Menu lock

Cars & traffic

Jun 10, 2015

How long before driverless vehicles take over cities?

Driverless vehicles offer huge potential benefits but it’s unlikely they’ll materialise soon. Besides the regulatory and commercial obstacles, it's still early days for the technology

A Google self-driving care encounters two cyclists (red objects) at night. That retro arcade game look is charming but I imagined in my techno naivete that it'd all be much more...advanced

NYT columnist Joe Nocera is enthusiastic about the potential of driverless cars to eliminate the 1.2 million deaths on roads worldwide each year. In an article republished by Fairfax on the weekend (Google’s driverless cars will make roads safer), Mr Nocera says “the sooner they are a reality, the safer we’ll all be”.

Driverless cars have enormous potential to change the way we travel, at least in theory. They could indeed virtually eliminate road casualties, as well as multiply road capacity by as much as eight times according to some estimates. They could also “create” productive in-vehicle time (including sleep), reduce the size of the national vehicle fleet, and significantly lower the cost of private travel (see Are driverless cars coming?).

But while I think governments should already be thinking about how to deal with them in the future, I don’t think driverless vehicles are going to take over the roads of our cities for a long time yet. I certainly don’t see them having much impact within the 5 year and 10 year time horizons I commonly see cited by the industry.

Most observers seem to think getting the technology right will be the easiest obstacle to overcome; the major problems will be how social institutions adapt to the new technology.

One line of thought argues that the complicated legal and social adaptations required by driverless vehicles, especially during the transition period when there’s a mix of driverless and (much more dangerous) human driven vehicles on the road, will be extraordinarily difficult.

Tyler Cowen notes that almost all States in the US assume the existence of a driver who is legally responsible for the conduct of the vehicle. Politicians will find it difficult, he says, to avoid clamping down on autonomous vehicles in the early years when the inevitable mishap occurs (almost certainly not caused by the driverless vehicle, though).

There could be demands to shut down the cars until just about every problem is solved. The lives saved by the cars would not be as visible as the lives lost, and therefore the law might thwart or delay what could be a very beneficial innovation.

Joe Nocera notes the difficult privacy issues associated with the collection of data. He says consumer advocates are right to press Google – and all the big tech companies – on privacy issues. The profligate use of personal data has become a big concern for many Americans, he says.

Another issue is more straightforward but no less difficult. Even if the technical and socio-legal barriers can be overcome, driverless cars won’t reach a tipping point until they’re affordable.

Building and maintaining maps is currently expensive and in the early years at least the vehicles will be too. While it’s reasonable to think the cost issues will be addressed with scale (and theoretically driverless vehicles can be lighter) that will take time.

But another reason why driverless vehicles might be some time coming goes against the accepted wisdom. Despite what the boosters claim, the technology is actually a long way from being viable. The constraints are lucidly explained in this article published in October last year on Slate, Driving in circles.

The author, Lee Gomes, says Google’s driverless cars can’t yet operate in heavy rain and snow, avoid potholes, find a space in a multi-level garage, or “tell the difference between a big rock and a crumbled-up piece of newspaper” on the road.

As of last month, Google’s 23 Lexus RX450h SUVs have clocked up just over a 1 million miles in autonomous mode on public streets. Sounds very impressive, but Lee Gomes says those streets are the same ones in Mountain View California that’ve been mapped in copious detail and are “driven over and over again”.

Google has developed a clever and expensive 3D mapping system to control the cars; they can’t move without them.

These maps contain the exact three-dimensional location of streetlights, stop signs, crosswalks, lane markings, and every other crucial aspect of a roadway. That might not seem like such a tough job for the company that gave us Google Earth and Google Maps. But the maps necessary for the Google car are an order of magnitude more complicated.

The company’s technology requires that entire cities and nations are mapped in high detail and kept up to date on the fly before driverless cars could be universally deployed. That can very probably be done but it will take time, political support, and money.

Unlike with the early history of the car, driverless cars will have to be technologically near-perfect before they can fully reap the promised benefits. In the meantime, manufacturers will loudly proclaim various technological advances in conventional vehicles, but the theoretical benefits won’t be fully realised until human drivers are completely unnecessary (and probably banned).

I think the most prospective application at this time might be in large trucks because a lot of freight traffic can be confined to a small number of major, limited access routes that can feasibly be mapped, regulated and monitored e.g. interstate transport on motorways. (1)

There’s a separate issue about whether the likely changes wrought by a world of driverless vehicles would all be positive. That’s an interesting and important issue I’ll discuss shortly.


  1. But then there are a number of driverless trains already in operation (e.g. this is just subways), so it can’t be assumed that driverless trucks operating on limited routes would necessarily be competitive in the future.

We recommend

From around the web

Powered by Taboola


Leave a comment

Your email address will not be published. Required fields are marked *

30 thoughts on “How long before driverless vehicles take over cities?

  1. Roger Clifton

    MRJ, I too am inclined to get bloody-minded on behalf of automatic doors. Perhaps they could nudge a sluggard a couple of times, then alert a remote conductor to assess whether to apply (remotely) a sharp tongue or enlist the assistance of other passengers to bundle a drunk out of the doorway.

    For that matter, I rather like the idea of cow-catchers on the front of automated trams. But that’s for another story.

  2. michael r james


    Not true Roger. The driver was always there, not least because opening/closing the doors was never automated and required a human, who happens to sit in the “drivers” cabin.

    Pity, because if one thing annoys me about the LU is the slowness of the trains in stations, with the bloke hanging out of the driver’s cabin, hesitating to close those doors as riders shamble around. Automatic closure on strict (and short) dwell times is what they bloody need. If a few pax are “lost” now or then, well that’s the cost of making the system perform properly 🙂

  3. Roger Clifton

    London’s Victoria line was automatic when it was opened in 1967. The public was alarmed at the concept of a train without a driver so they were promised that a driver would stand there with his hand on a “dead man’s handle“. He had nothing else to do but try to stay awake. Maybe he’s still there.

  4. drsmithy

    A well-mapped city centre might support a driverless taxi fleet. The perimeter of their range would logically consist of train/bus stations and parking lots for more rustic travelling beyond.

    The problem with the mapping isn’t the streets and fixed obstacles, it’s the dynamic things like people and animals.

    Thus serviced, the city centre could then ban non-professional drivers as being too unreliable to tolerate among the autonomous fleet.

    This will just put a massive cost premium on travelling into the city outside of a public transport schedule, unless you’re going to deregulate taxis – in which case there’s no reason to ban private vehicles at all, because if a driverless taxi can do it, a driverless private vehicle can do it. There might be an argument for banning human drivers. But, as above, mapping the streets is the easy part – so it’s likely that dense city centres are the last place we’ll see autonomous vehicles anyway.

  5. michael r james

    #24 wxtre at 4:32 pm

    Headlines are very abbreviated summary of the article. Only cars are mentioned in the article, and I would claim are the only intended subject of the headline. The lede:
    [Driverless vehicles offer huge potential benefits but it’s unlikely they’ll materialise soon.
    I don’t think driverless vehicles are going to take over the roads of our cities for a long time yet.]

    The “they’ll materialize” obviously eliminates driverless trains as they materialised many decades ago. Airport shuttles probably were first but I believe the first proper city Metro that used them was in Lille in 1983 (and with aligned platform doors). It was the precursor to the first line in Paris (#14, 1998, 64m pax) and in 2011-12 Paris’ busiest Metro line #1 which carries 210m pax p.a. So almost 300m people already use such trains in Paris.

  6. drsmithy

    The company’s technology requires that entire cities and nations are mapped in high detail and kept up to date on the fly before driverless cars could be universally deployed. That can very probably be done but it will take time, political support, and money.

    The good thing is that the more autonomous vehicles that hit the road, the smaller this problem becomes, as each vehicle can be trivially fitted out with cameras and a network uplink to keep these maps updated practically in realtime.

    Vehicle autopilots on freeways and other limited access roads should be readily available in about 10 years. I think it will be a lot longer until we have autonomous vehicles in high density areas. Probably more like 20. Be surprised if it’s more than that, though. The biggest impediments are likely to be legal and political.

  7. wxtre

    #23 michael r james

    The articles headline states – How long before driverless vehicles take over cities?

    Vehicles are classified as; wagons, bicycles, motor vehicles (motorcycles, cars, trucks, buses), railed vehicles (trains, trams), watercraft (ships, boats), aircraft and spacecraft. It is relevant.

  8. michael r james

    #22 hk at 10:08 am

    [There already are many cities where short sections are totally automated with computerized central control.]

    Really? Where are these “many cities” with driverless cars?
    I don’t know of any, unless you mean trains but we should not be even discussing driverless trains anymore than we would discuss pilotless drones in an article about driverless cars on ordinary roads.

  9. hk

    How long before driverless vehicles take over cities?
    In short: never.
    However, autonomous vehicle technology will be applied more and more. Richer and better managed cities than Melbourne according to some futurologists will have nearly all PT transport “driverless”. There already are many cities where short sections are totally automated with computerized central control. Several cities have longer sections of the rail based transport system more or less under total electronic control. Whether institutional inertia would ever result in fail-safe platforms and access to trains and trams within the next ten years or so is worth raising in the political arena.
    The aged and impaired can between from technology that limits impact collisions between peds, cyclists and vehicles

  10. michael r james

    #20 AD

    Then why cite the least credible piece of boosterism?

    Anyway as I wrote in #19 (and believe it or not, #17) it is definitely not the point. Adding an extra lane on a two-lane freeway nominally increases capacity 50% yet, as real-life and Braess’ Paradox shows, often leads to more congestion (and not after a long delay). By contrast closing a road can lead to less congestion. We’ve had some recent dramatic demonstrations of this in LA during some major freeway closures during long-overdue works (“carmageddon”).
    Even where some relief occurs it is usually transient. One of the more dramatic examples is when the first Sydney harbour tunnel was built, the bridge it was intended to relieve had the identical congestion only two years after it opened! This obviously had absolutely nothing to do with city growth or any real increase in demand. It’s just what roads do. Convincing everyone that self-driving cars will relieve road problems (by 0.5x, 2x or 8x) will have the same effect.

    FWIW (not a lot on this blog) I will repeat that it is all a massive distraction.

  11. Alan Davies

    michael r james #17; Soctrates #18:

    multiply road capacity by as much as eight times according to some estimates

    I’ve put in the rest of the sentence since you evidently didn’t see it. You can find the source of this claim by following the link I provided in the article (here it is again). For the record, I can’t say it’s wrong but it sounds boosterish to me; even Google is only claiming double.

  12. michael r james

    But, but …. Socrates, AD wrote it in black & white, and may have even cited another economist or some “expert” report. You are a terrible no-good sceptic.
    Then again, these are the same road-lobbyists (not you of course) who continually claim that building new roads and especially adding lanes to existing roads will ease congestion. So you mean to tell naive little me that all these promises of spending countless billions on roads (or self-driving vehicles) won’t relieve congestion?
    I’m crushed.


    Following the logic, either way one cuts it, the only solution to city urban planning is to have effective public transport. We’ll never decongest the roads, no matter how many we build or widen (real-world model: LA or any Sunbelt or Chinese city) and no matter the technology. Indeed the only proven methods to limit road use are PT and road congestion itself (which, funny enough, LA is also the new model).

    Incidentally most existing traffic modelling will not use the trick that self-driving vehicles can achieve: programmed to drive in tight formation with other cars in a packet or quantum. Currently on major roads, it is the traffic lights that tend to force cars into quantum but of course, that, by definition, is undesirable (and when humans are driving such stationary quantum results in very poor co-ordination of “take off” by the whole group). If the cars themselves form such tight quanta then their passage thru junctions can be highly orchestrated to be very efficient. If the packets (of cars) are tight enough and spaced appropriately then the time the lights at a junction are green can be much shorter than today so that the packet can pass thru, on both cross-roads thru the junction as if the junction barely exists.

    Do I think it could work so smoothly as models imply? No.
    Do I think it would ever come to pass in my lifetime? Possible but highly improbable*.
    Do I think it will yield anything like eightfold capacity? No.
    What I think is that it is entirely irrelevant to the real-world solutions that our cities need. It is yet another distraction, one which the road lobby (and our PM) will continue to deploy to argue we shouldn’t spend any money on PT.
    *The best implementation of this fantasy is probably in the movie Minority Report where Tom Cruise gets in his self-driving e-car, speeds along aerial super-freeways, exits and drives up the facade of his apartment building to park so he can step straight into his apartment. As it happens Tom Cruise’s ex, Nicole Kidman has a low-tech version of this: the Starrett- Lehigh Building in NYC has a truck-size elevator that residents can drive their cars into and drive off when they reach their floor, into a “sky-garage” within their apartment.

  13. Socrates

    I am a traffic engineer and support the concept and soon reality of driverless cars. Nevertheless this claim is false:
    [multiply road capacity by as much as eight times]
    No they will not. Even if working perfectly, at urban road speeds (where we have the capacity problems) driver-less cars might double or triple potential maximum car density per lane. The trouble is, it is not the mid-block “link” capacity of roads that determines urban road capacity. The problem is always at intersections and junctions, whether controlled by traffic signals, signs, or priority rules (merges, on-ramps, etc). we might get an improvement of up to 50%, which would be very significant. But I would be amazed if it was better than that. I have never seen any mathematical proof of driverless cars achieving the four-times capacity benefit as a network wide effect.

  14. michael r james

    [multiply road capacity by as much as eight times]

    Ha, if the road lobby believes that and advocates for driverless cars then I will not argue with them. The logical outcome is that we should immediately stop building new roads, especially West Connex and any version of EWL. In fact any new freeways and especially hyper-expensive tunnels. After all, ordinary roads will function like freeways once cars are forced into quantised packages that will mean green lights at every junction–for both directions!–and no waiting almost anywhere. (well, we’ll overlook the parking issue at endpoints).

    This will free up massive funds for building real public transport and cycle & pedestrian routes etc. And the notion of Shared Space will actually work even in backward countries run by rightwing business interests. Melbourne’s trams should be given absolute ROW including green lights so that they might function as planned instead of being the world’s slowest tram network.
    The most important thing is to stop spending on roads immediately as it will obviously be a total waste in just a few years time!

  15. Tom the first and best


    The automatic cars are not yet good enough to detect temporary roadworks traffic lights and presumably police hand signals. They will need to be able to detect temporary roadworks traffic lights and the police will need some way of directing them as well.

    I agree that driverless have serious potential for taxis and buses (and also trams).

    Automatic trains are a lot simpler than automatic cars because they operate on fully segregated tracks (much less prediction and detection required), with externally controlled segregation (signalling) and most of them have platform screen doors as well. They are essentially technologically closer to horizontal lifts. This is why they have been around for decades and more autonomous automatic vehicles have not become available..

  16. Andrew O

    While a driver may be ‘legally responsibly’ for their vehicle, the current Australian practice is to accept ‘Sorry mate, I didn’t see you.’, ‘The sun was in my eyes’ and ‘This is the first person I’ve ever killed.’ as being acceptable excuses for reckless and negligent driving.

    It would be nice if the widespread introduction of autonomous or semi-autonomous vehicles, or even just of advanced safety systems, could demonstrate that routine safe driving IS possible, and lift the expectations for ‘manual’ drivers.

    With higher expectations we could optimistically see higher penalties and when combined with ubiquitous recording, increased caution and decreased fatalities.

  17. pjrob1957

    I have just been talking to a friend who lives near silicon valley and he says its so positive about driverless cars there that it is looking like the next big boon to the economy and the actual economy there is looking up because of it.
    He gives the impression that its to hit us pretty soon.

  18. cbp

    Freeway lanes specifically for driverless cars seems perfectly feasible to me within the not too distant future. Imagine a partitioned lane where driverless cars can travel at much higher speed limits than currently allowed. As you say, this would be of huge benefit for trucking and commuters, making the costs worthwhile.

  19. Socrates

    I agree with Mark Farrall. This tchnology is advancing faster than peopel realise. The test is not when will the technology be perfect – the question is when will it be safer than most human beings. I would argue that it has already reached that point. Look up C-ITS tchnology and you will see it is already being trialled.

    Apart from safety, this technology has great potential in taxis and buses. It is already working well in driverless trains (France, London, Canada to name a few). A third to half the running cost of PT is the driver. Think about what this will do to the economcs of public transport.

  20. wxtre

    Singapore has a driverless rail network. Autonomous railway is being implemented on other networks.

  21. Jacob HSR

    How long before electric cars take over petrol tax indexation?

  22. Joshua Saunders

    MB – I’m in all in favour of alternatives. I just consider that it is highly unlikely that either our politicians or the largely suburbanised population will move quickly to modes of living and public spaces where cars are not the preferred and primary transport option. Small steps.

    In the meantime, unfortunately, cars are king. In which case, I’d prefer to live around competent self-driving vehicles (if that is possible) to jalopies driven by aggressive, incompetent and/or bewildered drivers.

  23. M Bourne

    The problem you describe is a symptom of our car dependant cities. Sure, welcome fully driverless cars when (if) they have the awareness and smarts to navigate a chaotic city centre without defaulting to FULL STOP, but don’t neglect the options that can be built now… ie, we should sustain the push for walkable cities with effective public transport. Don’t be distracted or seduced by what manufacturers claim is “only a few years away”.

  24. Norman Hanscombe

    Driverless cars [and planes/rockets] were a popular prediction at least as long ago as the early 1940s Sunday Newspaper Comics Section, and I wonder why Crikey keeps re-discovering the fad.
    Roger Clifton’s Post #4 raises the only remotely practical suggestion, and even it is unlikely to survive the unavoidable changes to lifestyles which will occur when our cheap non=renewable resources run out.

  25. Joshua Saunders

    Road safety campaigns mean don’t speed, don’t drink drive, even don’t use your mobile (but at least 20% still do and will continue to). True, real driver education, and stressing the absolute need to focus and watch the bloody road whilst driving, and drive defensively and considerately, just doesn’t seem to register for a large number of drivers. And the notion that people should just drive carefully around vulnerable road users such as cyclists or pedestrians (even if they do annoy you) is anathema to many Australian drivers.

    The result is, 1200 deaths per annum. Only ~60% of these deaths are of occupants of cars trucks and buses, and a totally unacceptable ~150-200 pedestrians a year, presumably every single one of whom has been hit by a motorised vehicle. In what other field of human activity would 1200 preventable deaths per annum be an acceptable outcome?

    And whilst the numbers have actually improved significantly over the last 10 years, the raw number of fatalities among those over 60 years of age is actually increasing. But which State or Territory government will be brave enough to make it harder for those suffering the effects of age to keep their licences? None.

    Bring on the self-driving cars. They won’t be perfect. But they will be MUCH better than many Australian drivers.

  26. Roger Clifton

    A largely-autonomous vehicle on a boring stretch of road doesn’t need much attention from a driver and a stationary vehicle doesn’t need a driver at all. A puzzle need only take up human attention when the vehicle alerts him/her to it.

    A roomful of remote drivers could supervise a much larger fleet, taking a sip of coffee between puzzles and pressing the pull-over button when they are too busy.

  27. Roger Clifton

    A well-mapped city centre might support a driverless taxi fleet. The perimeter of their range would logically consist of train/bus stations and parking lots for more rustic travelling beyond.

    Thus serviced, the city centre could then ban non-professional drivers as being too unreliable to tolerate among the autonomous fleet.

  28. Mark Farrall

    The all or nothing thinking in the ‘tipping point’ statement misses the point that driverless cars a combination of tens to hundreds of functional changes and components.

    Many of these components have already been delivered (AEB) and more are being delivered each year (automated highway driving is coming out next year on higher end models).

    So it will take a while for the full benefits of driverless cars to arrive, but many are already here and more are arriving each year.

  29. Jason Murphy

    Hindsight blinds us.

    We know technological change is lightning fast. But we forget that it happens in categories we didn’t predict.

    Every time we make a specific prediction about the future of a certain technology, another technology leap frogs it to become the new sphere of excitement.

    Flying cars, cures for cancer, jet packs, domestic robots, PRT, electric cars, voice recognition and commands, even online shopping. All these are trends we over extrapolated.

    Meanwhile technologies we never expected leapt ahead – smartphone apps, the microbiome, social media, drones.

    We go wrong when we use the pace of development of successful technologies to anticipate the future of the technology we’re excited by.

    In summary, your car will park itself in 2015, but we’ll all be dead before fleets of driverless cars roam our cities solving real tranport problems.

  30. pjrob1957

    Still a puzzle as to how no one has noticed the elephant in the room on this one.
    The safety features are the most significant idea and we can expect a vulnerable road user will be safer from vehicles operated by robots but therein is the problem no one is talking about.
    All it takes for a thief or prankster to stop one is to get in its way.
    Unless they brought in with laws and technology making it difficult to do this, I cant see how it all can work.
    Am I missing something?
    Our roads are dangerous now partly because we recognise an individual operating a vehicle, either by intent or falability, to be capable of running us down and have aquiesced to that reality.
    What will roads be like with thousands of quite carefull robots sharing the space with now casual humans?

Leave a comment

Your email address will not be published. Required fields are marked *