IAG: Driverless cars 20 years away

David Harrington from IAG channels Upton Sinclair: “It is difficult to get a man to understand something, when his salary depends upon his not understanding it”. From The Australian:

Australia has 700 pieces of different regulation that need to be changed before driverless cars can take to the roads, according to Insurance Australia Group, which says the advent of self-driving cars is well off into the future.

Speaking at an investor day in Sydney yesterday, David Harrington, IAG’s head of strategy said only 2 per cent of all vehicles on the road in Australia and New Zealand had “assisted” driver technology — meaning the take-up of fully automated cars would progress far slower than some believed.

“So, we’re actually quite early even in the phase of driver assistance and so while we believe in driverless vehicles, they’re still quite a long way down the track,” Mr Harrington said.

“But we think that curve is going to move very quickly and so by the time you get to 2030 we expect that 50 per cent of all vehicles in Australia will have assisted-driving technologies and by 2040 it’ll be more like 90 per cent.

“It takes on average 19 years or 20 years to turn over the fleet in Australia and that’s how long it will take people to buy new cars, replace old cars and have cars that have this technology.”

David is right in one respect. Australia is a long way behind other countries –  the Infrastructure Australia paper from last year seems to recommend that we stay that way, and the Federal Government is doing nothing except considering a small cut to the luxury car tax.

I’ve written on this a few times lately, I would recommend listening to the execs from companies that are near the front (e.g. GM or Waymo) rather than anyone from an industry that will be decimated by driverless cars in a country that is waiting to see what everyone else does.

My view is that driverless taxis look like they will be cost competitive with public transport (see here for some calculations) – suggesting most people will never own a driverless car.

Quick stats for the day:

  • Waymo and Jaguar are launching 20,000 self-driving taxis in 2020/21.
  • GM are looking to launch “mass production” of self-driving taxis in 2019, probably targeting similar numbers.
  • There are another dozen or so companies not far behind.
  • Total number of yellow cabs in New York:  13,587 servicing 241 million customer trips per year.

I’m not sure when driverless cars will hit in earnest, but I’m investing like they will be here much sooner than 2040.

Damien Klassen is Head of Investments at Nucleus Wealth.

The information on this blog contains general information and does not take into account your personal objectives, financial situation or needs. Past performance is not an indication of future performance. Damien Klassen is an authorised representative of Nucleus Wealth Management, a Corporate Authorised Representative of Integrity Private Wealth Pty Ltd, AFSL 436298.

Follow me

Comments

  1. Consider the hook turns in postcode 3000. Even if driverless cars can drive on American roads, they would have to learn about hook turns and trams.

    It is far easier to convert one lane of road into two lanes of PRT. The PRT system at Heathrow airport is very safe and driverless. I doubt any of its pods has ever crashed. The pods just need to be put on steroids via a powerful lithium ion battery and electric motor.

    • All nonsense talk. In the 80s we kept getting told we were about to get flying cars, and robots to do the ironing. It’s about as likely to happen in the colonies as a fast reliable internet connection. Although I understand that Reusa has a home service robot to assist him with vital duties (latex).

    • Sorry mate it’s 2018 and the hook turn was last century’s flavor. Nowadays it’s all about P TURNS!

      P turns has got to be the mother of all band aid solutions imo.

  2. IMO people underestimating the amount of worms in this can:
    – Legal compliance costs. Different legal requirements in countries and even states within countries. Remembering that the software algorithm is predetermined. What would normally be considered an accident by a human driver will now be considered premeditated. The cost of coding and testing laws from different jurisdictions will be astronomical.
    – The complexity of the technology. Take a look at the progress (or lack of) in the self-driving car racing tournaments. Note these are controlled environments.
    – The inevitable terrorist GTA software upgrade

    • rob barrattMEMBER

      Not just the terrorist hacks. How about the Robot Road Rage mod. Just splice it in and watch that baby rock….

    • Before cars, roads were a shared space. Before cars, your identification was government papers, not a car license.

      Money solves a lot of problems. People stepping in front of your machine? Call them “jays” and get the government to make it illegal.

  3. reusachtigeMEMBER

    I really don’t care as I’m happy with my current driver. He’s good to chat with, keeps things confidential, and takes my enemies for a drive when I need him to.

  4. Just consider how much political malarkey was involved in the conceptually simple legalisation of SSM. Our sclerotic lawmakers will take decades to develop a workable legislative framework for self-driving vehicles, and the starting gun will only go off when self-driving taxis are being used in earnest somewhere in the world.
    If anything, Harrington is staggeringly optimistic.

    • C.M.BurnsMEMBER

      the lobby group driving for SSM was not for profit.

      the lobby groups that will drive for driverless cars will represent huge, for-profit companies and global giants. there will be a huge financial incentive and all sorts of lobbying money behind the push for driverless cars. it will happen much sooner than people want to believe, because of the huge profits that lie at the end.

  5. drsmithyMEMBER

    My view is that driverless taxis look like they will be cost competitive with public transport (see here for some calculations) – suggesting most people will never own a driverless car.

    From a consumption perspective a driverless car is functionally identical to a taxi. Taxis have not displaced widespread private car ownership, so it is hard to see why driverless cars would, all else being equal.

    A larger driver for abandoning private vehicle ownership is likely to be increasingly dense residential environments, in particular the lack of parking.

    Anyone with kids should be able to tell you fairly easily why they won’t be giving up their private vehicle.

    • Even StevenMEMBER

      Why would it displace private cars given taxis have not? In one word: “Cost”…

      • Taxis are cheaper than private cars already, so clearly “Cost” is not a driver of this.

      • drsmithyMEMBER

        Let me put it another way: most people don’t make a transport decision based on running cost TCO. It’s the intangibles that attract them to private vehicle ownership.

        EDIT: Also what bjw said.

      • For a little while I had a baby, a car seat, and no car whilst being the member of a Car Share initiative. Every time I wanted to go somewhere, I re-installed the car seat. It was even crazier than it sounds.
        There is no chance that driverless taxis are going to be of interest to the vast majority of parents of children aged less than 7 (and given different child car seats for different age groups, the loss of boot space, and the loss of room for any other passengers, precious few driverless taxis will ever have child seats installed, just as, as far as I know, zero taxis have them installed now).

      • Even StevenMEMBER

        “Taxis are cheaper than private cars already, so clearly “Cost” is not a driver of this.”

        Wrong. Cost is a huge factor. I accept convenience is also a huge factor.

        We drive one car (not two) because of cost. Even though occasionally it is inconvenient.

        We don’t use taxis because they are still too expensive to replace the amount of driving we do in our car. Uber is *almost* changing that. If we ever wanted to do longer trips, we would then just hire a car for the relevant period.

        Rego, Insurance, Servicing, Repairs… and the big one… Depreciation.

        You’ve got to be kidding me if there aren’t already loads of people who are already near the tipping point of “its just not worth it”.

        Make it more convenient, ubiquitous, cheaper. The take up will be huge.


      • If we ever wanted to do longer trips, we would then just hire a car for the relevant period.

        Used to do that for an family holiday. Nightmarish, even before kids. Absolutely worth having a car in front of the house to avoid that particular horrorshow.

      • Even StevenMEMBER

        @ Robert

        Why so nightmarish? Take a driverless car to the local car hire place. Pick up said car, return home. Pack for your week of camping. Simples.

      • Like I said, I tried it multiple times with current car hire services. If you don’t know, you just don’t know.
        The cost structure would need to change radically, so that it didn’t cost an extra hundred bucks to get the car 24 hrs prior to the trip, otherwise by the time you’ve screwed around getting the car, and then packed the car from scratch on the day of travel, and, of course, installed two child seats, you’ve lost half your travel time, and have just packed a car in full sun in the middle of summer, and are about to leave mid-afternoon on a 38 degree day.
        Been there, done that, never again .

        and the big one… Depreciation.

        Depreciation is voluntary – it can be mostly be avoided by buying a second hand vehicle five or more years old.
        If you experience depreciation noticeably greater than your rego costs, it means you like having a new car in your parking space to signal your self-worth, an opportunity driverless taxis won’t give you.


      • What if someone hires your car and vomits on the back seat

        Exactly as car hire companies do now – make the customer sign a document agreeing to a $800-$1000 insurance excess, and offer to waive it for about 20% of the car hire cost. If the customer fails to pay the excess waiving fee, charge them the full excess if any damage or uncleanliness can be found, no matter how minor.

      • HadronCollisionMEMBER

        What Robert said

        Are we talking driverless cars or car sharing/hire cars

        Very different

        As someone living in the country with other people living in the country, driverless cars != useful or on the horizon

        Especially rural areas

        A man needs a ute to get around the farm
        And pick sh$t up from the farm shop

        However I did say upon seeing iPhone 2 it would never take off due to lack of keys so disregard anything I say about anything

      • Nokia is a prime example of what happens when you do not innovate and invest in R&D. They have gone from a market leader to trading on some retro boutique niche. When someone thinks phone, they think Apple, Samsung, Sony. They don’t think Nokia, Blueberry, Motorola….at least, they don’t anymore 😉

      • drsmithyMEMBER

        People driven cars will survive decades after self-driving cars that work perfectly are available and accounted for by legislation.

        I wouldn’t be remotely surprised to see cars piloted by puny humans legislated away at that point, corner cases (offroad, rural) aside.

      • Nokia is a prime example of what happens when you do not innovate and invest in R&D. They have gone from a market leader to trading on some retro boutique niche. When someone thinks phone, they think Apple,

        Funny thing is that the notion that an iPhone was the cool new thing to have gave it a big kick along in the beginning – now that people as uncool as me have them, their sales are falling…and cool people have to buy retro Nokias to signal true hipster fashion status, hence the revival.

    • So you guys have never lived in a place like London, NY or Singapore? I am not sure why you can’t imagine a world without people owning cars!!

      Cars will be linked to each other (and trams) so hook turns will be easy. Accidents will be non-existent so you don’t need seat-belts (you don’t have seat belts in a train or tram). Also cars like a Volvo have ‘pop-up’ booster seats built in.

      • drsmithyMEMBER

        I can quite easily imagine scenarios without people owning cars, I lived without one myself for years and it’s already commonplace in very high density areas.

        But I struggle with the idea that NOBODY will own private vehicles.

        And this suggestion that accidents will magically disappear is ridiculous and counter to everything we know about engineering, programming and human behaviour.

      • And we’ll all be travelling in our flying cars, with robot maids keeping the house clean, and buying our spacely sprockets with a video phone.

        On that note, how often do you make or receive a video call?


      • On that note, how often do you make or receive a video call?

        Not sure where you’re going with that, but I’m sure everyone on this comments page has used Skype at least once, but hardly anyone uses it with any regularity, unless someone close to them lives O/S at the moment or similar. Video phones seems to be one of those technologies that was completely underwhelming once it got here.There is at least a chance, however small, that driverless cars will underwhelm the consumer just as much.

      • DodgydamoMEMBER

        I see plenty of people having FaceTime calls (on the windscreen mounted phone) while driving!

  6. Even StevenMEMBER

    I completely agree. The economics are so compelling (and also the eventual safety benefits) that this will be fast-tracked. Consumers will demand it and the Government will respond.

    • the Government will respond.

      Yes – as quickly as they do for any popular concern, including, for example, migration reform.

      • Five or six years to expand the existing long used and tested taxi legislation to include Uber doesn’t bode well for how long it will take to develop legislation for something like self-driving cars that is genuinely novel, and requires a bit of thought to anticipate the potential risks.

      • The long distance trucking companies will see the cost benefits of driverless and lean on the govt to allow it. The govt will allow it.

      • Damien KlassenMEMBER

        My expectation for rollout: (1) China or Singapore legislates to reduce the liability of the companies involved in crashes and encourages driverless cars on a large scale. (2) The US is offended that they aren’t first and follows suit. (3) The rest of the world trails behind.

      • Even StevenMEMBER

        And eventually driverless cars will be demonstrated and accepted as so safe, that driver cars will be banned in some jurisdictions. You heard it here first! 😁

        Admittedly, that will take some time, but it will happen…

    • Insurance costs will drive it. It will get to the stage where the accident rate for driver-less cars is non-existent that insurance companies won’t insure people driven cars.

      Think of the disruption coming for the car and car service industry once cars are electric (2025-2030 for cost parity) and driverless. There will be almost be:
      – no car service industry (almost no moving parts in an EV)
      – no petrol stations
      – no panel beaters (approaching zero accidents)
      – no car dealers (almost no private ownership)


      • It will get to the stage where the accident rate for driver-less cars is non-existent that insurance companies won’t insure people driven cars.

        What will they insure then? If self-driving cars are that accident free, no one will need to buy insurance.

      • Even StevenMEMBER

        @ Robert. Correct! There will be far less need for insurance. I think that is Damien’s point!!! Hence the Upton Sinclair quote.


      • – no car service industry (almost no moving parts in an EV)

        I guess EV’s won’t have wheels attached to suspensions or gearboxes etc. Or perhaps you’ve just never worked in an environment where rotating machinery driven by electric motors needed to be maintained.

  7. darklydrawlMEMBER

    One has to consider IAG’s own profitability and product offering when this technology becomes widespread.

  8. Or alternatively, lets see how pilotless passenger aircraft have gone. The technology for completely pilotless airliners has existed for quite literally a half century, along with flying an aircraft being a much simpler problem, given that traffic management is handled externally to the aircraft, and obstacles are virtually non existent and mostly non avoidable even by human pilots given the travel speed.
    and yet, all airliners still have pilots. I think the uptake of autonomous cars is being massively overstated if the history of the uptake of similar systems in the much simpler aircraft space provides any indication.

    For an insight into the length of time autopilot tech has been around https://en.wikipedia.org/wiki/Autoland, first certified aircraft 1968.
    https://en.wikipedia.org/wiki/Autopilot
    The first aircraft autopilot was developed by Sperry Corporation in 1912…..The installation of autopilots in aircraft with more than twenty seats is generally made mandatory by international aviation regulations.

    • What about crew-less shipping? Is there a technological reason why a cargo ship needs half a dozen or so bridge officers to steam in a straight line in the open ocean? (note that pilots board vessels as they come into port so reason a pilot and a few crew couldn’t do the same to navigate port entry on an automated vessel)

      • I’m ever so sure that the Somali pirates are going to give your whiz-bang automated ships a run for their money.

        Because – what are you going to do? Deploy lethal force? how? where from? how quick?

        (mostly) everyone can drive a car in an empty car-park…

      • Cargo crews are unarmed, and so present zero barrier to a pirate intent on boarding. Indeed, the absence of people working for the shipowner just means that the pirates have a bit less leverage.
        Indeed, in the absence of a crew, the owner could remotely scuttle the ship once all the pirates were aboard, which I suggest would be a powerful deterrent to future piracy, given the consequences for the pirates’ life expectancy.

        I guess my point was shipping ought to be easier to automate than cars, yet there’s very little progress, and piracy risk is not a serious barrier.

      • Ronin8317MEMBER

        Mostly because the ships are poorly maintained, and human crews are still cheaper than robots when something needs to be fixed. Also if no human on board, going onto a ship and taking all its cargo will not involve human confrontation, and a lot more people will become pirates.

      • “I guess my point was shipping ought to be easier to automate than cars, yet there’s very little progress, and piracy risk is not a serious barrier.”

        Moving ships/boats through water is actually extremely complex! The water moves, waves happen, the air moves, the weather changes unpredictably. Try plotting a course on a yacht capable of 10 knots when the tide changes the water speed from +5 knots to -5 knots with a 5 knot cross wind (this is a vector calculation). Then half way through the cold front appears and changes the wind speed and direction and the waves pick up which change the direction your boat is pointing.

      • “Try plotting a course…..” Maths like this is what computers are extremely good at, far better than people. Identifying random objects that can have multiple shapes and appear at any angle or location and then predicting what this object will do and where it is likely to be in the future is what computers find hard, and is the exact problem automated cars are required to solve, while also being one humans solve as a matter of course, and therefore mostly fail to understand just how complex a task it is.

      • “Any calculation you can do, a computer can do better.”
        Abstract maths, almost certainly. Complex applied problems, not so much. You ever seen a robot try to walk? something a 2 year old manages without too much trouble at all.

      • Either way it’s been compulsory for ships to have working charting software for nearly decade, years after such software was carried by all but the slackest vessel.

      • drsmithyMEMBER

        Abstract maths, almost certainly. Complex applied problems, not so much. You ever seen a robot try to walk? something a 2 year old manages without too much trouble at all.

        Ever seen a person try to fly a modern fighter jet without a host of computers sitting in the middle making it possible ?

      • Given the pilot and the G load that they can withstand is the performance limiting aspect of the aircraft, why do you think it is that they still have pilots?

      • drsmithyMEMBER

        Given the pilot and the G load that they can withstand is the performance limiting aspect of the aircraft, why do you think it is that they still have pilots?

        That’s completely irrelevant, but I’m sure you know that.

        To answer the question, fighter planes still have pilots mostly because the air force is full of people who want to be in Top Gun.

        How many drones do you think there are airborne in the Middle East right now ?

      • The reason is completely relevant. People are infinitely better than machines at making decisions based on incomplete information or in unexpected circumstances. Causing a vehicle, be it plane or car to follow a pre determined course is something a computer can do better than a human, but interacting with and reacting to the outside world that it has no control over is something that is far more difficult. Regarding drones, I expect they would be seeing much less use if they were being operated against an enemy that posed any sort of threat, as they would be being wiped out. How many autonomous vehicles are the military operating on the ground, where the threat is much greater, and more friendly lives could be saved?

      • drsmithyMEMBER

        The reason is completely relevant.

        No, it’s not.

        Your claim is that a computer couldn’t possibly do anything as “complex” as navigate a boat (though a quick Google for “autonomous sailboats” indicates that they can).

        My point is that modern fighter planes cannot be flown WITHOUT a computer to manage the control surfaces. Humans lack the cognitive ability and reaction time to do it.

      • wow smithy, do you have any reading comprehension skills?
        Ive said planes are fully capable of being flown autonomously, and have been since before transistor based computers were available. Despite this, all the passenger planes in the world are still being flown by pilots, and not autonomously. In regards to autonomous cars, driving the car is a simple task and anyone could write software to do it. The hard part is getting it to drive somewhere without hitting stuff.

      • The hard part is getting it to drive somewhere without hitting stuff.

        Which they seem to be capable of doing at least as well as the typical person.

        Your example was setting and maintaining a course on a boat. That was what I was responding to.

        The plane comparison is silly, not least because of people’s irrationality around flying.

      • “Which they seem to be capable of doing at least as well as the typical person”
        Please point me to an actual demonstration of this, I am yet to see one. I’ve just heard a lot of promises from people with vested interests.
        Also peoples irrationality is one of the reasons there won’t be any autonomous cars driving around freely for a long time yet. This is why the plane analogy is so apt. It won’t be technical limitations that ultimately restrict autonomous cars, it will be legal and social issues.

    • Even StevenMEMBER

      Thank you for the history lesson. History is no replacement for understanding the rapid adoption and deployment of technology.

      I am pleased there are – apparently – many people thinking like you. Otherwise Damien / MB Fund wouldn’t be able to capitalise on such investment opportunities to the same extent.

    • bjw678, yep.

      The autopilot on QF72 went psycho:

      It went 150 feet down in two seconds.

      He remembers wondering if his life was going to end. Seconds later, he finally felt responses to his control-stick movements and he is able to bring the aeroplane back to 37,000 feet. But then, the plane dove again – 400 feet in just over 15 seconds. He and the pilots realise that one of the three flight control primary computers (PRIMs) is faulty. He was concerned about how he would safely manage an emergency landing, but knew that continuing on to Perth could be even more devastating for the injured. Mr Sullivan declares a mayday and puts ‘Learmonth Airport’ into the computer for navigation – which shows an error. At this point, he says he became enraged and started ‘cursing like a drunken sailor.’

      He said he lowered the jet’s nose and powered to idle as he began the final approach without all the necessary instruments

      https://community.infinite-flight.com/t/the-untold-story-of-qf72-what-happens-when-psycho-automation-leaves-pilots-powerless/116641

      Madness. There should be a manual override lever.

    • Ronin8317MEMBER

      The human pilot is the backup. Would you fly in a plane where a frozen pitot tube (used for air speed), which happens very often, means everyone on board will have to die? A pilot is and crew is not that expensive at all in the scheme of things.

      • would you go out in public when the slightest glitch in any electronic system related to driving an automated car could virtually instantly result in fatalities as a car crossed to the wrong side of the road, runs someone over or crashes into a solid object quicker than any human is even able to react, rather than a mild loss of altitude as in the case of an aircraft with a failed autopilot?

  9. it would be interesting to see what actual problem self-driving cars are solving at what cost (including externalities), who is paying for it and what are unintended consequences.
    Also who is going to make money on them and how much of that profit is going to flow into the society.

    • This is what fascinates me about it – it’s costing 10s of billions to develop this technology, but given the typical use of a car is to transport an adult who can drive, what is the marginal benefit over the existing technology? Is there one at all? Certainly it’s an order of magnitude smaller than the marginal benefit of the first cars over horses.
      Presumably the R&D cost has to be recovered by an increase in the cost of a car – but why would anyone pay that cost if they could drive, and adults able to pay the cost differential who can’t drive seems like too small a set of people to recover the cost, and the driverless taxi model that ultimately sees a huge decline in annual car sales if it succeeds just compounds the problem – a car manufacturer who has invested heavily in self-driving technology must be assuming there are decades of private car ownership ahead to recover the investment.

      • it’s not only that but externalities as well: unemployment, increased usage of cars , pollution, natural resources, than legal issues where gov will need to step in and provide insurance, than all the issues related to social vulnerability via dependency on this and related technologies (gps, mobile network, …) – this goes as far as national security, costs related to hacking, viruses, loss of skills …

        there is a whole mountain of problem looming but why? to fix the problem that’s barely a problem or no problem at all – large percentage of cars are already owned just so that people can enjoy driving


      • large percentage of cars in Europe and Japan are already owned just so that people can enjoy driving

        And it’s virtually impossible to imagine that any of those people will give up their car, or take driveless taxis at a greater rate than they currently take taxis. So how do the auto makers recover their development cost?

      • drsmithyMEMBER

        But you could make that argument with every driver aid ever invented. Auto transmissions, ABS, cruise control, etc.

      • All those things make driving more enjoyable, and it’s never been claimed for any of them that they will spell the end of private car ownership, and therefore massive drops in the number of cars being sold, moreover, I’d conjecture that the R&D costs involved were considerably lower, and they probably didn’t have much external cost in terms of unemployment or government liability attached.
        Like I said, an auto maker that invests in this technology is going very long on private ownership continuing at similar levels into the future.

        I note that the automatic transmission has been available in passenger vehicles since around the end of WWII, but manuals haven’t bitten the dust entirely, so the use of that example implies a very gradual uptake of self-driving cars.

      • drsmithyMEMBER

        All those things make driving more enjoyable, […]

        So do autonomous vehicles. You can now recover the hours of your life otherwise spent unproductively steering a car around. And that’s not even looking at the safety benefits – road deaths haven’t trended steadily down for decades because drivers are getting better.

        The percentage of people who enjoy driving for the sake of driving is small (and I say that as one). The percentage of people who enjoy the things driving their car delivers (independence, status, convenience, etc) is quite large.

        Like I said, an auto maker that invests in this technology is going very long on private ownership continuing at similar levels into the future.

        Well, no disagreements from me there. I personally think the widespread death of private vehicle ownership is grossly overstated.

      • That argument frames driver as a marginal benefit, initially offered at a sizeable markup as an extra and therefore with incomplete market and only gradually becoming universal.


      • The percentage of people who enjoy driving for the sake of driving is small

        True enough, and I’m not one of them (although I hate being a passenger in a car ten times as much as driving one) but it is large enough to support multiple manufacturers who only service that market as premium luxury good, and in the event that driverless cars become widespread, I suggest it will be large enough to support one or two additional manufacturers who service a sub-premium luxury good market (for people who like driving but currently just drive their commuter vehicle to drive around on the weekend), just as you can buy things like quad bikes and motor boats for a working class budget. (I personally will have no use for such a vehicle, and will probably resist getting an automated vehicle as long as possible due to a combination of stinginess and paranoid need to be in control, as long as driverless cars come in before my eyesight falls victim to inherited macular degeneration)

    • Damien KlassenMEMBER

      Cost, safety and utilisation.
      There are 19m cars in Australia travelling 15,000km on average per year – cars are parked 95%+ of the time.
      The average household in Sydney (with 2 cars) is spending $22,000 per year on cars. https://www.aaa.asn.au/wp-content/uploads/2018/03/AAA-Affordability-Index_Q4-2017.pdf

      Fewer cars = less money spent on depreciating assets, more roads (less street parking needed), lower property prices (converting indoor parking spaces into other uses), better safety (eventually), plus making a dent in the $22k per year.

      Like it or not, there is plenty of economic motivation.

      I agree that driverless does not make sense for car manufacturers as they will all lose massive volumes. But, the cat is out of the bag now. If they develop driverless they lose 80% of production. If they do nothing they lose 100%.

      I completely agree that this will create massive issues for employment. There were lots of blacksmiths that lost jobs when horses were replaced. In the long term it will be a good thing. In the short term there will be major problems.

      • Seems logical.

        Smacks of Smart Phones to me. People were so caught up in existing technology (everyone already had a PC and Nokia), that they couldn’t see the mass appeal of a phone that effectively combined a suite of technologies.

        We will see driverless vehicles sooner rather than later, and looking back, we will wonder how we ever got around with 1000s of lunatics driving a tonne of steel at a 100 km/h.

      • So short all forms of auto making, as if you are right 90% of current auto making capacity will be redundant by 2035, and the imbalance will be addressed by car makers ceasing to be; and the answer to my original question is that there is no expectation the R&D investment actually pays off.

      • Damien KlassenMEMBER

        The R&D is a hard one. Its hard to build a business case on spending the money, but I guess its easy to build a business case to try and protect existing revenues…

      • 100% will be a return on R&D investment. Any automaker that doesn’t invest in automation will go the way of the dodo, having lost a critical market edge.


      • Any automaker that doesn’t invest in automation will go the way of the dodo, having lost a critical market edge.

        9 out of 10 will go the way of the dodo anyway if the scenario posited comes to pass – why not save the money and cash out now?
        In the mean time I suggest that a niche market for human driven cars will exist long into the future at 1%-10% of current volumes, if the posited scenario comes to pass, just as vinyl records continue to be made. Deciding to go after that market and ignoring the AI thing is a completely valid strategy.

      • That’s not how a large multi-national corporation works. They don’t look for ways to fill a shrinking niche with dated technology. They at all times seek to maximize returns by claiming market share. To claim market share you must be competitive. I reiterate, any automaker that doesn’t invest in automation fast enough will be left in the dust.

        Once uptake of driverless cars takes off, no one is going to buy a car that doesn’t have this feature built in as an option.


      • That’s not how a large multi-national corporation works. They don’t look for ways to fill a shrinking niche with dated technology.

        There are pretty obviously dozens to hundreds of large companies whose strategy is to fill out of the way market niches with specialised offerings.


        Once uptake of driverless cars takes off, no one is going to buy a car that doesn’t have this feature built in as an option

        Because obviously no one in the history of auto manufacturing has ever bought a hopelessly impractical car for way too much money just to tell the world they can.

        http://autoweek.com/article/car-news/guess-where-these-two-gold-infused-rolls-royce-phantoms-are-headed.

        Just as a two seater sports car that goes 200 km/h faster than is legal to drive anywhere is a status symbol now – especially with less practical sequential gear shifting – if self-driving cars become standard, human driven cars will be a status symbol.
        And even if they don’t, you’ll be able to buy the technology from one of the broke auto makers in 2030 for 1% of what it cost to develop it (the cost of buying out the redundant auto maker).

      • I can’t think of a single market leading corporation that willingly gave up it’s market share to fill a niche in terminal decline. If so, they no longer remain a major player in their respective industry.

        IP theft may be common in China, but it tends to be frowned upon in developed economies. They either pay a premium for the right to use someone else’s hard fought for tech or they spend hard earned and lost sales playing catch-up.


      • IP theft may be common in China, but it tends to be frowned upon in developed economies.

        Usually if you buy the company, you get the IP. And no doubt in a world with car sales dropping like a stone, the car makers who did develop the tech will be prepared to licence it to you if you can agree on a price.


        I can’t think of a single market leading corporation…

        By definition, there aren’t all that many market leaders. And if you’re a market follower in a particular market, moving to an under-serviced market it better than going bust.

      • Which means you’re paying them for their R&D + mark-up.

        Yep, which is to say any major automaker that doesn’t invest in automation R&D will slide into oblivion

      • Well, unless they develop an alternative market, and given a 95% drop in car sales, pretty much everyone who DOES invest in R&D will go bankrupt also, hence why I’m guessing that buying out someone who’s got the tech will be pretty cheap, and why if you’re currently second or third tier, getting out completely could be the best option.
        There’s only going to be one or two winners – if you can recognise it won’t be you ahead of time, you’ve got some time to think of an alternative.

      • How is the economic motivation any more significant than currently provided by taxis/ public transport/ uber?
        And even more relevantly, if economics was at all relevant, then BMW Audi, Mercedes Tesla, Ferrari and all the other “luxury”car makers would not exist as everyone would be buying the most cost efficient vehicles rather than these overpricced status symbols. I find it ridiculous that you think that somehow when driverless cars appear, if they ever do, that it will all of a sudden become important what total cost of ownership for cars is when it clearly has never been before.

      • drsmithyMEMBER

        22k/yr is 60/day.

        That doesn’t sound like much to pay for the Uber equivalent of private car usage.

      • @DrS – note that the $22k was for two cars.
        $30/ day is really hardly anything to pay for the convenience of your car, ready for you, with your stuff already inside it, able to loaded with whatever you can fit in it a week in advance if that’s what you want to do, customisable to signal your unique personality, and able to be driven with zero notice.

      • why people confuse driverless technology with car sharing model?
        who guarantees there will be less cars? car sharing (with and without driver) is already an option and not very successful one. This will especially be the case if car prices become so much higher due to R&D, reduced volumes, …
        In addition, driverless cars will make more crowd (everyone will want to get to the CBD by a car if they don’t need to pay $50 for parking)

        people are not rational beings and they don’t do things because it’s more economical (especially if gains are only marginal). In many cases technical solutions are not really the best ones and they always crate other often much worse problems. Just have a look into into current existential issues of humanity (pollution and global warming, nuclear proliferation, privacy and spying, near death or liberal democracy, mental health issues caused by social media, … ). Very few of technologies causing these problems solved any real problem in the past (think of IT for example) yet created major problems that may cause great suffering in the future

        Building a channel tunnel may seem like a good solution to a problem of traveling 4 hours but spending all that extra money to provide people on ferries nice meals and entertainment would have costed much less and it would make people more satisfied (people are now already as unhappy with 2.5h commutes as they were with 5 hour commutes)


      • why people confuse driverless technology with car sharing model?

        To make it seem a much bigger deal than it actually is. Without the almost-free-driverless-taxi-leads-to-90%-fewer-cars-on-the-road aspect, the many people who normally drive about twenty minutes a day or less (e.g. drive to the nearest train station and get PT rest of way to work, or live in same suburb as work) will barely notice the difference. Hence my scepticism that it will be ragingly popular overnight to pay an extra $15k-$20k to have this tech in your car as soon as it’s available.


        In addition, driverless cars will make more crowd (everyone will want to get to the CBD by a car if they don’t need to pay $50 for parking)

        Not sure about that one – congestion in the last say 3-5 km into Melbourne is already high enough to make driving to a train station a better option for many, and finding a place to drop off outside my building in the city is already next to impossible, and probably same for many others.

  10. I’m wondering about the practicalities – say you take your driverless car to work, does it drop you off and park on its own? What if you pay for parking; wouldn’t it be cheaper to send your car back home? Then it has to come back to pick you up so it’s on the road for twice as long as if you drive it and park it, leading to more congestion.

    • Damien KlassenMEMBER

      Agreed – another reason for driverless taxis, not driverless private cars. If a driverless taxi = $3 per trip, would you bother owning your own car?

      • Personally, I’d still bother to own my car unless the self-driving car arrived at the door in less than a minute even if I didn’t have kids.
        Because I have kids, it has to arrive in under a minute with my preferred child seat configuration for me to give up my privately owned car.

  11. ” driverless cars will hit in earnest, ” accidental metaphor. Maybe not as long as 2040 but real mass takes longer than most innovation/IT forecasts and that goes back to the railway booms of the 19th century which took another 20 years to turn profits.

  12. Ronin8317MEMBER

    The economic of driverless personal car currently doesn’t work. If cost is the only issue, everyone would be driving a Mitsubishi Mirage. It needs to start with auto-pilot for trucks and public transport, as they spend a lot more time on the road. (which Tesla is doing)

    • If cost is the only issue, everyone would be driving a Mitsubishi Mirage.

      Hit the nail on the head, Ronin.

      If all people cared about was the economics of a purchase, we’d all be living in tin shacks, riding 5 man on mopeds and eating no frills baked beans. People buy things because they enjoy them, because they make their lives more comfortable, easier. No one needs an audi, or a bmw, or a range rover, yet people buy them none the less, despite vehicles that are magnitudes cheaper and of comparable utility.

      I don’t need air con or cruise control, yet I’m not clammoring for a cheaper car lacking these features (even were they still available).

      • You’re conflating someone with the rare luxury of avoiding a commute, with someone who lives that special hell that is the daily commute. The main reason people by a car is to commute from point A to point B, not for the fun of driving.

        We’re talking about mainstream….you can still buy a horse, but it doesn’t mean 99% of people are riding one to work.


      • The main reason people by a car is to commute from point A to point B, not for the fun of driving.

        If you’re spending more than what it costs to get an entry level Korean car, you’re after more than transport, especially if you’re buying a new car in a world awash with second hand vehicles.
        And horse ownership isn’t even as rare as the kinds of cars I was thinking of.

      • Driving is a chore required to get from point A to point B. They are buying these cars, with their flash features, to make this chore as tolerable as possible. Just imagine a feature that removes the most onerous aspect of this chore, the driving part itself.

      • The driving part is hardly the most onerous part of automotive transport – sitting in the car is. At least if you’re driving it’s not totally boring and you feel a sense of control.

      • If I didn’t have to drive, I could be working. Or watching a movie. Or crushing people’s arguments on MB 😂👍🏻

  13. The big implication is there will recessions in around 2030 in any country with excessive exposure to automating when big automakers fail overnight in the face of a 95% drop in sales. Every Australian should get on their knees and thank Uncle Tony for ensuring that won’t apply to Australia.

  14. Not going to happen unless there’s a signal embedded in the road that a car can track. And who guarantees that signal? GPS or GLONASS whatever can be turned off. Or ‘selective availability’ could be reintroduced as per the 90’s. Or – and this is the big one – a kid with a battery operated milliwatt transmitter could sit on a hill and swamp the GPS signals and just enjoy the carnage below. So much more fun than dropping bricks from bridges.

    • drsmithyMEMBER

      Indeed. It’s not like vehicles could fail safe, slow down, pull over and stop, right ? Soon as the GPS signal goes they burst into gigantic explosions, action-movie-style.

      • Not quite sure if you’re being sarcastic or agreeing with me.
        So the kid on the hill presses the button and your vehicle tries to pull over because it is now out of control. But there is nowhere to pullover to. Just 2 lanes and you’re in one of them. I’m happily driving along behind you and you are stopping. Now what.

      • drsmithyMEMBER

        Well, obviously, just plough into them at full speed, like you do today if a car in front of you slows down.

        You do understand that there are autonomous cars, *right now*, that don’t need a GPS signal to drive around, right ?

      • If you can give me a link to that non gps vehicle and it’s proven historical safety record. Then I stand corrected.

      • drsmithyMEMBER

        I don’t think any of the autonomous vehicles currently being operated use GPS for local avoidance, just for navigation.

        Or, in other words, even if their GPS goes away they are able to drive themselves safely to somewhere nearby and stop.

      • Drsmithy – I used to design life safety equipment and was a member of a Standards Australia technical committee for that product.
        It was our job to make sure to the best of our ability the product would not fail when called on. So we looked for any possibility of design or operational failure and addressed (regulated) those shortcomings via the standard.
        So that’s my background and all I’m saying is I see a various safety issues that have to be addressed re autonomous vehicles.
        That’s not to say it’s not a good idea. Perhaps we could try it first on our local trains and trams.
        Anyway thanks for engaging with me – appreciate it.

      • Well, apologies for the sarcasm, but the suggestion that autonomous cars which, today, don’t need or use GPS for obstacle avoidance becoming dangerous without it is silly. Vehicles with adaptive cruise control and AEB have already been on the market in Australia for years. Even lane-keeping technology is starting to filter down into the mid-range. That’s already got you probably 80-90% of the way to an autonomous failsafe.

        There are already multiple completely autonomous trains operating worldwide for 10+ years.