Two weeks ago posted some potential doomsday scenarios that could arrive from the introduction of driverless cars. They were intended solely as a bit of fun although some took them a bit seriously. This week the Herald on Sunday went big on driverless cars talking up the potential for the technology, including an op-ed and dedicating their editorial to it.

We read today of a future of driverless transport with a mixture of excitement and foreboding.

Excitement, because it offers the prospect of fewer accidents, fewer injuries and fatalities, fewer cars parked on city streets, fewer cars everywhere.

For along with driverless cars we will get the ability to use them more efficiently, ordering them with a phone app rather than owning one that is parked somewhere most of the time. They will drive closer together at permitted speeds, co-ordinating their movements more safely than human drivers can manage.

I think on the safety aspects alone driverless cars will become an essential part of the transport mix. I can’t think of any other area in our society where we so casually accept over 300 people being killed and over 11,000 being injured every year. While they wouldn’t have to be electric, I also suspect any roll out of driverless cars would speed up the adoption of electric vehicles which so far has been close to non-existent. This would obviously help reduce emissions and therefore help improve both the environment and people’s health.

Where I think the benefits are less certain are in many of the claims made about just what impact driverless cars will have on transport issues such as congestion or travel habits and preferences. For example, some experts now believe that while the benefits of cars driving closer together and co-ordinating their movements will increase road capacity, much or even all of that could end up being negated by vehicles making additional trips.

I’ve been thinking a bit recently about the potential use case for driverless cars and one thing that’s struck me – and for which I haven’t really seen commentary on – is just how things would work in reality. What concerns me is that in many ways we could be heading towards a situation where driverless cars represent combining some of the worst features of driving with the some of the worst features of public transport – particularly the part where you have to wait for a service to turn up. That would hardly be the transport revolution we’ve been promised.

Let me explain. There are two main scenarios that have emerged as to how driverless cars could be integrated into our future. One is an extension of the status quo where individuals own vehicles and the second is that we move to a taxi type model where the vehicles are owned by a company and you just book a trip when you need one. Below are some of the downsides to each of these scenarios that may make driverless vehicles worse than what we have now. Before going into them I’ve assumed places like Albany, Botany and Manukau and many others are bustling from all housing and commercial buildings that have been developed on what became unneeded carparks – in the process going a long way to helping addressing housing issues.

Botany showing the existing buildings in red and parks in green.

Private Ownership

This is an extension of the model we have now. To many people cars can be more than just a transport tool, they can reflect individual personalities and act as an extension of our homes while at the same time being a part of the family – we often give them their own bedroom after all. It’s for these reasons I can see a lot of people still wanting to own a car rather than use the taxi/sharing model many predict.

Positively driverless cars may allow households to reduce the number of vehicles they own, for example instead of owning two or three cars many families may only need one. Perhaps it drives one parent to work early in the morning before quickly returning home to pick up the other parent and child, dropping the latter at school before driving before dropping the second parent at their workplace in a different part of town. It would then drive home and park waiting to be summoned for the afternoon.

It’s the latter part I’m particularly concerned about. Let’s say you’re at work and want to go home so you pull out your phone and open an app to summon the car. Do you then have to sit and wait for the vehicle to come and collect you, possibly from quite far away. Instantly we’ve added one of the most frustrating aspects of PT to the private car, having to wait for service. It could be a bit like walking to a bus stop and seeing a half hourly bus just pulling out leaving you with a long wait for the next one.

One of the strongest features of cars is that they essentially have an unlimited frequency. Will people put up with potentially having to wait a while for their ride home? Obviously some people could pre-plan their travel times so the car turns up as they’re leaving work but many people don’t work to a set schedule.

Shared Ownership/Taxis

Many people are predicting that in the future no one will own a car and instead we’ll all shift to a taxi/sharing/Uber model where we order a trip on our phones and a car turns up to whisk us to our destination. One big predicted advantage is we’ll be able to get better utilisation our of our vehicle fleet instead of them sitting idle for 95% of the day. Just how many driverless vehicles a city will need is unknown although suggestions from study on come cities suggest perhaps a low as only 30% of our current fleet. This means less vehicles overall although at peak it doesn’t necessarily mean less on the road at one time.

Because there are expected be so many vehicles there should be coverage across much of the city however as you’d expect the best coverage will be in areas with higher density. This is quite well shown by this map Uber released last year to mark their millionth trip in NZ. It shows trips taken with the service over a three month period (they say start and end points have been altered to protect privacy).

Uber Auckland Trip Map

Close to the city wait times for a service will be fairly fast however those living in suburbs further out people will likely have a longer wait unless companies dedicate a lot more resource to providing coverage – again introducing the wait time component that PT has.

The other PT like aspect is that of course you don’t own the vehicle. I suspect like taxis today there would be different levels of service available but tied in with the point above, the more expensive/exclusive services could result in longer wait times to travel anywhere.

Another aspect that gets lost is the extension of the house aspect private vehicles have.How many people that drive use their car to store stuff should they need it. Perhaps kids gear for after school activities or your golf clubs for that round at a council subsidised golf course – basically things you might not want to be lugging into work to sit next to your desk all day (if you work at a desk).

One last set of thoughts, what about animals. For example, many people drive their dogs to the beach or dog parks and of course to see a vet. Hopping into a car that smells like a wet dog has just been in it is hardly going to be a pleasant experience, especially if you don’t like dogs or worse are allergic to them. Similarly, what about parents with kids, will these cars carry a few car seats around in case one is needed with parents needing to go through the process of setting it up in the driverless car before every trip.

The situations above and many others could quite easily be addressed through specialised services targeting different parts of the market. Of course to do so is likely to involve additional costs, both financial and time.

I can’t help but feeling that driverless cars will require more than just a technology change but changes to how we think about transport too. I suspect ultimately we’ll ended up with a mix of the two scenarios above – which in itself is a departure from what we have now. Once available, adoption by and the disruption to the taxi market will happen fairly fast but that after an initial burst adoption will slow down as private buyers only get them when it’s time to buy a new car. Given the average age of vehicles in NZ is over 13 years that will take a while.

Share this


  1. One thing I’ve been pondering: if people are perfectly willing to pay a large amount of money up front and over time for the convenience of their own car (often a second, third, fourth in the houshold) why will the suddenly not want their own car when they become driverless.

    Saying a share care would be cheaper isn’t a satisfactory answer because it’s already cheaper to not own your own car, it seems people will pay handsomely for some factor of private ownership, be that convenience, security, style or preference for something personal and private.

    My guess is most folks will still own their own one, and might start buying extra ones for their children.

    1. Your point is excellent. In Singapore, which must have one of the best PT systems in the world, people will pay over $100k to own a small car, which must be scrapped or exported at the end of 10 years. They certainly are not paying this amount for a cost effective means of transport.

  2. This is excellent. Great to see some deeper thinking on this. Imagine the congestion if everyone ordered their car for 5:00pm. I think for commuting people seriously might find PT more convenient and more reliable.

    1. The biggest argument I’ve heard from driverless car supporters (or more accurately anti-PT pro-car) is that driverless cars will make public transport (ie trains and buses) redundant.

      What you’ve described re-iterates the fact that driverless cars are not much different than human driven cars in terms of their physical characteristics. They still occupy the road spaces and hence will still block roads especially since many of them will still be single occupant during peaks times. No matter how efficient you make autonomous driving ability there will still be congestion and PT still needs to complement that the same way it does now.

    2. I just pictured a worst case scenario where everyone has their own private driverless car sitting at home and orders them to come pick them up at the office at 5 and then take them home. The number of car trips taken during rush hour has suddenly been doubled! and you will have rush hour traffic in both directions. I think I will want to stick with PT. Even if many people use a taxi service there could still potentially be a lot of cars on the road.

  3. Another thought is that utilising improved safety technology isn’t reliant on a car being driverless. Tech such as radar assisted braking or audio warnings could be adopted for human driven cars too. Claiming safety benefits will only arise if cars are driverless is therefore false.

    1. There already are. We had a product famil Volvo with auto braking for pedestrians at work a couple of years ago that we tested out on the interns. It worked, they survived.

  4. The market will decide and that is fine. If PT is cheaper then some will stick with PT. If it is quicker then others will stick with PT. People will balance time against cost and decide what form of transport works best for them. If I order an uber at peak, the cost will be much higher than off peak. I must be willing to pay the premium. If I subscribed to a pool car scheme I would have to be flexible with the time I use it. You would need a smart system to plan trips in the most efficient means, like ambulances use, just more complex. If I own a car then what difference does that make if I am already driving a car? The main advantage of a driverless car is safety and being hands free to do other things such that the time sitting in the car doesn’t have to be wasted. Being stuck in traffic is no longer a waste of time for many people.

    Overall there will be safer roads and fewer vehicles on those roads. But perhaps more trips.

    Regardless, it will be a decade or two before these vehicles become more than an expensive novelty.

    1. It’s the potential for human driven cars or trucks to hit the driverless car that requires it. Driverless cars will only ever be an additional transport option, not a replacement. Co-existence needs to be factored in.

  5. I think the flow down of the technologies developed for driver-less cars to driven cars will mean many of the safety aspects will be realized for driven cars. I think this will further slow the move to driver-less as one of the reasons for driver-less will have already been realized.

  6. There’s a scenario you haven’t mentioned: Driverless vehicles arrive, but they’re not evenly distributed for a long period of time.

    I’d expect driverless vehicles to be most appealing to vehicle fleet operators, who currently keep their vehicles on the road for a lot of the time and pay drivers to move them around. Bus operators, trucking companies, taxi companies, etc. Once the technology is proven, I’d expect these types of organisations to be early adopters.

    This is a scenario in which:
    * The cost of operating PT services falls by perhaps 20-30%, making it viable to raise frequencies almost everywhere
    * The cost of road freight falls by a similar margin, and operators are freed from some scheduling issues
    * Taxis get cheaper, but as Matt noted that’s mainly going to benefit people in dense, mixed-use areas.

    1. This would only happen if driverless cars were allowed to operate unoccupied. I suspect that would be a long way off. I also doubt someone would be hundreds of thousands of dollars of freight without an occupant inbthe vehicle. Seems kind of easy to clean out that truck.

      1. You probably have a good point on their still being an occupant, at least initially. however a driverless truck could be locked when it leaves and it isn’t going to stop until it gets to its destination. Perhaps it could be hijacked remotely, but then any occupants will suddenly become hostages.

  7. The more problematic future aspects of driverless cars relate to how they interact with pedestrians and cyclists. Already the car-makers are keen to place further restrictions on active transport users:

    The reason is obvious:

    is that in a city where driverless cars are programmed to avoid cyclists and pedestrians these cars wouldn’t be able to move for all of the cyclists and pedestrians riding and walking in front of them.

  8. One of my biggest concerns is how the transition will be managed from having a tiny minority of autonomous vehicles sharing the road with conventional cars through to a situation where the autonomous cars are in the majority.

  9. You seem to imply an either/or model – either we’ll keep owning, OR we go the sharing route. The reality will be a mix.

    So in the first scenario you describe where the owned car shuttles the parents and children to work/school… rather than returning home, the car could be sent to park near the first parent’s work (obviously this means there’s still a req for carparks) so when they are ready to leave there is no wait. If the other parent unexpectedly needs to leave early, then they summon a sharing service (just as they would nowadays with a taxi or bus).

    There is third (also complementary) model that is missing from your post as well… short or long term leasing of vehicles. For the parents with small kids who need car-seats scenario, perhaps they could lease a car for 5 years then after that relinquish the car and go back to sharing services. Similarly, if you’re going on a road trip, or need to move house, you rental an appropriate vehicle for the duration then go back to car sharing afterward.

    Having all three models available should eliminate your questions around convenience (although agree that congestion question remains).

  10. Given one of the supposed benefits is a reduction in parking, with driverless cars able to drop off and pick up and constantly circling for the next ride, I wonder about the choreography of these movements in city streets, and the effect on capacity. We will likely get long driverless taxi ranks along the kerb as people get in and out of the vehicles. WIth all these people movements in and around vehicles, the car software will be cautious and slow for safety reasons. Any proposed benefits in capacity from driverless cars will be lost (or be even worse) in this environment – and I wonder if this congestion would flow further back into the street network.

    Also, driverless cars seem like a giant IT project, and we know those don’t have a great track record of being delivered on time, budget, or being stable. The potential for gamification and hacking will be ever present.

    1. I certainly don’t buy the theory that driverless cars will end the need for parking. That would effectively mean car owners or users would get a less convenient and much less efficient outcome. Whats far more likely is that driverless cars are used just like cars today. Only side benefit is safety. Maybe a shard ride model is developed too. Having all cars circling on the road is absolutely insane and counterintuitive. It’s energy inefficient all to save on parking? Having them park far away on private car parks reduces the utility of owning a car and is also less efficient. Shared cars are just a slight steo up from good public transport.

      Driverless cars will probably end drink driving, speeding, provide route efficiency, and increase safety. That’s it.

  11. Glad you mentioned electric cars Matt – a lot of the media coverage of driverless cars imagines a kind of techno-utopia, where people whiz around in fleets of driverless, electric, environmentally friendly cars. But driverless cars and electric cars are completely different technologies. They could influence each other in unpredictable ways. There’s no particular reason why driverless cars need to be electric, or vice versa. I’d note the following points, which I might try to flesh out into a post:
    1) I’d agree driverless cars will probably reduce the total number of cars needed (I wonder whether NZIER have considered that in their projections for Ports of Auckland?)
    2) Taxis, Uber etc will be early adopters of the tech. Drivers are one of the big costs of the service they provide.
    3) They’ll also be early adopters of electric vehicles, as they have already been for hybrids. They do high mileage, so again they get cost savings out of it.
    4) Driverless tech will start to be widely available at some point. There’s no clear consensus on when that will happen. Maybe 5-10 years. But then it will take time to percolate through the fleet – we wouldn’t want to junk our existing cars just because driverless ones arrive. For most of us, what advantage do we get from it? I would think it’d be quite limited.
    5) Electric cars are further away from widespread adoption, maybe 10 or 20 years. They’re available now, but still a long way off being cost competitive.
    6) The thing that will bring down the “cost curve” for electric cars is mass production. But if driverless cars mean we need fewer vehicles, could that make it harder for electric cars to reach those economies of scale?
    7) As for environmentally friendly, this post argues that the emissions benefits from electric cars aren’t that big for most countries, unless they also switch to renewable electricity. This point doesn’t relate to driverless cars specifically, but it’s one of the themes I’m exploring in the other post series. Public transport is often a better way to reduce emissions.

    1. I’m not sure I agree about electric cars being 10 to 20 years away.
      I think the kind of hybrid that can do the first 30 odd kms on electric and then switch to petrol will take off very quickly. If I had one I could do at least 95% of my travelling completely electric. They wouldn’t need a huge battery so there shouldn’t be a huge extra cost compared to a standard petrol car. Of course it wouldn’t be any use to taxis, delivery people, etc, but for the average commuter it could potentially save a lot of money (especially if they start to charge a realistic carbon tax on petrol).
      In terms of the environment, electric cars might not be that much better in terms of total carbon output in some countries, but in terms of local pollution within cities they would make a massive difference. Can you imagine the difference to Beijing if it had 100% electric vehicles?

      1. Yeah it would be pretty great. Though the problem with Beijing is being located right next to dense industrial areas in China – you don’t see other large Western cities with similar amounts of cars being nearly as polluted.

        Cars do contribute but just are not as big of a factor in pollution as with things like burning more fossil fuels to supply more electricity for the new electric cars. In terms of environment, sustainability is key and until use of fossil fuels goes down pollution will remain close to the same as before. There is also the issue of dealing with an increase in new types of waste which is harmful to the environment such as used car batteries of the electric cars/buses/trucks a decade down the line of mass adoption.

      2. Agreed that plug-in hybrids are an important step on the way to fully electric vehicles, and that they will become widespread earlier. Even so, they’ll take a long time to become a big part of the fleet, unless there are major incentives in place. Part of the reason per-kilometre costs for EVs seem so low is that they’re not paying for road upkeep (as per

        Reducing local air pollution is a good thing, especially for big cities, but you still cannae change the laws of how much space things take up, as per Peter’s post this afternoon!

    2. I agree conflating driverless and electric is arbitrary. I see some competing things around driverless and electric in terms of whether driverless will help the adoption of electric.

      The primary thing about driverless cars is that we are likely to sweat the asset! Electric vehicles have limited range and require time to recharge. How much time will an driverless electric need to be out of service for in a given day to recharge?

      On the other hand you could imagine a system whereby charging stations are set up and electric cars simply direct themselves there to recharge when required. These recharging points could be located in lower value locations compared to where people would otherwise park a car, and of course the charging infrastructure could be shared between multiple cars. It would be specialist fast charging infrastructure although I am not too familiar with this technology to know what this means.

    3. The reasearch has moved on on electric cars since then John.

      Last year the Union of Concerned Scientists (UCS) in the published an updated cradle to grave analysis of electric cars and Greenhouse gas emissions compared to existing cars.
      Read it here.

      Headline finding:
      “Over their lifetime, battery electric vehicles produce far less global warming pollution than their gasoline counterparts—and they’re getting cleaner.”

      And in most states of the US, an electric car is still better even if that state actually has a low percentage of renewable electricity used to charge them.

      And recently EECA did the same for a NZ context and found even better results to the UCS findings.

      1. Yup, interesting analysis that one, and I’m planning on doing a post about it sometime (but I want to finish the original post series first!). There’s certainly some gains to be had in the US, and of course Europe is also moving towards a renewable grid, and NZ has a big head start. But electric cars can’t solve every problem (congestion, etc), and to lock in emissions reductions in the near term means investing in public and active transport too.

        1. Even if every bus went fully electric tomorrow and people walked and cycled way more and drove tons less, the true impact on CO2 emissions would be there, but not overly high given the relatively low portion of the vehicle fleet and VKT PT represents.

          Yes an electric PT fleet can displace a lot of car trips – but you need the PT routes and run them at frequency to be able to get there.

          With true driverless buses/trains you could easily roll out such wide coverage and run the fleets at high frequencies 24/7 if you wanted. That would give Uber and Co a real run for their money.

          But the true impact will be driverless freight – be it trains, trucks or a seamless automated combination of the two providing short and long haul capability door to door.
          Who would care how many times the container of goods changes vehicles to complete its trip if its all automated, is fast and no humans have to be paid to be involved in the process?

      2. The headline should be that if the US switched its private vehicle fleet to EV, the worlds lithium resources would be depleted in 7 years. Given that the expected life of a battery is 6 years, the EV revolution looks set to be a non starter.

        1. They made similar claims about Platinum and Paladium when the US mandated catalytic convertors in the ’70s. Didn’t happen. Won’t happen here.

          In any case, Lithium makes up less than 1% of the average Lithium Battery by weight, and Lithium is not a rare earth – its 3rd most common element in the Universe after all.
          As for the 6 year life span you claim they have? I think the jury is still out on that, in any case the outlandish replacement costs touted even if they did have never come to pass, and the batteries can be recycled to make more down the track if it proves necessary.

          So I find your bland assertions, like the rest of your bland comments in this blog to be totally ignorant and extremely biased towards the nothing but the Fossil Fuels you so love.

          1. As an aside to this it is crazy there is no consumer accessible battery recycling in NZ. I’m sure there will come a time when we start mining our landfills for all the valuable resources we throw away currently.

          2. Yep, 21st century coal mines, with all the dangers of the 19th century ones (like methane pockets, land stability issues, subsidence issues for surface buildings), and then some (tons of plastic clogging up all the mining equipment).

          3. Way to go Greg – attack the person not the argument x
            If you are not too bothered about the Lithium problem, turn your head to the rare earth magnets used in EV motors. You might be surprised to know that is a limited resource as well 🙁
            But if you really want to knock the EV argument to touch, just ask yourself where is the infrastructure. You think you can park your EV overnight away from home and just plug into a strangers electricity socket? Or do you think we are going to rip up the cycle paths to plant EV charging stations outside every home?
            The oil industry has a bit more cash than Tesla and it is investing in Hydrogen. The infrastructure does exist for that delivery as you will just fill up at a petrol station and pull the Hydrogen pump rather than Diesel or Petrol. The added benefit is refilling Hydrogen will be a 5 minute job while re charging your EV from empty is 10 hours +

          4. Your comment was about not enough lithium for 7 year demand of EV batteries and now you’ve been called out on it you want to change the topic to rare earths, because well, they’re rare aren’t they, its in their name right so it must be true?

            We’ve poured as much if not more rare earths into electronics used in mobile phones and elsewhere over the last decade or so, which are recycled and the rare earths recovered way less than the rare earths used in EV’s will be, especially as in your estimation there’ll be so few of them ever made right?

            All those comments about charging and such will never work – were all made about other technologies we now take for granted.
            It is true that hydrogen could power EVs using fuel cell technology, but that future of that is even more cloudy.

            The only problem that Hydrogen actually solves is fast refilling, But it comes with a ton of downsides, not least as NASA and go found way back when, its a very dangerous material when liquified, costly to make and hard to store without a lot of energy being spent on doing so.

            However, the fast charge problem with EVs is its only drawback, range is being solved already and faster charging will be solved with newer battery technologies, well within the future timeframes for when driverless cars will become commonplace.
            [2030 is the current likely estimate for full Level 4 “completely driverless” automation], so that problem will be solved within 15 years, perhaps not easily, but it will be.

            Hydrogen? Not so much, its been looking for problems to solve on the ground since way before it went to moon with NASA in the 1960’s, if it was going to work in the mainstream back on earth we’d be driving Hydrogen powered cars now.

            I assure you mass flying cars will never arrive until driverless tech is fully mature on the roads, because flying cars are too hard for most people to drive, and must be fully automated to make them workable.
            So thats a long way off except as a niche product at best.

            Fleets of fully driverless EV, Ubers and Lyfts on the roads on the other hand – commonplace in 20 years?

        2. The batteries last way longer than 6 years – google Tesla and check their warranties. Talk to taxi companies about all their older 2nd hand import hybrids – they aren’t replacing batteries at any great rate. Fast charging isn’t an issue either. Tesla are setting up battery swap out stations – 90 seconds. Google, Youtube, whatever. Do some research.

          1. I stand corrected Ricardo, Tesla and Nissan warranty on batteries and electric drive train is 8 years. Still, how many petrol tanks need replacing every 8 years at a cost of thousands?

          2. James – I think you’re assuming that the length of the warranty is the same as the economic life, but that’s not generally the case. For instance, most new cars have a three or five-year warranty, but the cars don’t need replacing after that length of time.

          3. As I understand it, from talking to the Tesla dealer in Palo Alto, the problem is the batteries degenerate. The warranty is that the batteries will retain 75% (I think that was the figure) of their life for up to 8 years. So there is a drop off in range prior to the warranty expiring.
            Tesla are a smart company, the next generation of cars will have batteries that can be replaced by cell. They will then be able to just replace the damaged cells and not the entire battery bank.

  12. The more I think about driverless cars, the more I think they will increase PT use. Private car ownership will drop significantly – I doubt many households will own two cars, and a lot will have none. Driverless taxis will supplement the PT system – you will be able to afford to take a taxi if your trip is not well suited to PT. At peak there will still be congestion even with driverless cars, and fares for driverless taxis will be a lot higher at peak, so good high speed PT will be very well used.

    For the doubters that say that driverless cars will be the end of PT, ask yourself: do you really think they will be turning off the London underground in 30 years time? Or the NY subway? Or even the Sydney train system? Can any of these cities handle significantly more cars on their streets regardless of whether they are driverless? Why do we keep on thinking Auckland is different?

    Of course many have tried to predict the future of transport and failed…

  13. Another option – You own a car and for the 95% of time your not using it you “supply” it to an uber-type service.

    Not sure your point about coverage. Obviously frequency/convenience in lower density areas will tend to be lower than in the city, but what the actual difference will be is the question. Are we talking waiting an extra 15 minutes or 3 minutes? Even in the outer reaches of Auckland there are still cars whizzing down roads at relatively high frequencies. If driverless request-based PT gets 50% mode share there will still be a heck of a lot of demand. Whether it is a last mile service, a full length service, route based or request based can be left up to the market. I think the focus on “cars” (somehow being the opposite to PT) is unhelpful. There will be vehicles of varying number and size, varying in location, varying in service type according to market demand.

    Oh and even more than now we will need market based road pricing!

    1. “you “supply” it to an uber-type service” – So what happens to all the crap that the average person leaves in their car? Will you have to leave your car pristine each day before handing it over?

      I can’t see the average NZ male, who defines their self worth and masculinity by the car they drive, giving that up. They also are unlikely to want to give up actually driving the car.

      Driverless cars seem like a lot of things – a fantastic idea until you involve actual human beings along with their egos and insecurities.

      1. Who knows, – a lockable boot space? My family owns a motel unit ata beach that is effectively a bach (to us) but it rented out when we are not there. There is a lockable cupboard in the kitchen and a storage locker in the carpark. Anyway its just another ownership option. People will respond to monetary incentives. I actually think it is more likely that very few driverless vehicles will be privately owned. Maybe by people who cant drive. As for the your description of the average kiwi male… I know a lot of kiwi males and none fit your description. My guesstimate would be that car “enthusiasts” are a small minority of kiwi males. Or maybe I just interact with non-average people?

  14. A thought re congestion… with no need for a driver (or even driving controls) there is no reason for driverless cars to keep their current form factor. A lot of the sharing service cars could be the size of a tuk-tuk with a capacity for only one or two people (as we all know, most car trips are single occupancy anyway). This would make their use of parking and road space much more efficient. So if roading infrastructure doesn’t change too much from now, a lot more traffic could be accommodated.

    1. Good point – much narrower cars that don’t need to drive in lanes like stupid humans do could seriously increase the capacity of roads. But it probably won’t help much until all of the human drivers are extinct (gotta be 30+ years away)

  15. How are driverless cars going to work in the US where litigation is so common? Who is going to take on the insurance risk – Google? No chance! Who wants a computer deciding your fate in an accident?
    Driverless cars are as nonsense as EV’s. They will be a tiny footprint in the history of land transport.

    1. Probably would have said the same 100 years ago about the idea of letting any old person get propel a 1 tonne piece of metal at speeds of 100kmh in uncontrolled environments. Laws and legal systems adjust to technology, albeit slowly.

    2. California is already proposing regulations ( regarding this. There will need to be a driver present who can take over the controls if the situation arises (ie they must be paying attention and licensed) and will be liable for any accident. Once the US has applied these rules other countries are almost certain to follow suit.
      Personally I don’t think that they let the no control option be allowed on the basis that someone needs to take the liability of driving, as James said the big companies producing these vehicles won’t/

    3. What an oxygen thief. So due to litigation issues we carry on killing 1.2 million around the world every year with the current system?
      Presume you’ve shorted Google seeing they’re throwing money down a hole?

  16. I wonder how this emerging technology will effect the perception of driving as a ‘right’. A review of current laws is required to allow driverless cars on our roads at present, especially unattended ones shuttling themselves between ‘jobs’. If accident rates fall to virtually zero for driverless vehicles, I could imagine that triggering another legislation review whereby driving becomes less of a right, requiring greater restrictions and more stringent training. Perhaps we’d get to a point where humans only get to take control of a motorised vehicle on closed circuits; track days for the enthusiasts and them parks for the nostalgic.

    In their attempts to shut down the used import business and protect sales of new vehicles, the motor industry lobbies heavily to have new car features included in rules for importing vehicles, and uses safety as a large part of the argument. If new, safety-based legislation comes about that would effectively restrict their sales market, how would they manage to spin against such rule changes?

      1. Except part of the point of driverless cars is that they don’t need a human driver – they can carry passengers unable to drive (because they’re asleep, or drunk, or children, or have a suspended licence, or never learned to drive). They even need to be able to do dead-running: operating with no passengers in them at all.

        Unless they can be sure of working 100% of the time without human intervention, their effects on transport are going to be minimal. It should make driving somewhat safer and less boring, but other than that, none of the main supposed benefits of driverless cars are possible as long as the cars still need an alert, sober, licenced driver in them to go anywhere.

  17. Driverless tech will break the current business model of paid parking. Why pay to park, when the car can just drive around by itself until it’s needed? Once driverless vehicles become available, we’re going to have to either (a) make all parking free; (b) implement full-strength by-the-metre road pricing; or (c) ban the exclusive private ownership of driverless vehicles. Whichever solution we choose, the implications for cities are huge.

  18. What if your semi-autonomous car could crawl down the motorway in heavy traiffic or in the slip lane cue, while you read facebook/news etc.? This would make car commuting less unattractive for some who have had about enough and are willing to consider PT to avoid being stuck in traffic, even if the time is similar or slightly longer. I think we could see semi-autonomous cars remove some of the worst things about driving everyday, which could slow down the uptake of PT. That said, I agree that other than removing some of the (human behaviour induced) bunching that causes traffic jams on motorways, driverless cars won’t help congestion so for many PT will still be the quickest and certainly cheapest option.

    I think the step from heavily semi-autonomous to full autonomous will be very hard to achieve. I am a massive Elon Musk fan, but I have serious doubts around his claims of how soon they can achieve full autonomous driving (without back-up driver and it be allowed on the roads). That said, if 10 years ago you told me one person would successfully launch a space cargo business, I would have told you, that you are nuts, so only a fool would underestimate the man.

    Don’t underestimate how far we are with semi-autonomous now though. My car is 3 years old, cost less than $60k and has radar guided cruise control which is faultless on the busy/tight Auckland motorway system and allows me to never touch the brake/accelerator even during heavy traffic. The only downside is that it cuts out below 30km/h. More expensive cars now also have lane keep auto-steering and I assume this works just as well? To get this working below 30km/h is probably more of a legislation hurdle than anything else. If my mass produced 3 year old car can do that, I wonder what the google car which probably has a super computer in the boot can do? That said, all the computer power (currently) available in the world would struggle to let a autonomous car know what to do when presented with a non-standard situation like a busy shopping centre carpark etc. The complexity of motorway/slip road driving is many orders of magnitude more simple a problem to solve than how to deal with a new situation with other cars moving every which way. Think about it…. when having a mexican standoff with somebody else for the one remaining car park, don’t you look into the whites of their eyes to see it is is worth challenging for it? How will computers do this? The cars will need to talk to each other and that requires massive industry agreement on standards etc….

  19. Current cars are designed now to be actively used for a small percentage of the time, they’re not built (or easily maintained) when run in a 24×7 operating environment.

    GM is how actively pursuing a future where they forget about mass adoption by the masses like the old ICE cars of today.
    And instead focus on a future where driverless is added to cars for a few “fleet” type operators, like Lyft (which they part own), and Uber and such.

    GM made such an announcement at the Detroit Auto show recently. Which was widely mis-reported as “GM would be doing a driverless trial real soon in the US”.

    See a Wired article about their plans here.

    Because GM knows that driverless, electric powered, permanently connected and expensive will all go hand in hand (and yes GM are actively linking driverless and electric cars together).

    They know you’ll have a car which costs 3+ times the price of the human driven ones, for one that can do 3 times the mileage (i.e. 24×7 operation) and won’t need the level of maintenance.
    And building those kinds of cars is a lot more expensive. And it will probably depreciate a lot faster than current cars do as well, so you’ll have to use it 24×7 to make money using it.

    And check out GM’s latest moves with driverless here, which suggest GM’s purchase of Sidecar may have gotten them a hold on a key patent which may impact Uber.

    And If GM – one of the worlds largest car makers is not considering mass adoption of driverless cars by people as their future and instead delivering a mobility service using such cars?
    Then why is the MoT and other government agencies pursuing an agenda that assumes that NZ will be?

    I still think the tech is quite a way from being in use here in NZ, simply because of the cost and complexities involved with delivering a complete driverless solution to the market.

    1. GM recognise that the future is in Moblity as a Service business with car ownership declining and the profits being made on moving people from A to B with GM made vehicles and services.

      1. Yes true, and they recognise it will be delivered by fully autonomous vehicles, running on electrons not fossil fuels.

        Because they know that the running costs and maintenance will be much lower, and many of the current issues with regards range and speed of recharge of electric cars will be solved, and well within the timeframes that these vehicles will need to be delivered in.

        So I think the deliberate conflating of these technologies by GM gives you an idea of roughly when they think these technologies will be mature enough for the market.
        Hint: Not any time soon.

  20. Another factor overlooked (especially for NZ) is the ability to tow a boat or trailer, and that in the both NZ and the US the largest sellers are not cars but light trucks. No reason that can’t be self driving, but how effective will they be in an off-road situation such as a construction site or farm?

    1. It could be quite handy to have your truck reverse the trailer down to the boat ramp while you unload the boat, then it goes and parks it self.

    2. My thoughts are the any sort of situation like off road, construction sites etc will require a human to take over for those parts. Therefore you still need the human interface part of the vehicle.

    3. If the Uber/Lyft model became the norm then I can see bidding also becoming part of the process. The result is personal transportation will become less egalitarian and more based on who can afford to pay for the vehicle to get there quickly. eg while you’re waiting for your car to arrive someone bids more because they’re in a hurry, now you’re offered the option to out bid them or wait for another vehicle – this isn’t a problem is you’re located where there is a high density of vehicles but could be significant if you’re in a more distant suburb.
      The other thing then is Toa Greening mentioned the network AI that can go into the vehicle network, this too can then become a matter of who pays more (similar concept to tolls) will get to go where they want faster.

  21. One other point is that the amount of data being generated from existing and new traffic infrastructure is also being developed and eventually AI traffic control systems will control the traffic routes for all vehicles to ensure optimal traffic flows. The AI control system will be able to work out new solutions to improve traffic flows which are not currently considered.

    For example directing all driver-less vehicles into certain lanes on the motorway for efficiency/safety. To analysing peoples travelling history and offering some commuters payments to go a slower route to work while others pay to go the faster route. To recommending new pedestrian lanes, public transport, driver less vehicles and even roads.

  22. Can someone comment on the risk of hacking driverless cars might be? I understand motor companies (Ford, Holden etc) will have brakes and steering to permit human intervention but google designed cars will not. Could we have a bit of mischief from smart young kids on a hacking challenge?

  23. Just looking at the taxi model for a second. On the people with smelly dogs/messy kids (or even just people commuting to and from dirty jobs who don’t have their own vans) point: while people are happy to get in a car which smells like their dog, they probably won’t be to get in a car smelling like anyone’s dog, so it’s not even as simple as ‘dedicate x proportion of the fleet to general usage and 1-x proportion of the fleet to smelly/dirty people’. Even if it were, there would be a number of contradictions e.g. because the dogs-allowed cars would need more opex on cleaning, the cost to run the service would be higher, but because those cars were more likely to be smelly you’d have to charge lower.

    The standard arguments behind why Zipcars etc. work today is that, due to online booking, excessive mess can be traced directly back to the person who caused it, but that doesn’t help the person in a rush who has booked a driverless taxi which turns up smelling like baby sick.

  24. Will driverless cars be great for walking and cycling?

    Assuming the primary rule is that driverless cars are not allowed to kill (or injure) people:
    – pedestrians will be able to walk across the road at any point and driverless cars will stop for them.
    – cars will see you cycling along and not allow the door to be opened into you.
    – cars wont drift into the cycle lane, or get road rage if you pass them.

      1. I wonder what the “self-driving” Renault will make of a flock of sheep!

        The Google system is all about learning from all interactions across many vehicles so it can be trained to deal with a flock of sheep, just as it was trained to deal with a track-standing cyclist after initially getting stuck.

  25. Why would someone who lives alone want to spend on the fuel to send their car home to park? If I had a driverless-car, most of the time I would expect to part it outside of where I was. Especially if I’m out shopping, and want to place things into the boot for safekeeping several times during the day.

    (Right now, I don’t own a car because the cost of inner-city parking outweighs the convenience of my own vehicle vs public transport waits. Even so, I do rent a car occasionally and the main benefit is not having to plan the day at a micro level like I would if I used taxis.)

  26. The practicalities of any system coming into New Zealand is that we will have to wait for them to trickle here second hand from Japan. We are already seeing it with hybrid Toyota Prius and Camrys, which are coming to dominate the better taxi fleets, but these are way out of the price range of the average family with their Serena or Estima people movers which dominate the South Auckland supermarket carparks. There is also a problem with the batteries in these second hand imports, and we have already had publicity about unqualified people doing botch up jobs on the batteries to keep the vehicles on the road, only for them to fail in spectacular fashion. Another issue with driverless cars, is that while they may not hit anything, they are still vulnerable to being whacked from behind by a manned vehicle, driven by someone who doesn’t have the reflexes of a computer. All the discussion on driverless vehicles revolves around city people using them to commute to and from work, but my vehicle (a 23 year old 4 x 4) gets used for a lot more than that, and what about the country folk coming into the city. Would I buy a hybrid or fully electric vehicle? At present they are way out of my league price wise, and I can’t see that changing in the foreseeable future, so I will stick to my Bighorn, which is a perfect fit for me, my work and my leisure driving.

  27. Driverless cars, surely, will have to be – will only be – electric. And probably solar powered to some extent. I can’t see industry putting in all the effort to provide petrol – or hydrogen – automatic refuelling stations for driverless cars. Too dangerous – or too reliant on humans.

    If you want a driverless service to rock up to your house and txt you that it is outside waiting, you’ll also want it to go away and recharge itself when it gets low on charge before it picks up its next passenger. That leaves us with only 3 options – either a friendly human plugs it in (but why would you if it is not yours?), or it plugs itself in (feasible, if it can sense where a public recharging spot is), or it just parks up in the sun and a PV roof panel recharges the car. When you think about it, the last option is the only really valid option.

  28. If everyone will ditch their private cars and share driverless ones, that will mean a reduction in cars. And therefore, a reduction in the road tax take. How will the motorways be funded? Or will new ones be needed?

    Driverless cars as a disrupter is bordering on being a myth. Imagine everyone moving to uber overnight. Might address congestion on the roads initially, until the demand at peak results in an increase in supply, leading to congestion and road pricing….. in the meantime, smart authorities will have expanded PT networks making the cost of using the driverless cars redundant.

    At best it will address drink and dangerous driving but by then, most people will be taking cheap PT even that advantage will be diminishing.

    Time to move on. This space is being doninated by anti-PT types who believe driving a car is the ultimate freedom but in the same breath advocate handing that over to a machine. Its just moving from the front seat to the back in the same vehicle.

    1. Time to move on? No. This is a transport blog, and people enjoy thinking about, blogging about, and talking about transport. Better to thrash out possible problems and possible solutions here in the ether, than to blindly rush ahead or regulate in the face of ignorance about a subject that hasn’t been thought properly through.

      1. I’m all for the debate, it’s why I am here. But if you are looking for a solution, perhaps define the problem first. If its congestion, driverless cars are not the answer based on the above post and the numerous ones before it.

        I’m not saying the technology isn’t great and that they would not add value – for safety, drunk driving, etc,. But in the end, its replacing cars with cars. Only the driver is changing.

        For congestion, there are better, more immediate solutions that air could be better wasted on, IMHO.

        1. The “problem” is that driverless cars are coming, whether we like it or not, so we need to prepared for whatever their effects might be. Potentially, much much more than “just the driver” is changing.

          1. Which is a different set of potential problems specific to driverless cars, than general congestion. And I would agree, one that needs thought.

            My “time to move on” comment was in relation to those suggestions they will be a solution for congestion and therefore the need to defer any spending on PT. In the context of general traffic congestion, it’s replacing like for like.

        2. I wrote a discussion paper in 2013 which set the challenge on how to get the motorway moving in less time and cost than other methods. I investigated a number of different scenarios including driver less cars and came to the conclusion that the form factor of the peak vehicles needed to change to single commuter vehicles.

          The best paper I have read on driver less cars is by KPMG.

          “An essential implication for an autonomous vehicle infrastructure is that, because efficiency will improve so dramatically, traffic capacity will increase exponentially without building additional lanes or roadways. Research indicates that platooning of vehicles could increase highway lane capacity by up to 500 percent. It
          may even be possible to convert existing vehicle infrastructure to bicycle or pedestrian uses.”

  29. Victoria Crone seems keen on driverless buses “Her pitch is that Auckland may miss the [driverless] bus unless it has a leader who is chasing the future.” but if there’s no driver, who’s going to answer the question “does this bus go to (the place shown on the destination display) or “how do I get to . . . . ?” a significant proportion of the driver’s role is to provide information, assist elderly on and off, settle school kids etc. who will check the bus for lost property if there’s not driver? seems like a bandwagon where the only musician is playing a kazoo (badly)

Leave a Reply