve crime.
But think of the revenue lost to them.
Maybe the elimination of ” driver crime ” would see the number of police needed, halved..
Think of the state budget savings.
But think of the job losses.
My head hurts , gunna mow the lawn now.
As a cyclist these autonomous cars will be a godsend. The current road laws are great for cyclists, but the problem is that most motorists are ignorant of them, or intentionally disregard them.
With autonomous cars I’ll be able to confidently ride using an entire lane at whatever speed I like, even a leisurely 10 km/h, knowing that autonomous vehicles will respect the law and not overtake until safe to do so, and give me adequate clearance, even changing wholly into another lane when on a multi-lane road, as the law demands.
Imagine riding peacefully like that without the fear of road rage!
This could be a great boon to cycling numbers, to the extent that roads may become bike-dominated rather than car-dominated. It could spell the death of mass car-use as cars become practically limited to the max speed of bikes, yet without bikes’ ability to filter through stopped traffic.
Simon,
“With autonomous cars I’ll be able to confidently ride using an entire lane at whatever speed I like,”
No you won’t. The Huggy “cyclehunter” (TM) will hunt you down, pick you up from your bike with giant robotic arms and deposit you safely on the side of the road.
If you are wearing Lycra, it will tenderly tear it from your body and stuff it into its Lycra fuelled power plant.
We will launch thousands of these machines as part of our program to clear the roads for our massive 6 wheel drive all terrain pizza delivery vehicles.
Huggy
Simon, at the risk of thread derailment, I will actively assist Huggy in the development of the Cyclehunter if cyclists take the opportunity to participate in mass abuse of the law as you’re implicitly proposing.
I & U @ 18 – taxi plates for accessible taxis are heavily discounted to take into account that they may be less profitable and cost more to setup. If there were unlimited taxi plates for little to no cost there’s a lot less incentive to make extra upfront capital investment.
Given that many people just drive to and from work, potentially during off peak periods, everyone who wants could make it so their car could be a taxi while they’re not using it? Or perhaps people would have long term leases on people’s cars during non peak periods.
I think it would be feasible for a lot more families to drop down to one car too. The car could drive itself between the users of the car (one of the big problems with sharing cars at the moment is moving the car itself).
Simon.
To counter the malevolence of thousands of Huggy ‘cyclehunters’ I am working on a prototype of the riderless bicycle or ‘lane drone’ that will lure his ‘Max-imus machine’ into an IED – immediate exit by drone- otherwise known as the ‘afghani clinch’. Looking for preferably Lycra free riders to test the prototype…
Pablo: we have the prototype all ready for you.
Dam it Robert. It appears Mr Yamaguchi has beaten me to it.
Back to the drawing board.
So, the invention of the driverless car automatically transforms one-hundred-years-plus of horseless carriage ownership into either a golden age of rational, selfless co-op shared ownership of personal transportation, _or_ it means the demise of non-commercial ownership of said transportation because the taxi companies have the consumer’s back, what with nobody ever again having to pay for another one of those price gouging cab drivers.
Because? Technological determinism is going to just overturn private car ownership/automobile industry capitalism how? Driverless tech will take cars off the road more successfully than, say, the adoption of Northern European/Japanese style mass transit?
Robert, this is not going to happen the way you dream it will. It can’t. Making it possible for non-drivers to steward their own motor car, all alone, is not something an environmentalist policy wonk should be building up into the fantasy solution to either fix gridlock or rationalise (downsize) the effects of consumerism.
(And this is even granting the ability of driverless tech to take off the way flying cars never did, pardon the pun.)
Wilful @27: If you slow down when you know a wombat may be around the corner you are probably driving too fast to deal with the unexpected wombat around other corners.
I think autonomous cars will develop from semi-autonomous cars. For example, we might start with simple systems that do things like control how close we get to the car in front, stop us going above the speed limit unless we push the accelerator very hard and protect Wilful from wombats.
Then we we might automatics that do things like control lane changes, avoid potential collisions etc. Then at some point we might start allowing fully autonomous on freeways then….
Much of the thinking about cars start with someone asking “what can we do make the cars we know and love better? It may be more productive at times to go to different starting points.
For example, consider electric bikes. They weigh about 25 kg, have a range of approx. 30 km, max motor power of 0.2 kW and max speed about 30 km/h. They also allow you to decide how much, if any, exercize you want to get during a trip. Downsides include exposure to the weather, negligible protection in the case of accidents and……
What might we do to make electric bikes better suited to urbane transport?
The problem with your 42, JohnD, is that someone is developing a machine to devour the clothing of the riders of small minimalistic and slower road vehicles. Why?? Well aparently, as some whackey form of climate change adaption, it is to force people to stay permanently at home using their NBN connection for work, and everything else, while buying home delivery pizza indefinitely for sustenance.
It won’t work of course. Because no sooner had this been announced when a major supermarket chain launched the proposal to build a network of under ground tubes through which people could travel at collosal speed in vacuum powered pods.
And all you wanted to do is ride casually in the sunshine. Seems it will never happen.
Chris,
“everyone who wants could make it so their car could be a taxi while they’re not using it?”
Yes, but you could equally say that, at present, everybody with a car could be a taxi driver whenever they have nothing better to do. But they can’t, because the State prohibits it: apparently for some arcane reason to do with subsidising accessible taxis, which I don’t really understand or believe.
As NickWS alludes to, the existing role of the motor car often has more to do with cultural, regulatory or commercial constraints than technological ones. One cannot assume that these constraints will be removed just because of technological advance.
There is one major flaw withthe car drives itself idea:
Some really demented and insane person or group will infect millions of the vehicles with a virus or a worm that would set up the (in its most benign form) a massive grid lock that would take forever to untangle.
Huggy
“What might we do to make electric bikes better suited to urbane transport? … ”
Well, perhaps we’d have to start by polishing their manners, and ensuring they cultivate a suave demeanour … … “
I&U: Some of the barriers to better people transport are cultural. Unfortunately a person’s form of transport is associated with status, masculinity, sex appeal etc. “up came the squatter mounted on his thoroughbred…”
On the other hand, the development of better people transport is inhibited by a failure of regulation to take account of changing technology let alone encourage potential developments. Existing regulations are often irrational. For example, call something a motor bike and a superbike that offers no protection in a crash is OK. Call it a car and there is a whole raft of requirements aimed at crash survival with no recognition crash avoidance technology.
Our regulators should be looking hard at what could be done to encourage the development of light weight, low energy requirement vehicles.
Phillip @46: There is nothing urbane about a spell checker.
I & U – on reflection I think you’re probably right about taxis. I think it would have an impact on the feasibility of car sharing schemes though.
John D – I think the if the rate of motorcycle usage was as high as car usage then there would be a lot more work put into trying to make them safer (or the environment safer for their use). It’ll be interesting to see if there are safety changes related to big increase in the use of scooters. In the meantime it should help the donor organ shortage problem.
Chris: It would be interesting to know to what extent unsafe riding, falling over during skids (because they only have two wheels) and lack of crash protection contribute to the high level of motor bike related injuries and deaths. It would also be useful to know how often motor bikes avoid accidents because they are more maneuverable and fit through tighter spaces. The answers might have some bearing on narrow track vehicle design and the potential safety of light weight cars.
I read some time ago that large cars had more accidents than small cars but the occupants had a greater chance of surviving the crash.
Can’t happen, won’t happen…
Not, that is, until much lawyers’ blood is spilled resolving the question of who is to blame when something goes wrong.
Driver? Nope – The autopilot was engaged.
Manufacturer? Nope – The vehicle was designed in accordance with the appropriate ADR’s (Australian Driverless Rules) and, besides, it was sold months ago.
Designer? Nope – When was the last time an aauto engineer was pilloried because the car he designed was a pig to steer on a wet road? Multiply this by a thousand.
At present, every time there is an accident, the police have the task of assigning blame. This requires that a real, live human being be found guilty of breaking a law.
How can a car get insurance?
Perhaps the car itself will spend time in the slammer.
Let’s see how far California (#10) gets before we get too excited.
John B: What a misery guts. Perhaps anyone who suffers a a result of accidents involving cars driven by humans could seek damages from the lawyers that blocked the introduction of safer autonomous vehicles?
never say never, John Bennetts,
http://www.smartplanet.com/blog/smart-takes/whos-to-blame-when-an-autonomous-vehicle-crashes/18146
John, “autonomous” vehicles will, for the first few years of their existence, be like planes with autopilots – routine operation by the computer, with oversight by a human.
Many of us already use autonomous control in varying degrees.
Hands up those who use the cruise control on their car?
I do and I love it. I would like it even more if it had the vehicle spacing control available on some high end vehicles. Add ABS braking and I am feeling safer again. I would like it even further if it was GPS linked for road speed limit information.
Now here is the gap. Steering control.
Jump forward. We are seeing a number of cars now offering auto parking.
So we already happily hand some control to automatic systems. I suspect that we will progressively hand more control bit by bit towards autonomous control as tese become available in mainstream cars, and as we become comfortable with their presence. And the legal process will largely follow the same route, find its way small incident at a time establishing the pathway of precedence as confidence increases.
An after thought on the “who is at fault”.
In NSW accidents are judged on the situation. If a vehicle crosses a median strip and “head-ons” a vehicle on the other side then regardless of who or what was in control the responsibility is clear.
If a car rear ends another, again the responsibility is clearly laid out in the law. Who pays? well insurance companies have a “you cover yours and I will cover mine convention” where vehicles are equally insured, and loss recovery is a different process. I believe that there will be a “who benfits” test where there is an autonomous vehicle failure, as a primary criterion. If this is backed up with suitable compulsory insurance then most situations will be covered. A commenter to another article makes the point that AV’s are likely to have, and cancertainly be required to have, quite specific data logging capabilities. This will simplify the “what happened here” debate.
There was a very good comment to this article
http://www.scientificamerican.com/article.cfm?id=google-driverless-robot-car
Which I have copied here
“Suppose 10 human-generated fatalities are replaced with five robot-generated fatalities, is that an ethical trade that society wants to make?” that is a very interesting comment. Ethically, it is a reasonable trade-off but people will see it differently. People will see those deaths as additional as though the number of people killed has increased. It is similar with medicines. A medicine may help a large number of people but if a small number of people have bad reactions to the drug, the general public sees those problems as additional issues and not part of the checks and balances. It is as though the drug has only caused the bad reactions and not helped anyone. The anti-vaccine and naturopath crowds are famous for this kind of reasoning and I suspect driverless planes, trains and automobiles will experience the same thing. The general public has no idea of the number of people killed by automobile accidents. But the first time someone is killed in an autonomous car everyone will know and autonomous cars will be credited with one kill not one million lives saved. The real challenge as always is not in making smart cars, but making smart people.
Robert@6:
We already have facilities that could work for them in many parts of Melbourne. They’re called bike lanes. What stops most motorists from using them is that they don’t know where the left hand side of their vehicle is. The bold ones drive in the bike lane until they reach an obstruction (usually a parked car), at which point they block the lane and wait. If you put those motorists into a single-width vehicle where both front corners were visible they would have no problem using many bike lanes.
A pragmatic approach for narrow vehicle introduction would IMO be to allow them as motor vehicles initially then if they take off start splitting inner urban multilane roads so that the outermost lane becomes two narrow lanes – a Copenhagen lane for bikes, a kurb-style barrier, then a narrow motor vehicle lane only available to motorbikes and NVs. Occasional width restrictions in that lane would discourage oversize vehciles from trespassing.
That approach would probably get bicycle groups on your side, as many of them are partly environmentally focussed.
For a really radical approach, require that autonomous vehicles be less than (say) 1.5m wide and weigh no more than 200kg per passenger unladen in order to get approved. Would have to be global, but it would solve a bunch of problems.
Robert’s right Moz,
We need a critical mass of them before it is even slightly sensible in making changes to how our roads work. Look at motor bikes, for instance. For all of their numbers the only accomodation for them is in parking arrangements, best in Melbourne…pathetic every where else. People have to demonstrate that they are prepared to travel in that manner first.
BilB, that’s why you let NV’s sell themselves just as cars first.
As for acommodating motorbikes, I’m not sure whether we see the same things on the road. What I see is motorbike users habitually using bike lanes (even more than cars), lane splitting and jumping lights. Having bike lanes and wide motor lanes facilititates those behaviours and does get explicitly mentioned by motor traffic engineers when they’re doing “community conslutation”.
It’s important to consider “what does happen” as well as “what is lawful”, since it’s very rare for road users to obey the law (or even know the law).
For all of their numbers the only accommodation for them is in parking arrangements, best in Melbourne…pathetic every where else
BilB, if parking officers and councils would just unclench, parking for motorbikes really isn’t a problem, it’s pretty easy to put them out of the way on the footpath, almost everywhere.
But wire rope barriers , great for cars, deadly for motorcyclists.
Just so that we know what we are really talking about. The question is, will enough people be able to pack their ego’s into this car
http://en.wikipedia.org/wiki/File:1-Liter-VW_(525150348).jpg
in place of this
http://www.free-extras.com/images/mitsubishi_suv-6946.htm
Bilb: If you define narrow track as vehicles that can safely drive two abreast in a single lane you don’t need many before congestion and hold-ups will start to be reduced as a result of narrow tracks driving side by side and, like motor bikes, continue to move by taking advantage of narrow gaps in stopped traffic etc.
It would be good if innovative governments started setting the laws required to allow advantage to be taken of narrow track vehicles now.
The normal cruise control system (that controls the minimum speed) is OK for roads where you can safely stay on the speed limit most of the time. However, it is not much help for roads where you need to change speed often. It is also has the potential to cause more serious accidents when people nod off on a long drive.
Cruise control would be more useful if it was a semi-automatic system that put an upper limit on speed. (The driver would control speed up to the limit and would only be able to go above the limit in emergencies by pushing much harder than normal on the accelerator.) Such a system would avoid the problems of the normal system mentioned above while reducing the risk of exceeding speed limits. I could imagine such as system evolving to include braking when driving down hills, controlling distance behind the vehicle in front etc. etc. At some point we would reach a point where it would be safe to put speed control back into auto.
BilB @ 60 – people don’t buy cars (especially the first car for a household) for what they commonly use it for, they buy it for what they use it for sometimes or even what they *might* use it for.
So although most of the time its just one person in the car, they buy a car big enough to carry parents, a couple of kids and a friend or two. That’s unlikely to change unless people simply can’t afford to do so. For second cars which are just used for commutes I think there’s a lot of potential for change, especially if the smaller cars are a lot cheaper.
Chris,
Agreed. And a second car is very expensive, what with tax, insurance and parking costs, even before you factor in the cost of the car itself.
My long-term bet is plug-in hybrids. They are, effectively, electric vehicles for daily commutes and petrol vehicles for the exceptions like holidays and emergencies. So you get two cars for the price of one and a bit.
Only slightly apropos the theme of this thread, I found this to be an interesting essay: http://www.lrb.co.uk/v31/n11/andrew-ohagan/a-car-of-ones-own
John D,
The GPS that came with my android phone running “route66″ software has speed measurement and road speed limits fully functioning. And it is quite accurate, tested on the return drive Sydney to Temora. So road speed changes are not a problem. With the broad band radar integrated systems now available for high end vehicles a car in a flow will naturally follow the speed of other cars to the limit set with in the road maps as monitored by the GPS satelites. The route66 display also shows by way of a circle of varying diameter aroung the displayed GPS position the accuracy of the calculation achieved from the satelites.
All of the features you are suggesting are already off the shelf, and more.
Chris,
Mostly true.
Upon leaving school and strating work most people in our world acquire a small vehicle which becomes their first. Then comes the family and the scenario that you talk of.
In my experience it is the commute vehicle that does the highest mileage, and consequently the highest environmental damage. However with the commute vehicle being the vw 2 seater it will become the family car that is the least environment friendly.
I was just discussing with a friend the possibility that we could buy canola oil from Woolworths retail, convert it to biodiesel to power the 2 seater and still come out cash ahead compared to a conventional petrol powered small car running cost.
Bilb: Does your GPS actually control speed including adjusting speed to take account of bends etc.?
What I really want right now is something that at least stops me going above a set limit while allowing me to go slower if appropriate. Who are the suppliers?
Have a look here, John,
http://en.wikipedia.org/wiki/Autonomous_cruise_control_system
then it is a matter of sifting through the manufacturer’s information.
Also look here
http://www.gizmag.com/sentience-adaptive-intelligent-cruise-control-driverless-car/11223/