DESIGNING A DIGITAL CITY

Craig Nevill- Manning, who j oined Google i n 2000 as a senior research scientist and started i ts fi rst remote engineering office i n New York i n 2003, i s one of New Zealand’s most successful tech exports. As engineering director, he has overseen projects l i ke Google Maps and Google Shopping, furthered Google’s philanthropic efforts i n disaster zones l i ke (fi ve- and- a- bit years after i t appeared as an option on the search engine). But i n 2016, he moved to fellow Alphabet company Sidewalk Labs, which i s reimagining cities i n order to i mprove quality of l i fe. Its vision i s that “by combining people-centred urban design with cutting- edge technology, we can achieve new standards of sustainability, affordability, mobility, and economic opportunity” and, as head of engineering, he and his team are currently attempting to bring a slice of waterfront l and i n Toronto kicking and screaming i nto the digital era – and then take those i nsights and i deas to the rest of the world. Ben Fahy talks with him about a more fl exible future, how autonomous cars could change everything, the role of open data, what New Zealand cities are doing well and the conflict between digital efficiency and messy humanity.

Ben Fahy :

Google's already solved a pretty complex problem: organising the world's information. Is this the next step, and is fixing cities harder than what you've done before?

Craig Nevill-Manning:

It’s certainly very different. I'm going to be pedantic for a second, but while we are related to Google, we're definitely a different company. I had to quit Google, I lost my badge, and we are in a totally different office, in a different part of New York. I go to Toronto once a week, sometimes twice a week, and we're hiring a lot of people in Toronto as well. In my old life at Google, organising the world's information was a fun problem. I don’t think that’s really been solved yet, but I feel like Google's made a pretty good start.

So what attracted you to Sidewalk Labs?

It feels like this gap between what's currently done and what's possible when it comes to technology in cities. And it's exciting to think about how technology could actually improve quality of life in cities, in a whole bunch of ways. And there are a number of reasons why all of this technology hasn't been applied before, and I think Sidewalk Labs are in a good position to help make that leap.

What are some of those reasons? One of the things that people talk about is that planners don't really understand technology and maybe technologists don't understand cities. Is that tension you're playing with?

We use this phrase so much within Sidewalk Labs it's a bit cliché, but you’re right, there's a gap between technologists and urbanists. And so we think a lot about how to bridge that divide. I love cities. I'm a big fan of cities. I love New York, I love Auckland, I love Toronto. They're great places to be, but before joining Sidewalk Labs, I didn't really have a sense of what made them tick. How they worked, how you'll get things done, how you might even go about improving them. So from the technologists' side, there's that and from the urbanists' side, in general cities have used technology in really interesting ways in the past. Think about aqueducts in Rome or the power grid starting in places like New York City.

I love the story of the London sewer system. In a rare example of long-term thinking, they said ‘Let's triple the size of it’ after they got the report back. They admitted that they didn’t know what the city was going to be like in a hundred years and that sewerage system was still working until relatively recently.

It's all technology. It's all things that improve quality of life, improve health, improve wellbeing in cities and so cities have done that a lot in the past. They are adopting digital technology over time. One of the challenges, one of the divides I guess, is that if you're a software company and you say ‘well, let's try this new idea’, you can launch something as a beta and get it out there and tell people this may

or may not work. Try it out and give us some feedback, and then you iterate towards perfection hopefully. Or you just grab the idea and you say ‘well, that was worth trying but it turned out not to be a good idea in practise’. Cities don't have the luxury of doing that. If they don't pick up the rubbish on a regular basis, really bad things start happening. So there's a different appetite for risk for very, very good reasons in cities, and furthermore, as you know, to make a change in a physical city involves a lot of new infrastructure. It's very different from software, and so I'm stating the obvious, but …

Cities are also similar to software in the way that they're quite iterative. Things get bolted on over time, they grow and evolve and so the interesting thing with Toronto, where Sidewalk Labs is conducting its first project, is that there's almost a clean slate. You’ve got this chunk of a city that you’re going to try all of this new stuff in. What will we be looking at in five years? What's the vision of Sidewalk Labs in that area?

You're right. In many ways, this area in Toronto is a clean slate. It used to be the Port of Toronto for many, many decades. There was shipping in and out of that area. It got built up over time, and for a whole bunch of reasons, when it stopped being a port, it hasn't been developed as a neighbourhood in the city. Even though geographically it's right there, right downtown. Honestly, one of the reasons we are super excited about Toronto because we wanted to be able to develop infrastructure from scratch. Thinking about that in a new way, but we didn't want to do it out in the middle of nowhere, which would have been easier in terms of getting permission.

Like one of those Chinese cities where they just build it and hope that millions of people will eventually live there.

Right. Whereas we wanted to do it in a city in North America where there was already a great city life and urban environment so that you can essentially have that spill over into this new place rather than trying to create that from scratch. This also creates constraints because you have to do this with the enthusiasm of the city, in this case, Toronto. In terms of the clean slate, if you believe that in some small number of years, self-driving cars will be practical, and I happen to be fairly optimistic on this front, I think things are moving incredibly quickly and cities could actually look very different. So it's not a question of just how that will change how people get around from A to B, it's also about how you change the way you use land in the city. If you think about a neighbourhood where it's 100 percent self-driving vehicles and in that environment you've got vehicles dropping people off and picking them up. If they need to recharge or they're not in use, they can go into a basement or to the edge of town and recharge themselves. There's no reason to use precious urban land for storing cars, for cars sitting around waiting for somebody to jump and drive them.

A huge chunk of urban space is dedicated to cars. Do you know the percentage on that?

It's between 30 and 40 percent in main

If you think about a neighbourhood where i t's 100 percent self- driving vehicles and i n that environment you've got vehicles dropping people off and picking them up. If they need to recharge or they're not i n use, they can go i nto a basement or to the edge of town and recharge themselves.

cities, and if you think of how valuable city land is, storing cars seems like a crazy use of that land.

Parking is quite a lucrative industry though, isn't it?

That's right, but you can use land for other things once you don't use it for storing cars anymore. Also, they tend to be in places where all the people want to be. So, often cars are parked along the sides of the streets but if you get back that land, you presumably can make the footpath wider, you can maybe add another lane. But I think the more interesting thing is if you don't have streets and buildings already that would define what will work, what's the right answer? Is it actually making the street narrower?

Flexibility seems to be a key theme of some of the ideas that I've seen from Sidewalk Labs, like the ability for a street to be the same size, but then it changes at different times of the day or even buildings that can adapt. Is that something that has been key from the start?

Absolutely. Cities do adapt over time and you certainly see old warehouses getting reused as artists' spaces and then as apartments, but buildings and streets aren't necessarily designed to be flexible. And we think with digital technology, one direction you could move in is to appoint much more flexible streets and buildings. So that as uses change, the actual environment changes with you, and so maybe in a building that's over a number of months, but as you pointed out in a street case, it might be over the period of an hour. So maybe you have a street that is one way in the morning when it's rush hour, and then as a pedestrian plaza in the lunch time. It becomes one way in the other direction for the evening rush hour, and then maybe in between actually you shut down a lane because school's getting out and it's time to drop off and pick up the kids. And we haven't quite figured out exactly how you signal all of those things. We've actually done an experiment

where we just renovated a warehouse on the Toronto waterfront and in there we've got an experimental street of the future where you essentially have lights that indicate to pedestrians and cyclists that this part's sidewalk, this path's a lane going in a certain direction, and then that switches and suddenly the whole thing is a pedestrian plaza. We don't think we got it quite right in that experiment but we can start iterating on that idea when we're actually building the thing.

Do you have any other predictions around how we'll live? Will we be moving panels around our houses to fit our own day?

Buildings do change. If you think about the downtown buildings, they do change in their use over time, from light industrial warehouses to sometimes residential, to commercial. And even during their lives in a particular use, you want to move around the floor plans. We think we can make that much more flexible. So maybe as your family grows, you need another bedroom, you don't necessarily have to move out of the building. You can actually rearrange things within the building to enable you to stay in the same neighbourhood. That derives from a number of potential technologies. One has been around for a while – modular housing. Building out of modular components, and then be sort of mixed and matched. There's a really interesting technology around low voltage power. If you look around at most of the things that you plug into the wall, a lot of them don't require high voltage. They often have a transformer brick attached to them and then low voltage coming out the end. Can we actually have more low voltage supplied within buildings? That would mean that you don't necessarily have to have an electrician to move a wall around because you don't have high voltage cables embedded.

Will we even have power lines?

I think so. People are working hard on wireless power, but that's one technology that doesn't seem quite ready for prime time, unless it's the wireless power that says, "I put my phone down on a particular spot on the table and it charges." I think that at longer range, we haven't seen successful applications. But I wouldn't rule it out, to be honest, in the decades ahead.

What's the time frame for this? I know tech companies like to move quickly and Google certainly does. But cities don't really move quickly, they evolve over time. At the moment, the idea of smart cities is largely traffic lights and maybe a few sensors in some of the more developed cities. Lots of cameras obviously around, but at the moment there's a lot of expectancy and hype around it. Is the substance going to come soon do you think? Are we in the trough of disillusionment at the moment?

Well, I think in terms of smart cities more generally, I feel like there's a little bit of a false start. In the sense that a number of companies said cities should be smarter. Here are a bunch of sensors. Let's sprinkle them around the city and we're sure they're going to be useful for something. And it wasn't really driven by improving quality of life. In some cases they were but in general I think it's good to take a step back and say, “what are we trying to achieve?” And in this case of Sidewalk Labs, having a bunch of technologists like myself be in a room for all the working day with a bunch of urbanists, we think about quality of life and what that means for sustainability or for flexibility or for mobility and accessibility. And then we work back to the technologies that we think will be helpful, and so I think we're going through that now.

So it's becoming a bit more useful?

I think so, and the technology might not necessarily be a million miles different, but it'll be used in much more considerate way. I think one good example you mentioned is cameras. In various past instances, cameras were scattered around and people didn't always consider the privacy aspects of that. It's good to start from the point of view that people's privacy is sacrosanct and they shouldn't be being spied on by the city, and starting from there you can think about if you want to use cameras for understanding the flow of vehicles in a city for example, through an intersection. Can you do that without actually transmitting any of the actual footage from the camera but rather can you – and this has only just become technically plausible – have a device that is a camera and processor all in one? And all the processing is on the device, and then the only thing that's transmitted back is, say, counts of cars every five minutes. Preserving everybody's privacy but still giving you the upside of being able to understand whether there's congestion and whether there are near accidents.

That's something that has been talked about a lot with Google's motivations. Sidewalk Labs is a different company, but Google’s business model is based around collecting data and using it to make money. So who will own the data in the city environment? Will you own it? Will the city own it? What's your take on the data issue?

We are different from Google. And it's true that Google uses machine learning data to target ads. But that's not part of our business model. We're developing a city and it's not at all clear that any other business models from Google apply to Sidewalk Labs. Our business model is more around building real estate, we'll sell some of that real estate, we're operating a city. In terms of the data itself, there's a really strong open data movement within cities that has gained momentum over the past decade. What everybody realises is that data about what's going on in the city has some value. But it has value to a whole bunch of different players and so it is good to make this data open and available and then see what people do with it. Rather than try to make money of the data stream itself. So there are a lot of discussions that we're currently having with the city of Toronto, the residents. Nothing's been decided, but that's a pretty active area of discussion.

In terms of what you see coming, what are the things that excite you most?

I think the self-driving car piece, not just because of mobility, but because of the way it changes the way you think about how you actually design cities and how flexible they can be and how you use the land. I think another big area is sustainability. We want to make this place, for example, planet positive. We want to make sure we're taking carbon out of the atmosphere, essentially. And there are a bunch of approaches that we need to use in parallel. One of them is just being more efficient about heating and cooling and lighting. Building systems that can use machine learning to improve the efficiency of heating and cooling and those other energy uses. I think there's a lot of headroom there. So that I'm pretty excited about. And this is this is going to sound kind of brazenly futuristic, but…

That's what you're there for.

Exactly. But with robotic delivery of goods around the city, either with real robots or maybe it's drones, the technology's going to advance quite a bit, I suspect, in the next decade and beyond. But, again it's not so much that you're going get your packages from some online merchant faster, although that will happen. It's also about having to own less stuff. So, imagine I have a cordless drill, I use it maybe 10 minutes a month, but when I use it I really need it. I kind of need it for those 10 minutes. But maybe I don't need to own it, especially when I can get it in a few minutes' notice.

Especially if you don't have to deal with anyone to pick it up or you don't have to have that awkward transaction.

Yeah, so maybe I share it with a bunch of people or maybe there's a sort of a central reservoir of cordless drills. But imagine, if I could say to some sort of digital assistant, ‘I really need a cordless drill’ and then five minutes later I open my closet in my apartment and there's a cordless drill. Behind the scenes there's a robot that delivered it and scurried up the elevator and put it in. Maybe there's a closet in my apartment that has a door on the other side that robot puts stuff in. And so it's really, really seamless. Now, I don't need to own one so I'm saving some money. I don't need to find room in my apartment to store all the stuff that I needed occasionally. I don't need to remember where I put it. All that stuff sort of goes away and so there's a whole bunch of efficiencies there. It's not just that we're being lazy, although it will make life a little bit easier. It’s about using resources more effectively. We're not buying a cordless drill in most of the apartments around the city and finding some way to store them, we're just making smarter use of the resources.

So we may be going back to having a series of pneumatic tubes around the city, like we used to have in the office?

If you've been to C1 in Christchurch, they deliver hamburgers by pneumatic tubes.

Have you still got your cafe?

Yeah, we have a little New Zealand cafe in New York City, called Happy Bones. It’s quite small. It's only about 450 square

Cities l ast a l ong time and so one of the big, almost meta- challenges of cities and technology i s how do you design things so that i n 30 years' time somebody doesn't l ook back and say, ' Those guys are crazy.'

feet, so 45 square metres or so.

On the topic of Christchurch, what do you think of the process they've gone through? I wouldn't call it a clean slate, but a lot of people talked about the opportunity to create a new smart city. From what I've heard, that hasn't really come to pass. What's your impression of what Christchurch has become?

I haven't been to Christchurch in a few years, so I haven't got it updated on what's been going on there. I really like the look of the plans, they stepped back and considered the overall structure of the downtown and how to create new parks, which is great. I do think that technology has advanced a lot since the earthquake in 2011. Honestly, if you started fresh with Christchurch today, you might make even slightly different decisions. For example, self-driving cars were much less developed than they are today. But that also brings up this really interesting challenge, in general, which is that technology's going to continue to evolve, right? We know all about self-driving cars and drones and to some extent, robots, but what's going to be possible in 10 years, 20 years, 50 years? Cities last a long time and so one of the big, almost meta-challenges of cities and technology is how do you design things so that in 30 years' time somebody doesn't look back and say, “Those guys are crazy. They completely built everything around assuming that drones were going to be the big thing and it turns out there was a much better idea just around the corner and now everything's kind of designed for that old, out of date technology.” I think there's some amount of humility that you have to have when you're thinking about planning things.

Humans are not particularly good at predictions and we don’t seem to think long term enough these days. In Auckland at the moment there are quite a few issues around raw sewage going into the harbour after it rains, gridlock, housing unaffordability, but it's still a

very popular place. The infrastructure just isn’t keeping up with the growth. Are these problems just part of life in the city? Can you actually solve those things? There are probably things you can put in place to future-proof yourself a little bit. And some of these things are easy and doable and some of them are very, very hard. One of the things we've learned on the software side is that you can upgrade it over time. The fact that your phone every so often says, "Hey, it's time for an upgrade. I've got the latest version of the software." And suddenly your phone is a little bit more capable than it was before. That's maybe a little bit of a stretch to think of a city as a phone because that's really not the same thing. But there's probably some truth to the idea that if you can have things more software-run, software modifiable that you can probably adapt a little bit better. For example, the solution from the people who've tried to solve congestion in the past is to build more roads. Oh, the motorway's congested, let's add another lane. It turns out when you add more lanes to a freeway, you end up with what's called induced demand. More people drive and a few years later it's just as congested as before if not worse. It might be the case that the better solution is to understand the flow of traffic in a much more fine grained way using sensors. And then have a way of feeding back to drivers and to vehicles, so that people can make better decisions about whether they're driving or not. Enable, for example, public transit to take an express lane. You already see this to some extent with Google Maps and Waze taking data from people using those devices and feeding them back and saying, "There's a crash up ahead, you should take a side street." There’s more opportunity for those software solutions to enable the cities to constantly upgrade themselves not from an infrastructure point of view, but from a technology point of view. And that's your theory on the city as a digital platform, in a way. Having the data flowing through a city and using that effectively? Exactly. If I'm a software developer and I want to build an app on a phone, if I'm building it for iOS or Android, there's a whole bunch of pre-packaged APIs, these interfaces that programmers use to make it easy to do certain things, to use certain features of the phone. The questions we're asking are what are the APIs for a city? What should be the standard things that, if I want to build a better transportation system, or a more efficient way of heating buildings, what are the APIs that I should be able to take for granted that provide the information about what's going on in the city at an aggregate level? One of the phrases that I saw used was around turning Toronto into one of the most measurable cities in the world. For Google and lots of other digital companies, analytics are really important to guide decision making. Is that something cities need to do better? I think most cities already measure themselves. They keep track of things like congestion and usage, but there's an opportunity to make that even more effective. As soon as you say measurement in a city you always raise the spectre of surveillance and lack of privacy so I think you need to be very careful that you're not measuring so much detail that you're actually gathering personal information. You need to keep it in the aggregate, in a way that's unpersonalised. But when you can measure something better you can decide which trade-offs to make in a wellinformed way. I guess there's another tension here between technologists and urbanists, and it’s that urbanists are often looking at the human stuff; the interactions, not the algorithms. Dan Doctoroff [the CEO of Sidewalk Labs] was asked about the High Line in New York. It took a long time to get through, and a lot of people didn't want it. It required people who had a vision and it wasn't based on an algorithm or data, it was based on a weird human desire to bring an idea to life. Have you found that difficult, and have you moved in the direction of urbanists – to look more at human responses and that kind of organic approach to design? There are certain things that measurement and algorithms are useful for and, to your point, a whole bunch of other things that really require human empathy and experience and understanding and relationships. I honestly think if you walked into Quayside in Toronto and said, "Oh my goodness, look at all this technology," I would consider that a failure. You want to walk into a place and say, "Oh wow, what a great neighbourhood. Look at all this stuff going on. Oh, we should check that place out, I've heard it's great." And it's about the people, it's about the energy, it's about the creativity on the street. In the background somewhere, it is technology making that more efficient and safer, but that's not the goal. It's kind of like the sewers under the street that make sure that the right stuff happens with waste in the city. Nobody pays it much attention most of the time. It just has to work. But it enables all of these other things in the city and keeps people healthy. That's the way to think about the way technology should be used in the city. Another criticism of the smart city is that it's aimed towards wealthy cities, or wealthy suburbs? How do you stop more inequality being created with these technologies? We really set right up front the goal that this would increase affordability and would be an inclusive place. For example, in Toronto, we want this neighbourhood to reflect the diversity of the greater Toronto area, which is incredibly diverse. It turns out that Toronto is actually probably the most diverse large city in North America. Immigrants make up 50 percent of the population, which is amazing. And long may that continue. And so one of the things we have to figure out is how do you make this

place feel like it was built for each of the folks that live in Toronto, that they can see themselves living in this place. Part of that's affordability, but there a whole lot of other cultural aspects of that. Where you want to build a place with diversity and attract a diverse set of residents, a diverse set of visitors. And we hope that some of these technological approaches will actually help on the affordability front. If we can drive down the cost of getting around, if we can drive down the cost of heating and cooling homes, etc. There's a good opportunity to pass that on to people who might not otherwise be able to afford to live in downtown. I think adding parking into a development adds around 30 percent to the cost. Right. So you take that out and the cost comes down. And we're having that debate in Auckland at the moment. There are some progressive developers who have said, "We're going to have a couple of shared cars and there's no car parking." And then they get complaints from the locals because they think the residents will take up their parks on the street. But there’s demand for that type of development and they tend to inspire more public and active transport. There are some things that need to be broken for the city to get better and more efficient. I think your point here and earlier on when you were talking about the High Line is that this back and forth between people within a city is critical. It's very rarely the case that somebody's first idea is the perfect idea. And these things get shaped by that public discussion. And sometimes the idea turns out to be a terrible idea after that discussion and it doesn't go ahead, but even in the case where it does go ahead, it's often significantly modified. And so you have to have that public engagement and public discussion because I think that's what ends up creating great places. Sometimes efficiency is the enemy of great places isn't it? You want that randomness, and that serendipity. Whereas you look at the tech mindset, they are often trying to limit human interaction. But the more you interact with friends or people in general, the more likely you are to live longer. That kind of thing is being thought about more now, with more shared spaces within developments and better placemaking in public spaces. Instead of putting fences up, actually trying to talk to your neighbours and create communities. I don't know how often you get back to New Zealand, but do you see anything that we're doing over here that is standing out, that you would say is world-leading, or is just interesting? In general, New Zealand is incredibly friendly. I get back there once a year, but whenever I have visitors from the US who go to New Zealand and come back, they say, "I can't believe how friendly everybody is." So there's a baseline of people being pleasant to each other, interacting, like you say, in public spaces. Combine that with New Zealanders' comfort with technological change, so many technologies get tried out on a small scale in New Zealand before they get tried

In the background somewhere, i t i s technology making that more efficient and safer, but that's not the goal. It's kind of l i ke the sewers under the street that make sure that the right stuff happens with waste i n the city.

out in the rest of the world. I think about EFTPOS in the ‘80s. So I think that New Zealand cities are going be up with the fastest cities in terms of adopting new technology. In terms of those spaces, I think that Christchurch is a good example of where there's been a lot of thought about the new downtown in a modern age. And presumably it won't be perfect, but cities never are. They're always in the process of becoming something else. But I think New Zealanders are generally thoughtful about that. Do you have a favourite city? I chose to live in New York, so I'm a big fan of New York. But honestly, cities are so diverse around the world. Toronto's amazing in the sense of its international aspects and diversity, sort of the Canadian-ness of Toronto. I love visiting cities. In fact, when we go away on vacation, it's almost always to some city elsewhere in the world. I think cities are endlessly fascinating. There's a huge urbanisation happening all around the world. People are moving into cities faster than ever before. Once Toronto is up and running, how big is the opportunity for Sidewalk Labs? And will you attempt to take those insights into other cities? The whole point of this project is firstly to make a fantastic place within Toronto, but then, as you intimated, take those ideas and seed other cities around the world with them. Once the ideas have been de-risked a little bit and it’s clear that they can work and have a positive impact, cities do tend to adopt things relatively quickly. So we're hopeful that happens. And you're right, urbanisation is happening at an incredible rate around the world, so it's actually super important that we do a really good job of designing cities and take advantage of all the technological opportunities there. That's our hope, that the longer term impact of this won't be just in Toronto, but that it will be in cities around the world.