Facebook Google Menu Linkedin lock Pinterest Search Twitter

Advertisement

Transport - general

Jun 30, 2015

Would a world of driverless cars be all beer and skittles?

A speculation on some of the wider and less commonly cited changes - not all positive - that might happen in a would where autonomous vehicles have replaced human controlled vehicles

Share

Driverless cars will reduce the cost of travel by allowing passengers to use the time in more productive ways (source: The Economist via amor mundi)

A world of autonomous vehicles promises many benefits, including the virtual elimination of accidents, the opportunity to use travel time in more productive ways, and at least a doubling in the number of vehicles the existing road system can accommodate.

Most of the commonly cited potential benefits relate directly to the road system. There’s less discussion about the potential consequences of a fully driverless world in terms of the wider impact.

Here’s my speculation on what some of the possible implications might be in the event that autonomous vehicles fully replace human-controlled vehicles.

Industry changes

I can say with confidence that the job of ‘driver’ will be redundant. Relatively unskilled work driving taxis, delivery vans, trucks, and buses will disappear.

With close to zero road casualties – both major and minor – many jobs in ancillary industries like health, insurance, crash repair and policing won’t be needed.

A smaller total fleet size, lighter vehicles and higher road utilisation will likely reduce jobs in vehicle manufacturing and servicing as well as in road construction and maintenance.

These losses should be offset to some extent by jobs generated by more travel to activities (see below) and by avoided casualties. I expect there’ll be jobs in industries offering services and products to fill the in-vehicle time released from driving and now available for other purposes.

Land use changes

Time is the largest cost of travel; autonomous vehicles should make private travel dramatically cheaper by reducing the opportunity cost of driving. Rather than devoting all their attention to controlling the vehicle, drivers will be able to use in-vehicle time for productive purposes e.g. work, sleep. The design of vehicles will adapt to the new opportunities.

This will go beyond the activities that train and bus passengers can currently perform because occupants of an autonomous vehicle will be able to use the greater privacy and space for activities that aren’t feasible (or in some cases permissible) on public transport.

One consequence is that some people will choose to live further from key destinations e.g. they’ll live further from the city centre or they’ll take jobs further from home. Cities will become more dispersed, with more emphasis on commuting from country centres.

If travel costs are low, more firms might join or form clusters so they can benefit from agglomeration economies. In that event, the CBD and suburban job centres should get bigger while many jobs in dispersed suburban locations (the majority in Melbourne) should migrate to suburban centres.

There might also be a greater tendency for activity centres to specialise in only one or two industries. Rather than suffer the downsides of agglomeration associated with large diversified centres (e.g. congestion, higher rents), firms might selectively co-locate with other firms in complementary industries.

More travel

A key promise of autonomous vehicles is that they’ll extract more capacity from the existing road system. Some claim capacity will increase up to eightfold, but most think a doubling is more plausible.

Given other factors that have reduced per capita travel over the last eight years or so, the additional capacity released by driverless cars might be enough to put off new road construction for the foreseeable future.

The counter view is that if roads are less congested there should be more travel due to induced demand. People who can’t drive at present – like children and the disabled – will be able to make trips they don’t make at present.

Another factor that could increase travel is the liklihood that autonomous vehicles will be used as delivery vans. They’ll be sent empty to pick up goods from outlets developed specifically to serve cars without human intervention.

This could lead to more trips that currently aren’t made (“I can’t be bothered”) and to the deterioration of disciplines like trip chaining. People might dispatch a car for multiple trips rather than make the effort to get everything in the one journey.

A possible corollary is that the weakening in the benefits of complementary shopping for goods will lessen the attractiveness of enclosed shopping centres (malls); strip centres that rely more on services should be less affected.

Public transport

High capacity passenger systems will still be necessary to provide access to dense locations where autonomous vehicles would be handicapped by congestion. Greater concentration of activities in centres should reinforce the continuing need for public transport.

Autonomous vehicles should improve public transport by replacing circuitous feeder buses; in some cases they might also replace walking to local stops.

The sophisticated technology required for driverless cars could find new applications in public transport. Hybrid driverless vehicles could collect passengers from street stops (like a bus) and use shared rail lines to provide a relatively fast trip over longer distances (like a train, albeit a small one).

Many of the technical problems autonomous vehicles have to overcome would be simplified on a rail line, enabling a very large increase in capacity. Hybrid vehicles would still be limited by the speed of the fastest stopper though unless express tracks were built.

Actions

The liklihood of greater sprawl is one of the key implications of universal use of autonomous vehicles. As I’ve noted before, that might not necessarily be all bad news: if access to driverless cars is universal; if the size of the vehicle fleet is much smaller than at present; if cars are vastly more fuel-efficient, environmentally friendly and cheaper to operate; then the negative implications of sprawl would be significantly reduced.

Nevertheless, policy-makers concerned about other impacts of sprawl could discourage outward expansion and regional travel by putting a price on road space. That would also make people more conscious of using autonomous vehicles for low value (i.e. unnecessary) trips.

Advertisement

We recommend

From around the web

Powered by Taboola

23 comments

Leave a comment

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

23 thoughts on “Would a world of driverless cars be all beer and skittles?

  1. drsmithy

    “Decentralise” is a pretty broad term, what do you actually mean by that?

    More cities with less people rather than a handful with lots of people. Less people and/or more space -> less congestion.

    And they add significant overall trip time because far more money has been poured into roads and parking than trains and busses for nearly a century.

    No, they add to trip time because bus stops are rarely exactly at your source or destination, and unless you live on the same line as your destination(s), you need to change vehicles or modes.

  2. M Bourne

    21
    “Decentralise” is a pretty broad term, what do you actually mean by that?

    “people don’t use buses, trams and bicycles today because they add significant overall trip time.”
    And they add significant overall trip time because far more money has been poured into roads and parking than trains and busses for nearly a century. Time to remedy the situation, not continue with the same mistakes.

  3. drsmithy

    The London congestion charge started in 2003. Cameras are much cheaper now.

    Cameras always were cheap. The expensive part is putting them in and making them useful.

    There should be buses or trams on main roads. As well as cycleways along main roads.

    Firstly, please don’t misquote me.

    Secondly, people don’t use buses, trams and bicycles today because they add significant overall trip time.

    Thirdly, you haven’t actually addressed the point, which is that most people have little to no control over when they have to travel. Thus, taking steps taken to punish them based on when they’re travelling are unreasonable and unfair.

    A female that worked with me on St Kilda Rd used to drive to work. I also used to work in Port Melbourne and take the tram to work. While some of my colleagues used to drive in.

    And ? I’ve also driven, ridden a pushbike, taken a tram, taken a train, taken a bus and simply walked to work throughout my life, depending on where I was living at the time and what made the most sense.

    So if parking on St Kilda Rd was properly priced…

    By which you mean, of course, “too expensive for all but the rich to afford it”. Because if it’s one thing this country needs it’s yet another thing to make the astronomically high cost of living even higher.

    I do find it odd that someone who can barely disguise a desire to ban cars is worried about “odometer based discrimination”. Seems like what’s good for the goose is not OK for the gander.

    Here’s a better idea: decentralise. Less people means less congestion. Less enforced density means better quality of life. We have no reason in a modern world to pursue the absurd anachronistic idea that people need to trek into the CBD each morning and home again each afternoon.

  4. Jacob HSR

    drsmithy #19,

    The London congestion charge started in 2003. Cameras are much cheaper now.

    “people have no control over when they are on main roads”?

    There should be buses or trams on main roads. As well as cycleways along main roads.

    A female that worked with me on St Kilda Rd used to drive to work. I also used to work in Port Melbourne and take the tram to work. While some of my colleagues used to drive in.

    So if parking on St Kilda Rd was properly priced…

  5. drsmithy

    So maybe Alan should say “electric cars will make redundant petrol tax indexation” rather than cheer on petrol tax rises.

    Indexation should never have been removed from petrol taxes. It was transparent vote-buying by Howard. Supporting re-indexation is not “cheering on tax rises”, it’s sound economics.

    As for odometer based discrimination. No thanks.

    If I drive a lot on private property, eg, farm, golf course, race track, I will get overcharged.

    Not hard to put in an exception for rural dwellers and farm vehicles that spend the majority of their time off road. Such vehicles are a tiny minority.

    If you’re only spending a relatively minor portion offroad, then suck it up. Taxes can’t be tailored for every use case. No different to fuel taxation in that regard anyway.

    Set up cameras on main roads to monitor number plates and charge people for driving on such roads during the peak periods. (like the London Congestion Charge).

    And then you’re discriminating against all the people who have little to no control over when they have to be on the road – a far, far greater percentage (nearly everyone), and much more skewed to lower income earners – than those who drive significant distances on private property.

    Not to mention the substantial infrastructure cost in putting cameras everywhere.

    GPS trackers in cars are another alternative, of course, and loved by the authoritarian types. But it’ll only take about five minutes after those are legislated for things like automated speeding and parking fines to be implemented – and that’s before going into the gross privacy invasion of having your location tracked by Government.

  6. Jacob HSR

    drsmithy #17,

    So maybe Alan should say “electric cars will make redundant petrol tax indexation” rather than cheer on petrol tax rises.

    As for odometer based discrimination. No thanks.

    If I drive a lot on private property, eg, farm, golf course, race track, I will get overcharged.

    Set up cameras on main roads to monitor number plates and charge people for driving on such roads during the peak periods. (like the London Congestion Charge).

  7. drsmithy

    Electric cars and hybrid cars have grave implications for petrol tax indexation.

    Trivially solved by adding in an equivalent distance-based charge from a simple odometer reading as part of registration.

  8. drsmithy

    If cars aren’t crashing, they won’t need crumple zones. I expect cars will be a lot smaller and lighter. Perhaps more like golf carts or even motor bikes.

    No chance. Robot cars will inevitably crash, and it would be gross negligence not to design them to do so as with current cars.

    One good thing is that if cars are fully autonomous – that is, with no allowance for manual control at all – then they will be much safer in collisions, especially for drivers.

  9. drsmithy

    I was reading an article in the AFR recently that deliberated on the ethics of self driving cars and what would happen if the car had to make a choice between saving its occupants or saving someone else on the road. i.e to avoid a collision does it run over the footpath and potentially hit a cyclist/pedestrian? Or does it take the hit?
    These decisions are made by people at the moment, and they are accountable for them by law. But if self driving cars make the decision, who is accountable?

    Ignoring the accountability question – and I do agree it’s a big one, probably the biggest impediment to autonomous cars – decisions like this should be practically nonexistent. In nearly any situation where the decision is “hit a person, or hit a car”, vehicle speeds will be low enough that “hit a car” is the obvious answer. Until you get over 80km/h or so, vehicle collisions are, but and large, quite survivable for the occupants. Over that kind of speed, pedestrians tend to be relatively rare.

  10. Dylan Nicholson

    I have to say if anyone can work out how to install a skittles alley in a driverless car, I’m investing all my money in it.

    Slightly more seriously, it’s surely only a matter of time before computers are so much obviously and measurably better at controlling cars that human-driven vehicles will be outlawed except on dedicated courses. As I’ve said before, as a cyclist this idea appeals to me immensely (not least because the idea of spending 4 hours driving somewhere interesting to ride now becomes a lot more appealing!).

  11. drsmithy

    I struggle to see where congestion will be substantially improved, save for high speed roads where following distances can be reduced (which may not happen for other reasons – i.e.: passengers being rather alarmed at flying along a few inches away from the car in front). What’s the logic here ?

    The “fuel economy” of vehicles will likely cease to be a question around the same time (probably sooner than) autonomous vehicles are ubiquitous, so that’s a minor point at best.

  12. Jacob HSR

    Alan Davies #8,

    Electric cars and hybrid cars have grave implications for petrol tax indexation.

  13. Roger Clifton

    Now that we don’t have to sit facing forwarding more, we can sit facing back, which is much safer in the event of collisions. But then, if traffic is driverless, there will be no collisions anymore will there?

    If all four seats could be rotated, a full taxi could hold a meeting between busy people, record the meeting, drop off the occupants in order of priority, and send the recording and the bill to corporate headquarters.

    On a long commute, a young family could climb in, complete with breakfast yet to be eaten and clothes yet to be donned. Perhaps even with homework yet to be completed.

    For that matter, we don’t have to sit at all. A lone passenger could catch a nap during a boring commute, having switched the transparency of all windows to black. Of course, the ride would be smooth enough to snooze.

    Considering how expensive hotels can be, a visitor could opt to stay overnight in the same bunk, at a parking facility chosen by the taxi, and be driven to the ablutions block before breakfast. The midnight to dawn shift is normally quiet, so the taxi operator might be grateful enough to provide those facilities.

    Come to think of it, perhaps we can avoid paying rent…

  14. Alan Davies

    Tony Morton #5:

    I expect robot cars (nice term) could be more like first class on airplanes; space to fully recline and a flat surface for eating, computing, writing, etc. Add in some ‘stretch” space and interiors of existing small cars would accommodate that no trouble, esp for a low occupancy commute. Remove steering, pedals, most of the instrument panel, crumple zones, etc. Seats on multi person vehicles could face each other as on trains.

    With rented cars, travelers can choose the sort of vehicle they want for the particular journey e.g. single person for commute, large for a family. I’m not optimistic about renting though; more than now for sure, but I think the majority would still elect to own a vehicle, although the warrant for purchasing a second vehicle would be reduced. Owning vehicles will reduce the benefits from driverless cars.

  15. Oz (Horst) Kayak

    AUGMENTED & AUTONOMOUS VEHICLES
    Feedback from my cohort of 65+ strongly supports the Scott opinion 3#, particularly:
    “Augmented driving however (the use of censors etc. to improve reaction times/park/avoid driveway accidents/allow people with disabilities/elderly to drive) I think is a better use of capital than a fully autonomous driving experience.”
    Policy steps taken toward an autonomous integrated land-use and transport system can only benefit the safety, health and well-being of the ever increasing percentage of people with limited abilities due to age in our urban population.

  16. Alan Davies

    Jacob HSR #7:

    True; I’ve written a quite a few times over the years about driverless cars but I don’t think I’ve ever written about electric cars. I will correct that sooner rather than later.

    But I think they’re pretty different topics. Don’t need vehicles to be electric in order to be driverless. Only advantage I can think of is it might be easier to automate recharges than refills. I expect it’s a matter of time but there are still issues like charging time and range to overcome; that’s relevant given some of the more optimistic time frames touted for driverless cars.

  17. Jacob HSR

    You write way more about driverless cars than you do about electric cars.

    The Tesla Model S is on Victorian roads. Driverless cars are not.

  18. Jason Murphy

    If cars aren’t crashing, they won’t need crumple zones. I expect cars will be a lot smaller and lighter. Perhaps more like golf carts or even motor bikes.

    Expecting the average driverless car to have capacity for multiple passengers is wrong, I think, if we’re renting them. They’ll have space for two, or one and luggage. Most vehicles have only one person in them on most trips, remember. there’ll be a subset of multi-person cars you can obtain, but they will be much rarer.

  19. Tony Morton

    Point taken Alan. However, if the assumption is that people will be using robot cars to recreate their living rooms, that does presume that the cars will be privately owned by their users. Meaning they’ll need to be either stored at their destination or add to traffic while en route to being stored elsewhere.

    All of which means that at best, they’ll do no better at solving urban space problems than conventional motor cars, but more likely will add to these problems. More so if they take mode share away from public transport.

    For years we’ve been told the car is king because people like being in charge of one. I don’t really see people being that attracted to a future where they spend more time in cars but don’t get to drive them.

    Besides, no matter how comfortable you make the journey, having to get up at 5am rather than 6am to beat the traffic on the daily commute won’t be any less of a chore.

  20. Alan Davies

    Tony Morton #1:

    I think an autonomous vehicle should provide clear cut advantages over public transport in terms of the way in-vehicle time is used.

    Occupants will be able sleep lying down and with a friend; they can sit at a desk to work; they can practice singing or playing the guitar; they can study in silence; they can play monopoly on a table top with a friend; they can make personal phone calls, perhaps loud or mushy ones; they can watch a (relatively) big screen TV or listen to loud music without earphones; they can brush their afghan hound; they can do exercises on the floor; they can belch, snore, fart, argue, yell, put their feet up; etc.

    The big differences relative to public transport is more privacy and more personal space. The distinction between home/office time and travel time will be much reduced, consequently making travelling time more productive and less of a burden. I expect that many will use that opportunity to travel more e.g. taking a better job further away; a better dwelling further away; nipping down to Geelong from Melbourne for a party/show/restaurant- after all you can sleep on the way home and the driverless car takes you to your door.

    Zeebee Johnstone #2:

    Initial costs will be high but like everything else the cost per unit should decline as the R&D and set-up costs are recovered and with economies of scale. There should be cost savings from lower safety requirements e.g. lighter bodies, fewer airbags. There’s the option of hiring a car as needed rather than owning one too. I think the more productive use of time is the killer though; it should make the cost of travel by autonomous car much cheaper than car travel is at present.

  21. Scott

    I don’t think self driving cars will ever replace full driving. Where is the accountability? Who is responsible for the actions of a self driving car? The legal liability is a big question mark.

    I was reading an article in the AFR recently that deliberated on the ethics of self driving cars and what would happen if the car had to make a choice between saving its occupants or saving someone else on the road. i.e to avoid a collision does it run over the footpath and potentially hit a cyclist/pedestrian? Or does it take the hit?
    These decisions are made by people at the moment, and they are accountable for them by law. But if self driving cars make the decision, who is accountable?

    I think this is one of those areas where Google and like are spending huge amounts of money on something that will never get up.

    Augmented driving however (the use of censors etc. to improve reaction times/park/avoid driveway accidents/allow people with disabilities/elderly to drive) I think is a better use of capital than a fully autonomous driving experience.

  22. Zebee Johnstone

    I suppose much depends on cost. How much would these car trips cost? Can a single person travel for less than a train ticket, the same amount, twice as much?

    can’t really tell what the result will be till we know the magnitude of cost to the consumer.

    If lots more people are using cars for trips, wouldn’t that mean a lot more congestion on feeder roads? And so more wear on those roads? Ask anyone whose street has become a rat-run what it is like to find the traffic count shoot up. Probably not a huge issue for the car-dependent suburbs but a very big one for middle ring and inner city.

  23. Tony Morton

    Interesting thoughts here Alan. But there’s also a paradox of sorts at the root of all this, that goes to the unexamined question why all this hypothesised extra travel should be happening in the first place.

    The idea is that robot cars will cause people to spend more time in them, doing things in transit that they can’t do now because they’re fully occupied piloting the vehicle. Now, this presumes an expansion in the range of activities we can undertake without being in a fixed location – after all, it still does us no good sacrificing productive time at home, in the office or at the shop for less productive time in transit. But then, if the value of our activities is to become increasingly decoupled from location, why should we at the same time devote more of our limited time budgets to shifting our location?

    It seems clear that if robot cars do release latent travel demand by liberating people from the driving task, this effect is self-limiting. And to the extent we have any inkling where this limit will be, I’d suggest it’s likely not far from the status quo. We only have to consider the extent to which people will tolerate longer travel times by public transport compared with an equivalent car trip. Travel by public transport presents many of the same opportunities for productive activity as in a robot car, but most people don’t appear to give this very much value – they still want public transport to be time-competitive with driving their own car.

    Robot vehicles show a lot of promise for public transport itself, as you say, and also for a large part of the mobile workforce – tradespeople, sales reps and such. But again, these are workers whose skills are needed at their various destinations, not so much en route. For the rest of us, robot cars don’t solve the essential space problem that comes with living and working in cities. That hypothesised 2:1 capacity benefit won’t apply when roads are loaded to the point where traffic is almost stationary – and most of the research behind such claims also assumes a world devoid of pedestrians.

    Beyond potentially eliminating the professional driver as an occupation, and squeezing out a bit of extra road network utilisation at the margins, I don’t see robot cars making much difference to the shape of our cities or to most people’s willingness to spend more hours each day travelling between A, B and Z. What I do see is that when people do leave home, they’re increasingly wanting to do so on foot.

Advertisement

https://www.crikey.com.au/2015/06/30/would-a-world-of-driverless-cars-be-all-beer-and-skittles/ == https://www.crikey.com.au/free-trial/==https://www.crikey.com.au/subscribe/

Show popup

Telling you what the others don't. FREE for 21 days.

Free Trial form on Pop Up

Free Trial form on Pop Up
  • This field is for validation purposes and should be left unchanged.