How automation will drive property prices

Because I do not own a car, when traveling a bit outside Sydney’s inner suburbs, I need a ride, and I’m happy to use a hire car — and very happy to employ Evoke Limos.

And it’s not just because I know one of the founders of this startup. It’s not just because they have an all-electric fleet of Teslas. It’s because every time I ride with them, I get a glimpse of the future.

A few weeks, when driven to Terrigal, Evoke sent along a Tesla Model X — that’s a nearly two hundred thousand dollar vehicle — and we were soon speeding along the M1 freeway toward Gosford.

At one point the driver asked me if I needed to charge my smartphone, then took his hands off the wheel to fish around in the dashboard storage, finally producing the requisite charging cable, and plugging it into the onboard power.

I was shocked — for just a moment. Then I remembered that we’re driving in a Tesla, and I’d heard the ‘bing’ sound meaning the vehicle had gone into autonomous mode.

To be honest, autonomous mode in a Tesla seems almost like magic. There we were, cruising down the freeway at 110 kph, doing whatever needed doing, while the car minded the road, stayed in its lane, and avoided obstacles.

It feels like the future.

And it is — to some degree.

That’s where people have gone overboard.

We’ve recently heard the CEO of Mercedes-Benz claiming that within a few years all automobiles will be autonomous, no one will be driving anymore, and so forth.

Here’s the thing: Mercedes-Benz is really good at making cars. They’ve been at it longer than anyone.

But they’re not as good at making clever computers.

An autonomous vehicle is a very clever computer on wheels.

That part — the part they’re not looking at — that’s the hard part.

Because we’ve all heard about self-driving Teslas, most of us have been led to believe that’s a solved problem — cars and drive themselves.

Now it’s just a matter of putting the pieces together.

It’s not that simple. And the best way to show that is by telling you another story.

In May I was MCing an event at Sydney Town Hall. Normally I’d walk to Town Hall from my home — it’s a bit more than 20 minutes on foot.

But I was dressed in my sharpest suit and it seemed much more reasonable to hire a taxi to drive me the 2 kilometers.

Now if you know Sydney, you know that it’s a straight shot down Broadway and George Streets to Sydney Town Hall. Easy as.

Or, well, it used to be.

Before the endless and ongoing light rail project went off the rails keeping George Street in a state of dynamic chaos.

When I got into the taxi, the driver and I had a good, hard think about the best way to get to Town Hall.

Neither of us were new to Sydney. But the prevailing conditions had changed.

The more dynamic the road conditions, the more difficult it is to automate.

And they’ve kept on changing.

It’s not exactly clear, from day to day, which sections of George Street will be open to traffic, or in which direction the traffic might be flowing. That’s evolved as the light rail project has stalled.

When we got to the intersection of Goulburn and George Street, we had a good think as we waited at the light. Could we go up a block to Liverpool Street, then down to Clarence Street, over to Druitt Street, and up to Town Hall?

Neither of us knew if all of those roads were open — and none of it was at all well marked. So we gave it ago.

We got there in the end, but I came away from that very short and very difficult trip with the clear sense that if it tested the limits of both a long time Sydney resident and a long time Sydney taxi driver then it would be well beyond the capacity of any computer we would be able to put into an automobile for the next decade — or more.

In perfect conditions — on an uncrowded freeway with plenty of visibility — an autonomous vehicle can perform well. Change those conditions even just a little bit, and things fall over.

Change them a lot — and George Street is a fine example of that — and, well, there’s just no computer anywhere that can deal with that. There’s no signage, no direction, nothing a computer can look for.

It could make an educated guess — computers are not so bad at that, these days, but the further it travels from what it expects, the more likely it is to make mistakes — and mistakes that weigh two tonnes can be very dangerous.

We were promised a nirvana of self-driving cars ferrying us everywhere, but, that’s further away than we’d like it to be. Certainly further away than it looks when you sit inside a Tesla on the M1.

On the other hand, in reasonably good conditions — not too much rain or fog or wind — an autonomous vehicle can handle highway driving admirably well. There’s a lot of Australia that fits those conditions — just not the bits where the people live.

So it is very reasonable to presume that within the next five years we’ll start to see entire fleets of trucks going fully autonomous. But only for part of their journey. A driver will take the truck to the limits of Sydney or Melbourne or Brisbane, get out of the cab, then send it on its way to its destination — on the outskirts of Brisbane or Sydney or Melbourne — where another driver will get into the cab and take the truck to its final destination.

This is how you can tell it’s the future — because the future doesn’t operate in binaries. It’s a bit of this and a bit of that. A bit of computer and a bit of human, each doing the bits they’re best at. The computer is going to do a great, safe job on the highway driving, while a human is going to do a great, safe job doing the city driving. Everyone wins.

You might think that will throw a lot of truckies out of work; that part’s not yet clear.

We know that the economics of shipping goods across Australia will change dramatically as labor costs vanish — and that could mean there’s more work for truckies to drive trucks within our cities.

But here’s what we do know: there’s going to be a zone of really cheap transport of goods and a zone where transport costs stay close to where they are today.

That’s where it presents some interesting commercial opportunities.

I did a bit of planning work for Liverpool Council last year. They’re preparing for the Badgery’s Creek airport to open — but the thing I find far more interesting is that they sit across the junction between the Hume Highway and Sydney’s highway system.

They’re at that magic spot where highway trucks will need to pull off and get human drivers — or where human drivers will depart those vehicles to send them on their way to Melbourne and Adelaide.

It means they’re going to be developing commercial resources in an area where ground transport costs are going to become very inexpensive.

And it means that any commercial real estate in the area that depends heavily on shipping or access to flows of materials — particularly from the southern states — gains a big competitive advantage.

So this is one lens I recommend that you to use when you evaluate a real estate investment.

Take a look at the property. How smart do you have to be to reach it via road? Is it dense and busy and difficult or simple and easy and clear? Where does it sit on that scale?

Ask yourself if you can improve the property to make it simpler and easier and clearer. That’s going to be a great investment — if the the property is situated somewhere that will be accessible to autonomous vehicles.

Eventually, all of Sydney, even the most difficult bits, will be accessible to autonomous vehicles.

If we don’t change the City much, that could take thirty years. And that’s not my opinion, that’s the opinion of the best roboticists on the planet, folks who have been working in this field for their entire careers. They know this isn’t an easy problem to solve.

It will be solved. Incrementally.

Think of these as zones of computational difficulty — equivalent to Vernor Vinge’s “Zones of Thought” — and you can see why our dense, old, weird urban cores are the last to receive autonomous land vehicles. That said, they may be the first to benefit from autonomous drones — if we can stand the noise they make!

Imagine a map of Sydney — the outside roads automate first, then the newest and least densely bits of the suburbs, then the older suburbs, then the more densely populated areas, then the inner suburbs, and — finally — the CBD.

My neighborhood, for all of its accessibility to the City, is one of the most difficult neighborhoods in Sydney to navigate, and so it’ll be one of the last neighborhoods in Sydney to receive the benefits of autonomous vehicles.

There’s one way this future might look a bit different.

We can expect cities like Singapore and Shanghai, with strong command-and-control governments — to rip up and redesign their ‘difficult’ urban cores to accommodate autonomous vehicles.

That’s the other way to do this — rather than wait until the computers get smarter, you can simply make the cities more accessible to not-quite-as-smart vehicles.

There’ll always be a question about whether this is the right thing for the people who live in the city — which is why it’s less likely to happen in Sydney than in Singapore — but, for larger commercial developments, think about designing from the ground up for autonomous vehicles.

Even if they never leave the area, a larger development that can accommodate autonomous vehicles will inspire the kinds of services that we’ll find very hard to offer in our jumbled urban cores for many years — and that could be an attractive drawcard.

The future isn’t all one thing. It’s not all robots, and it’s not all humans. It’s a mix. Where we set the mixture between the two, that’s up to us — that’s going to be driven by our needs and by what we’ve planned for with our investments.

If we want autonomy — if it makes good business sense — then we can build for it, or rebuild for it. If it’s not important, we won’t need to change much.

But the location of every commercial property has value now not just because of its foot or vehicle traffic, but because of how easy it is to get an autonomous vehicle to and from it.

That’s a new factor, and one to consider closely in all property investments in the years immediately before us.