Fewer Accidents, Fewer Cars, Faster Speeds: The Future Of Self-Driving Vehicles On Australian Roads

What are Australian roads going to look like in 10 years? 20 years? Whatever they do look like, they're going to have more autonomous cars on them -- whether that's a Tesla or a Holden or a Ford -- and there will be more car-to-car communication than an occasional angry yell. We'll still be responsible for the actions of our vehicles, but we won't necessarily have our hands on the steering wheel or even own them.

Self-driving car image via Shutterstock

We talked to Simon Wilson, the national Automotive Contact Centre manager at Allianz Global Assistance, about what Australian roads might look like a few years into a future where self-driving cars are the norm, and where Aussies are driven by robots rather than driving themselves.

Obviously this is still a long way off, and what we even think of as a car might change significantly before then, but it's an interesting thought exercise nonetheless. The future of driving is going to look very different to what we see today. And it'll probably be a lot safer.

Do you think self-driving cars will have a big impact on the way Australians use their cars? Will they still commute on public transport or will working while driving be more commonplace?

SW: I think self-driving cars are a significant and exciting development in the future of motoring.

Two developments that immediately come to mind are the potential safety improvements and reduction of congestion on our roads. By having these vehicles networked together to make the most of existing infrastructure could be the only way we can keep up with the growing needs of our population and the increasing pressure on our road systems.

In terms of productivity, ride sharing and logistics are also likely to see an improvement.

I predict that fewer Australians will own cars when this technology is perfected. The reason for this is the likelihood that car ownership will move towards the already popular pay per use model, but taking it to the next level.

This will enable Australians to get where they need to go with faster pick up times and as these vehicle numbers increase, there will be wider coverage and also the increased safety that will come with vehicles being networked together. In terms of public transport, it’s likely that routes will be more flexible and more in line with demand.

The rise of companies in the sharing economy, like Uber, Airbnb, Task Rabbit and the Lending Club is evidence people are looking to make use of the spare capacity in assets that aren’t fully utilised.

What is the effect that you see on accidents and insurance from self-driving cars? If humans are taken out of the equation, how do you apportion blame in a robot-on-robot accident?

SW: This is one of the burning issues that regulators face before self-driving cars become a reality, and there is a long way to go.

Accidents are likely to reduce by a very high percentage as the technology is perfected, though there will then be a temptation to increase the speeds of vehicles once people and regulators begin to have faith in their operation. Think high-speed bullet trains – if we were able to safely ensure that the stopping distance of a vehicle was maintained in an automated way – how fast could you safely travel?

I think the secret will be in the way they interact. If we can ensure that these vehicles travel close to each other but at a safe distance, there will be all sorts of advantages – fuel economy due to the tunnelling effect, fitting more vehicles within existing infrastructure in a safe way. The potential time savings in avoiding traffic snarls is really exciting!

I think what will be really exciting for the insurance industry here is the information we will receive – knowing exactly how customers are using their cars and where and when they are travelling, we will be able to more accurately calculate their risk. This could potentially mean the customer will be able to pay more precisely for what they use, reducing waste in the transaction, and potentially making pricing cheaper in some circumstances.

Do you see any potential in the increased danger of semi-autonomous cars? I'm thinking back to the urban legend of the driver setting cruise control then going to sleep, thinking it would steer for him.

SW: Like any new technology, there is the potential for it to be misunderstood. A prime example from when I first joined the auto industry 11 years ago was the urban legend about ABS causing more accidents because people thought it would lessen stopping distance -- my personal view is that the consumer has the responsibility of learning the technology fully before applying it.



    I've owned one automatic car and went back to manual because it's more engaging while the automatic was boring and encouraged bad lazy driving habits. I don't even like being passenger in a car when someone else is driving and find it incredibly boring. In fact, I've now moved nearly all my transport to a motorcycle because it's more engaging.

    Self driving can suckit. Literally the only appeal I can see is for massive daily commutes, and high speed internet should eliminate that.

    Last edited 16/03/16 9:10 am

      Should... but it's Australia you're talking about. After the National Botched Network has been rolled out, I doubt many office jobs will change.

      I'm all for automatic vehicles. Terrible drivers and painful long commutes should be a thing of the past.

      How is high speed internet going to eliminate daily commutes? Have I been using the internet wrong?

        Rubbish. People still need to go places. You are not going to share a car with three strangers so they will travel in their own car. The suggestion people will spread out their use is rubbish. Everyone goes home at the same time. They are not going to shift behaviour to off peak vehicle demand. 4pm time to go home.
        if you have vehicles to meet peak demand you have vehicles that sit idle for hours...and in an emergency they need to be there. Not in ten minutes or next available vehicle in your area or fifty dollars demand surge to secure your vehicle.

          Yeah, this will cause more cars to be on roads. All the people that can't drive will suddenly own one of these as well as all the other drivers that are normally on the roads. People will want to own their own cars because who wants to sit in a car that one hundred other stinky people have sat in it before you. No thanks.

      Do you think boredom would be reduced if all of the occupants in the car can talk to each other (even look at each other) without someone having to concentrate on driving?

        No, I do not. The last thing I want is to have to engage with people in the car. The car is my Fortress of Solitude, where I can listen to whatever I want as loud as I want, and that's how I will want it to stay.

          I guess you could still do that. I was thinking more about @dknigs comment about being bored. If you can truly concentrate on the people around you, which you can't really do when you're driving, would that help alleviate the boredom?

          I'd rather read on my phone or watch a TV show than stare out the windscreen on a morning commute. I don't find trips into unfamiliar territory boring though.

            I'm one of those rare people who can multitask and talk to people while driving, and know when to just ignore them if needed. But no, if nobody is driving, nobody is obligated to talk to them instead of playing with their phone, so I don't think it will help. You'll just see driver less cars going around everywhere full of people with craned necks looking down.

      I think that's the idea, these cars are for commuting and regular journey. Where there is zero driving satisfaction. V8 comes out on the weekend....

        Ha, you don't think your non autonomous vehicle will be priced to oblivion with taxes, registration costs and insurance costs once autonomous cars are the norm?

          If a driver's license cost $3000 a year, I would happily pay it. I love driving that much.

    The sooner we can remove the human element, the sooner we can move to the third dimension with autonomous skycars.

    Take my love,
    Take my land,
    Take me where I cannot stand,
    I don't care,
    I'm still free,
    You can't take my car from me

      I don't agree with the sentiment, but you win an upvote for Firefly reference.

    Generally, I'm all for autonomous vehicles. I think it will greatly reduce the road toll and levels of frustration experienced when in the car. I look forward to not having to loan my car to the kids (though I doubt the technology will be ready by then). As a cyclist, I get the impression that software will look out for me far better than some human drivers ever will.

    There was a discussion a while back, I think on Gizmodo, where a few commenters raised really good points around the decision making made by the algorithms, specifically what action(s) an autonomous vehicle would take when someone's life is in danger. In a scenario where the car found it had to decide between avoiding a pedestrian (possibly a child) and avoiding a wall/cliff/obstacle, which option would it choose? Would you be happy for your car to make that decision? Who would be liable for the resulting deaths?

    I would like to think that, with car-to-car communication and infinitely better reaction times of software vs a human, these situations would very rarely, if ever, arise, but fringe cases must be dealt with before the general populace will accept the transition to autonomous vehicles. I'd like these questions to be asked of the car manufacturers. I'd like to know why Elon Musk thinks.

    Last edited 16/03/16 12:00 pm

      Software reaction times may be better but the ability of even the smartest software to anticipate and avoid potential emergencies will never be as good as that of a human. e.g. A car won't see a kid chasing a ball until he runs onto the road, where a human driver might see the ball fly over a fence, 30m from the road and 200m ahead, and anticipate the child 10 seconds before the car would.

      Another example would be the parked car near the top of the hill. A human driver is likely to see it and change lanes 500m or more before the parked car - e.g. from the crest of Gladesville Bridge you can see parked cars at the top of the next hill, Lyons Rd, more than 1km away - but the car itself won't react until it is almost on top of it.

      Autnomous cars will do their best work in peak hour traffic, the rest of the time I think they will be quite hopeless.

        You should really watch some of Googles talks about the car and the view it has.
        They cover things like this. Like a self driving car spotting a kid running across his yard 500m in the distance, and then when the position of the kid is obscured by an obstacle a human would think the kid stopped already, and the car has already slowed down and made a course correction.

        The self driving car will see further than you, see more things from all directions and predict where they will be in 10 seconds better than you will.

        Both situations you described would be better handled by the self driving car than by you. This technology is a lot more advanced than you think it is, definitely check it out.

          "... when the position of the kid is obscured by an obstacle a human would think the kid stopped already, and the car has already slowed down and made a course correction." This is the other side of it, dumb cars slowing down for nothing all the time.

      From my understanding these self-driving cars are currently set to be very conservative/cautious. So the scenario of the car avoiding a pedestrian but consequently driving off a cliff/into a wall I feel is a low probability. The car realises the unpredictability of the pedestrian and will slow down to navigate safely past rather than continue to barrel along at 50/60kmh. Chris Urmson even shows a clip of the Google car encountering a lady in a mobility scooter chasing a duck in the street.

      These cars "see" more than I think people appreciate what they can. Based on a video from Sept last year the Google car has 360° view out to "2 football fields distance" (roughly 200m). Sure there's a heck of a long way to go with this technology but the early signs are positive considering it will most likely be 10, 20, 30 years maybe? until these cars become a feature on the roads. Car manufacturers and tech companies aren't sinking the billions of dollars into this tech to see it fail and a key component of that is gaining trust about the product.

      If you haven't already, it's worth having a look at a couple of YouTube vids to appreciate what the cars are about. Both vids share some of the same stories/vision but it's interesting either way.

      Google Self-Driving Car Google Presentation -

      Chris Urmson: How a driverless car sees the road -

        I'll have a look when I get home, thanks. The 360° FOV is interesting. Thanks.

Join the discussion!

Trending Stories Right Now