Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Standing Committee on Infrastructure, Transport and Cities
Current and future developments in the use of land-based mass transit

METCALF, Mr Gabriel, Chief Executive Officer, Committee for Sydney

STEWART, Mr Sam, Policy and Advocacy Officer, Committee for Sydney


CHAIR: Welcome. Although the committee does not require you to give evidence under oath, I should advise you that this hearing is a legal proceeding of the parliament and, therefore, has the same standing as proceedings of the respective houses. The giving of false or misleading evidence is a serious matter and may be regarded as a contempt of parliament. The evidence given today will be recorded by Hansard and attracts parliamentary privilege. I now invite you to make a brief opening statement before we proceed to discussion.

Mr Metcalf : I have been in this role for one month. I've been running an organisation called Spur, which is the counterpart to the Committee for Sydney based in San Francisco. I have been working in the field of transport policy and specifically doing a lot of thinking about autonomous vehicles for a long time. The Committee for Sydney is an independent think tank representing 150 member organisations working across a range of urban policy topics. I want to acknowledge Sam Stewart, who made the original submission and developed the report on our thinking about autonomous vehicles, which I'm going to draw from in these opening remarks.

We did a report, released last year, where we examined whether Sydney is ready for autonomous mobility. Part of that involved a survey of companies. One hundred per cent of them answered that autonomous mobility will impact their assets but only 10 per cent have a formal planning resource dedicated to the issue—not surprising. This is something everybody is trying to figure out. We had a report with 26 different recommendations across governance, scaling, infrastructure, design, pricing, data, energy and insurance. I'm not going to speak to all of those but I will jump to what I believe are the key implications for the government at this time.

The starting place for us is to realise just what a big deal it is to see this level of technological change. It is probably most analogous to mass adoption of the automobile 100 years ago. I think what's interesting about that analogy is, first of all, that the cities did not necessarily handle the introduction of the automobile successfully. In some cases, the way the automobile was introduced into the urban fabric degraded liveability and created conditions of quasi permanent congestion.

The second thing to take note of is that different cities handled the automobile quite differently. Los Angeles and San Francisco both have the same technology of the automobile that they incorporated that technology in very different ways, resulting in a very different urban form, very different patterns of household expenditures on transportation. Tokyo and Paris London and New York are different cities with the same mobility technology that have integrated it differently into the way they work. While there is deep uncertainty about the exact timing of the rollout of autonomous mobility, it's clear that the implications for the way cities work are potentially quite profound. It's up to government to decide how it wants to use this tech. This is not something where government should simply be getting out of the way or simply enabling it and letting it run its course.

Our view is that the autonomous mobility revolution presents an opportunity to make cities more liveable and more convenient. We can potentially gain greater mobility for many people. We can potentially reclaim space inside cities from storing cars and from moving cars converting that to public space, to parks, to housing, to any number of uses that are more valuable to society than making space for cars. We can potentially make our town centres more walkable. We can potentially save households a lot of money from not having to own a car and from being able to use mobility as a service. However, all of those possibilities are not inevitable; they depend on the framework that cities use for the technology.

I will suggest three key points as high-level principles for what we would view as the way to optimise the potential of this technology. One: in the future, cities will still face geographic and geometrical constraints, more or less the same ones we face today. If cars are about the same size and their occupancy is similar to what it is today, there will be some spatial efficiency from tighter platooning of vehicles, but they're still going to take up a lot of space. If people abandon mass transit and switch to autonomous cars, that is going to be a recipe for much worse gridlock than we have today, because of the geographic constraints.

Two: our throughput, our mobility, is going to remain heavily reliant on mass transit, specifically the train lines and the rapid bus lines, even in the future. There are many trips that do not work for public transit; for those, autonomous vehicles will be of great benefit. But one of the best use cases for autonomous vehicles is to get people to the train station. We want to be thinking about a world where we are doubling down on our investment in mass transit and using autonomous mobility to make that a more seamless, easy-to-use experience. We do not want to be moving to a world where the opposite is happening.

Three: it is time to be defining the regulatory framework for point-to-point mobility for autonomous vehicles now, and I would suggest that we think about that broadly as a continuum, from scooters all the way to autonomous vehicles. The presence of taxis and Uber are part of that continuum; those are already here. The regulatory framework puts government in a role not of direct provider, the way it is with mass transit, but of regulator—the enabler of what will most likely be private services. Generally, innovations follow a classic s-shaped curve: there are early adopters in the beginning; there's a steep curve as the population adopts an innovation, and then it levels off at full saturation or full adoption. We're clearly in the early days of adopting mobility as a service that maybe Uber or the electric bikes represent. But it is probably important to be thinking through the design of the regulatory system at full adoption. I don't think we know yet whether autonomous mobility is going to be a natural monopoly or whether it will be a competitive marketplace with many providers. The design of the regulatory framework will be very different, depending on how that plays out. But it's not too soon to be thinking about the design of the regulatory framework.

I would conclude these opening remarks by observing that Australia is famous for adopting innovations quickly. My understanding is that it had the quickest rate of adoption of the iPhone of any country in the world. There is a chance that, once the technology is ready, it may move through adoption very quickly. Again, with all of the time uncertainty about the rate of development of the technology itself, I'm not the one to guess when that happens. But thank you for inviting us to speak on this today. We would entertain any questions.

Mr ZIMMERMAN: Thank you for joining us and the work that you're obviously already doing for Committee for Sydney, which has such a wonderful reputation in Sydney for its intellectual leadership on many of these issues. The premise that we should be thinking about the regulatory framework is an easy one to say. The harder thing is what that regulatory framework should be. We're, particularly, still in a time of autonomous-vehicle development where we really don't know what the cost structure shape of the industry might be like. There are plenty who will say that, except for the very wealthy, autonomous vehicles will be, for the foreseeable future, outside the capacity of an ordinary person to purchase. Overlade with that, I'm just very conscious of the political constraints that governments will be under. There will be no government that wants to go out at this early stage and say, 'We need to be heading for a world where you're not going to own your own vehicle.' I'm just wondering whether you have any thoughts on, basically, the type of crafting that we should be doing at this stage with all of those political and technological constraints about what's happening in the sector?

Mr Metcalf : My own view is that this is going to be a lot more beneficial if people are treating mobility as a service more or less like when you get out your Uber appropriate, it knows where you are, you say where you want to go and you're not owning that vehicle—it's taking you for a trip, just like a taxi. If autonomous vehicles are privately owned the way they are today, you're not going to see the big savings to households. My own prediction is that it's going to go that way because it will be so much cheaper to pay for five minutes of a ride than to purchase this whole huge vehicle and have it not being used all day. Just because we don't rely on horses as the primary mode of urban transport anymore doesn't mean horses have gone away—people still own them—so there will still be people who choose to own a car for fun—

Mr ZIMMERMAN: But you don't see many on George Street, although it's probably a bit easier now! I come back to the point, though: in terms of where we're shaping the regulations at this point—which was one of your three points, that we need to be doing it now—how do you do that when we don't know not only where the technology is going to land but also where, as I said, politically no government's going to go out and say, 'You're going to have to look at a world where you don't own your own car'?

Mr Metcalf : I don't think you need to go out and say that. I have a couple of thoughts about potential regulatory frameworks that make sense. One is: I would view Uber and potential other competitors as a good chance to start iterating on the regulatory framework. It is a very successful example of mobility as a service that exists today.

Mr ZIMMERMAN: What do you see as the primary goal of the regulatory framework? What do you think the regulatory framework should achieve?

Mr Metcalf : I have a couple of thoughts. One is around a basic city-planning question about where you want to allow vehicles to go. With the parameter of optimising for maximum throughput of traffic, you have an opportunity to have that not be the primary goal anymore. Instead, there is the opportunity to say that we will optimise to create a space for public life in the town centres, on the high streets and in the CBDs of Australia's cities. It's the most basic type of regulation, which is defining the space where vehicles go versus the space where people go.

The second thought is: you might have the opportunity to make a lot of progress on electric-vehicle adoption through regulating fleets rather than regulating personally owned vehicles. So you can pull in another goal of shifting the vehicle fleet away from gasoline toward electricity—presumably eventually renewably powered—through the way you regulate mobility as a service as well as through the way you regulate freight. In the long run, as you move through the adoption curve, I think we're all going to have to start to view mobility as a utility. Right now, in California, the energy utility is going through a bankruptcy, because it burned down the state. The lights still come on even when it goes through a bankruptcy. When a utility is something that people are going to use and depend on, like water or like power, there is a regulatory framework that ensures continuity of service through all kinds of disruptions. If we're going to say to people, 'You don't have to own a car, because you can get a ride to wherever you need one,' and if we're going to say to developers, 'You don't need to build parking, because people aren't going to own cars,' then we're in a world where people are going to depend on the constant availability of this service. That's sort of the end goal, and we're not there. I don't know when we'll be there, but that's where I would imagine we will be headed.

CHAIR: It seems a natural progression that people will just gravitate towards the most efficient option. While we might know where we are going to be in the future and we've got a pretty good grasp of where we are now, we don't necessarily look and see our history clearly. The horse is like a recreational vehicle, and I can see that most likely in the future the privately-owned car will be somewhat similar. It might still have some different uses. That's hard for some people to grapple with and accept. But the essential transport, the routine transport, of commuting, which is what creates the greatest congestion and the problem that we're trying to address, is through the first mile, last mile of utilisation of shared mobility, autonomous mobility. Whether it's a pooled or privately-owned autonomous or a privately-held autonomous vehicle that has many functions or a number of functions for each member of that family, we're reducing the number of cars, autonomous vehicles, being used, and we're maximising the mass transport through the best possible interface that we can achieve. Too often, it has been commented that we've built a great train line, but how do you get to it? What's the access? How is the bus coordinated with where you live? How can you get there? No-one has thought about it. It's like a golfer who's really good at getting the ball towards the green but no good at getting it into the hole, and that's where all the shots occur. That's what's happening with our transport. I think there's a real correlation there.

I think I'm understanding it, but it is the transition. We know where we want to get to, we know where we should be getting to, the question is: how do we get there? It's probably a much more difficult task for those designing autonomous vehicles to compete on that road space if two per cent of vehicles are autonomous and 98 per cent are being driven by us frail humans, who are human to err. That's a more difficult proposition than one with 98 per cent autonomous vehicles, where each of those vehicles could rely on each other to do the right thing, and there is only two per cent human error; that would be an easier scenario. It's more difficult to get started, but it is the transition that we talked to a number of times yesterday and with the idea: would there be dedicated lanes during peak hour traffic for autonomous vehicles? So you've got time and lanes. You would need a critical mass: autonomous buses versus trams and trains. And there is the cost of infrastructure disruption—all these things. Maybe you could talk to a few of those things.

Mr Metcalf : Sometimes the best way to adopt a new technology is not to be first but to be third—

CHAIR: We had that comment yesterday—but coming second was good yesterday. They were a bunch of sentimentalists yesterday!

Mr Metcalf : to let some of the bugs get worked out. It may be that the way to transition is to do things like take a part of a city that has low rates of transit ridership, and buses that are not very well used because they don't run very often, and put in a program where you subsidise people's trips to get to the train station and you let all comers provide that service and use the subsidy—taxis, Uber, new companies that don't exist yet, shuttles. You begin to experiment with reorienting the capillary network, if you will, toward the trunk transit lines. It may be that experimenting with reorienting those patterns is more on the path than experimenting with the technology itself.

CHAIR: That's the right term, is it—capillaries to lead to the main veins or arteries?

Mr Metcalf : Yes.

CHAIR: That is where the devil is, in that detail. In the embracing of transport technologies and meshing that in with, in the case of Sydney, the retro-fitting of infrastructure which is needed on several fronts, and then the land use planning around that of densification—we've looked at both that, which is what we seem to have concentrated on far more than greenfield sites and how would we design things if we had that clean sheet of paper. I don't know whether your committee has looked at those.

Mr Metcalf : Maybe I'll you get you in on this, Sam, but greenfield sites offer the potential for specifically designed infrastructure that enables autonomous shuttle buses or things like that. There is probably a lot of scope for experimenting with those.

Mr Stewart : Obviously there are a number of transport-on-demand trials currently being carried out by the New South Wales government. I suspect you'll see that continue. I think that what's already being recognised is that, as you pointed out, there is quite a big difference between how they work in a greenfield area versus a more established area or something like around Manly, for example—the trials there.

CHAIR: Are you referring to the Uber trials around Manly?

Mr Stewart : No, the shuttle trials with Transdev, which I believe is the operator that's managing that. It's an on-demand trial. The thing that I would caution against when looking at those trials that are already happening is that, with any new technology, there is a fair bit of hesitancy around how it will operate, but, when you are looking at this particular technology, if you're looking at just a shuttle that comes every five minutes or every 10 minutes, it's a completely different set of factors that you're examining versus platooning of autonomous vehicles when it's at mass saturation and there will be quite a lot of them in a confined space. I think when you're talking about infill area, the problem that you're trying to solve is probably quite different to what you'd be trying to solve in suburban areas. In outer suburban areas, it might be that you have more of these shuttles that aren't necessarily as fixed in where they're going, whereas, the more dense an infill the area becomes, the more likely you're going to rely on these trunk lines, to prevent everything being a free-for-all, for lack of a better way to put it.

When it comes to demonstrating, you mentioned the point about having separated lines on major roads for autonomous vehicles. I think, in a way, that's probably what will happen in that transition period. There is a real challenge with the different stages of autonomous vehicles. It's going to be very easy when it's 100 per cent autonomous and the human doesn't need to do anything, but in that interim period, where it's mostly autonomous but humans can press a button to stop it or slow it down or get involved, there are going to be a number of very different technologies on the road at the same time. I think what you're going to have to have is separated-out roads, because, if you've got people who are in private vehicles who aren't feeling as confident driving next to an autonomous vehicle that could be travelling quite fast and then stop, there will be a real hesitancy around people wanting to drive near something that they're not familiar with. I don't think that's a long-term thing; I think it's probably more of a transition thing. Feel free to jump in if you disagree.

CHAIR: To jump in there: one of the extraordinary comments in evidence given yesterday which I thought was very reliable was that autonomous vehicles would be set at, say, three seconds from the car in front. I think the example was coming from somewhere in Victoria, where currently cars were travelling much closer. When the figure was somewhere around 70,000 vehicles a day. By having this extension of a number of seconds, it reduced the capacity of the road enormously. That was seen as something that was necessary in an initial phase. The problem is that, as anyone knows if they drive on Australian roads, as soon as you are a safe distance from the current front, someone will jump in front of you. And it's a pain in the neck. But you can see that autonomous vehicles would invite a non-autonomous vehicle to break that. But, when you get a higher level of autonomous vehicles, they will be able to go closer together and increase the capacity. It was very interesting.

Mr Metcalf : Another cultural commonality between America and Australia is driving habits.

CHAIR: You tailgate pretty well! Sorry I jumped in there, Sam. You can complete what you were saying.

Mr Stewart : I think actually what I might pivot to in that is that, while you probably will have some degree of separation in terms of laneways, I don't think it would necessarily be a good outcome if we just had the same roads that we have today and one lane is for autonomous vehicles while the other is for traditional automobiles. I think that there is a real opportunity to have that autonomous mobility but still have it function as mass transit—as autonomous bus rapid transit, for example—if there's not going to be heavy rail along there. I think that, even if you take into account the potential for platooning—for eventually having smaller distances between cars—even assuming the most optimistic design of a car that's owned by a private individual, it will have built into it for safety reasons quite a large amount of space. To fit in individual mode is on every single cars—obviously you get economies of scale when you do have buses and you can have greater standing room and greater efficiency. I think that there is a real opportunity here and I think Gabriel's point is absolutely dead on: the ideal situation is to get people to that mass transit infrastructure in terms of rail, but, even along those capillary lines, if you can retain the importance, even if they're fully autonomous, bus rapid transit and public transport in that sense, I think that's very useful in certain key areas where maybe you don't have a rail line or a rail line might not stack up economically but where it is an incredibly concentrated spot of congestion already and where the solution probably is going to require some degree of bus rapid transit even if it's fully automated. So I think that separated line is important, but I don't want it to be the only focus—if that makes sense.

CHAIR: We had a comment several years ago when we had a meeting with an authority on autonomous vehicles who made the claim that within 10 years in Sydney, to drive at peak hour, you would have to have an autonomously equipped vehicle. That might have been a little bit premature. But, if we're looking at this, have you given thought to how this does roll out—whether you start to have some major roads declared autonomous-friendly? Obviously highways between cities are easier places to have autonomous trucks, buses and automobiles, but, in the city area, where we really want to address congestion, is there any thought to say: 'Well, we've got main roads. Let's make this this main road, this main road and this main road autonomous friendly.' Maybe you will have some priority to laning up or timing up autonomous vehicles.

Mr Metcalf : Again, I think this is where being second or being third will be really helpful; let some other places work this out. But I would guess there are two use cases where it's going to be deployed first. One is freight trucking where the motorways have standard markings and standard lane widths. You're already seeing experiments where trucks have driven all across Europe. They may have a driver in the front and then a and then a series of trucks that platoon behind to begin with. I think that is already starting basically because of the predictability of the infrastructure—the design standards of those roadways. The other use case where I think you're going to see autonomous mobility roll out first is very lightweight, golf-cart-like vehicles that travel less than 20—shoot. I don't know how to convert from miles to kilometres per hour.

CHAIR: Thirty kilometres an hour is roughly 20 miles.

Mr Metcalf : Okay. They travel slowly enough that, if they hit a pedestrian, the pedestrian does not die. They're lightweight. If you could start from scratch, you'd want it—

Mr ZIMMERMAN: They could still do a bit of damage.

Mr Metcalf : Yes. They are slow enough that you might get hurt but you don't die. If you could start from scratch, you really would say that, in the CBD of a city, you're not going to have these very heavy huge vehicles moving around. You'd want people to kind of get in however they get in and then circulate in vehicles that are actually made to co-exist with pedestrians. And so I think you're going to see in places like college campuses, retirement communities, vacation communities or places with a more—

CHAIR: controlled environment

Mr Metcalf : more authoritarian form of government where they can make things like that happen. You may see that roll out first.

CHAIR: Okay. Let's leave it there. Thank you for your attendance today. If you've been asked to provide any additional information, would you please forward it to the secretary by Wednesday, 6 March. You'll be sent a copy of the transcript of your evidence and will have an opportunity to request corrections to transcription errors. Again, thank you both for attending. Welcome to Australia.

Mr Metcalf : Thank you for having us.