IG: You like to use the concept of “networked urbanism” instead of “smart city.” Can you define what network urbanism is and what the difference between those two terms is?
JT: I think the problem with the term smart city is that it has nothing to do with people. Or at least humans aren’t implied in that term. A city could be just the sum of its infrastructure, and that wouldn’t be a very fun city. I think the early wave of smart city thinking was very infrastructural, like, “let’s put sensors on everything and there is your smart city.” That is only half of the equation. Urbanism, which is the human approach to living in the city, is a better term. Sensors, apps, and technologies become activated by people doing interesting things, useful things, and sometimes bad things. It is the human actors in this mix that make it urbanism. Smart, to me, can mean a million things. Networked doesn’t imply good or bad necessarily, but it does imply a kind of scale. One person with an app is not really contributing to the urban vitality. A thousand people with that app, using a thousand sensors, makes it something new. I opened my talk at the Chicago Architecture Foundation by saying that the best cities have always been smart. It was just that the feedback cycle was slower. What changes now is the scale and the pace in which that information can be exchanged.
IG: What elements define this networked urbanism?
JT: One is using this technology to an end. And the end is to further the goals of urbanism. Urbanism has nothing to do with technology per se. It has to do with making a city more livable, more vibrant. There are a few millennia of examples of good urbanism, applying the core principles of urban planning and urban life. That’s the goal that we are going after. It’s not that we want a better map or a better app. We want better cities. But what’s happening now, and this is where sensors and humans come in, is that feedback loops are spinning much more quickly. There are many examples but one of them would be Waze, a community-based traffic and navigation app where drivers share real-time traffic information. It was acquired by Google in 2013, so, if you go to Google Maps and look at traffic, it is uncannily accurate. In the old days it used to be one sensor, if you will. It was a helicopter saying “the Kennedy looks backed up.” The morning news shows still do that. Now, we have Waze which basically knows where you are and how fast you are going, relaying that data to Google. It creates millions of sensors out there, human sensors. The feedback loops are quicker, sometimes using social media, sometimes using an actual sensor.
In the office we have a Raspberry Pi, which is basically a $40 computer. With it we track what is going on in specific areas of the office in terms of light, temperature, humidity, and barometric pressure. At $40 you can put these things all over the place, for example in a wheel well of a Divvy bike. Our challenge is to figure out how this data can help us to design better cities and make them more livable.
IG: Trashcans have sensors to alert when they are full and need to be emptied. It is interesting how these sensors attached to elements or activities that we already do, such as riding a Divvy bike, now can provide new sets of data.
JT: It tells you what it maybe wasn’t designed to tell you. That’s why I liken the Internet of Things, or networked urbanism, to data portals. Some of those interesting things have been built on city data, which let’s face it, some of that is sensed, that is the vital signs of a city. Some of the coolest things were things that the city could have never dreamed that they would be used in this way. The same thing goes for sensor data. We don’t know what people will use some of this for. We have to be careful, of course, with personally identifiable information. But the data that public infrastructure generates, which is paid for by taxpayers, should be readable. Kevin Lynch talks about the legibility of the city. He refers to it architecturally, from a planning perspective. But I see that term as being just as relevant in a digital sense. Just earlier today, Apple announced that it’s putting together a home object framework, so that your lights and everything can talk to one another.1 That doesn’t really exist at the city scale. I don’t think a company should own it necessarily, but there is real opportunity for making the city more legible. Imagine the navigation and environmental opportunities if the city could be very easily read. If networked public objects like bus shelters and bikes themselves had a networked presence that was query-able.
Types of Data
City Fleet Locations
Public Transit (Trains, Buses, Metra City-Specific Social Media)
Building Permits and Use-Types Business Licenses
Zoning and Parcel Improvements
Services Requests and Fulfillment Violation History
Health Code Violations
IG: Companies, whether public or private, might be collecting data with a goal in mind, but once they make that data public people can read it and use it in unexpected ways. There is that unknown component.
JT: In some ways that is the promise. That is the interesting thing about big data. You are collecting information at such a scale that there is no way of forming the hypothesis ahead of time as to what it will tell you. That is where machine learning comes in. Machine learning is there to be able to define the patterns without you collecting it with an idea in mind, that it will tell us this. Charlie Catlett runs the Urban Center for Computation and Data.2 They have put a grant application into the National Science Foundation to essentially hang several hundred black boxes filled with very cheap sensors on public infrastructure such as light poles. That platform is open to anybody so you can apply to collect data—mostly environmental data—from this sensor network for the next month. It’s like an open data platform for the physical world.
IG: Can making that data collection visible change the way we use public space?
JT: Absolutely. Reading a city is also about knowing what is sensed. For example, the Institute for Applied Autonomy’s iSee is a website where you put in a start and an end, and it routes you around the paths of least surveillance. Reality TV in the past 20 years has shown us that the presence of a camera changes how people behave; there is no question about this. However, data is being collected all the time whether we see it or not. When you are walking down the sidewalk, every person has the ability to record you with their phone, Google Glass, or any other device. I don’t know if that necessarily changes how I walk down the sidewalk, even though I know that I can be recorded. If you are out in public you are probably being recorded somehow. And I don’t know if that has changed peoples’ behavior necessarily.
IG: In your talk you mentioned that the presence of Wi-Fi in Millennium Park has changed how people used that public space.
JT: Yes. The first time I thought of this was experiencing the feedback from speakers. Sometimes microphones will have feedback and they create a “safe area.” The zone of their space is defined by this technology. In this case it’s an error. I think that my kids know, because they don’t have cell phones but they do have iPod touches, where all the free Wi-Fi is in the city. They know where the best Wi-Fi is and it does change their patterns. People who have cellphone networks know the good places and the bad places for Verizon or AT&T. I think we have only scratched the surface in terms of how these technologies change how people use space.
IG: It is clear that we are really good at collecting data. But how good are we at interconnecting that data? And not only data with data, but data with other areas of physical planning.
JT: We are not really good at all. For one, there is a technical issue. Information in some cases is not comparable. For example, the term for “land use” in Hong Kong and “land use” in Kansas City is not even the same. If you and I speak different languages but we are medical doctors, we can probably work on a body, diagnose, and fix it. There is a standard nomenclature and procedures. It is not the same with urban planning and that is a problem.
IG: When you look at large systems, or even different counties in Illinois, in some cases it is almost impossible to compare similar sets of data. We don’t have the means to compare this data, even if we know how we want to look at it.
JT: I think that is why you get so many technologists and data scientists interested in the city. They are interested because there are so many complex relationships. It is one of the most difficult engineering challenges. They don’t care about cities, they aren’t civic minded. They are interested in hard problems. In the office there is an astrophysicist working for us. He doesn’t know about buildings. In astrophysics, especially radio astrophysics, you are trying to find patterns in billions of data points. He is interested in cities for the same reason. But the standard nomenclature I mentioned earlier is just a technical issue. The other issue is that data literacy is very low in the architectural world.
IG: That was actually my next question, data literacy. Some people know how to access and work with data and they are comfortable using it. But others are not and are in a way left out of this. How do we address this issue both at a professional level and a citizen level?
JT: I think we need a better set of tools. What Google Maps has done for location, we need that for data. Imagine the ability of an average person to not even care where the data is, but to basically say “I would like to see economic activity in Humboldt Park versus Lincoln Park.” All that data is there, he or she could get that now, but it is not very easy. One way would be to create a middleware layer. The other aspect is that, especially in an increasingly surveilled world, it would be great if there was a way to access the data that you generate in one place. Health care is still trying to figure this out. There is a federal government initiative called Blue Button that whenever you see the button it signals that it is a health care website. It gives you the ability to download your data, the data that you created. I don’t know if you have ever tried to move records between doctors. It is a real pain in the butt. As you move through the city you are leaving a data trail and you could argue that that is yours. You can download all of your Ventra swipes. You can download all of your Divvy trips. But how about every time you swipe into a building, or something like that? That should be protected, but also it should be yours. I think that this is an opportunity for development: ownership of your data trail. I am sure you have heard that just last week, there were these lawsuits in Europe for the right to be forgotten.3
IG: Yes. I believe thousands of people have already applied.
JT: It is an incredible technical challenge. Another way to put it is that it is not the right to be forgotten, but the right to control your own data trail. Maybe you decide to delete it.
IG: Privacy is clearly an issue. We basically wake up every day and there is something new about it, like the NSA now collecting millions of faces from the web to use in their facial recognition programs.4 I know that the information you put on social media is public but…
JT: I think it would be wrong to suggest that the NSA is only using public data. From what I have read they are tapping phones. But, yes.
IG: Let me rephrase that question. How do we, as users, become more aware of the unwanted consequences of releasing all this information publicly?
JT: When we were little, people talked about media literacy. Training kids to read advertisements for what they were, to understand when you are being pitched. Nobody talks about that anymore because every child I know is fully media literate. For the most part, they know how they are being pitched. That initiative worked. I think we need a similar initiative for data literacy, around the consequences of an online world. The other problem is that for a vast majority of business online, like Facebook and others, it is in their best interest to obfuscate the privacy implications of releasing this data. The more people who are exposing more data on Facebook, the better for Facebook and its marketers.
IG: When a government sector, like NSA, is using this information secretly, and the private sector does not want to be clear about the use of the data they release, who should be in charge of pushing this initiative of data literacy?
Transparency Builds Trust
Accountability Builds Better Workforces
Analysis Builds New Processes
Open Data Builds Businesses
IG: You have worked both in the private and public sector. How is data approached from each sector?
JT: The public sector data is seen as being valuable mostly for transparency and accountability, which are things that companies care about, but they have moved past that. Data for companies is really about analytics, becoming more profitable, saving money, analyzing costs. Now, certain governments like ours have made that leap. But how many hack-a-thons do you hear about where they are actually doing analytics? Not many. They are creating little web apps. We still have a long way to go in governments really embracing data, really embracing machine learning and analytics to change the way things get done. Partially this is just inertia. So much of politics is instinct, gut, and anecdote. I am not advocating for a kind of robotic or computer controlled government. At IBM, I was one of the people who tested Watson [a cognitive technology that processes information more like a human than a computer], so I got to play it. The interesting thing about Watson was that is wasn’t connected to the Internet. It didn’t go out and do Google searches. It actually was given data and then it had to find the patterns of correlation between terms and things like that. That’s the promise of data and analytics that the government has just coming to. In general government is behind the private sector in using data.
IG: In terms of Chicago and data, the Chicago Architecture Foundation (CAF) just opened its exhibition Chicago: City of Big Data, for which you are part of their advisory board. What are the goals of the exhibition and what are the tools used to engage with people?
JT: The CAF organized it by scale: human, block, neighborhood, and city. I think of it a little differently. How does data, surveillance, and sensors inform city design? How is it actually changing the way that we design cities or how should it? It is similar to the work we do here at PositivEnergy Practice. The flip side is, how can we learn more about our cities by looking at it through this prism? The CAF is giving a physical walking tour where you are looking at the infrastructure of data: cameras, data centers, and cell phone repeaters. Once you see those elements, you start to see the city in a different way. I think that is good. It also goes to the data literacy point. The Internet does not just rain down from heaven. There is an actual built structure.
IG: That is what Andrew Blum talks about in his book Tubes. Data is physical.
JT: It has a physicality, it has a cost, and it has an aesthetic. If the old wooden water towers on buildings are still there, they are festooned with radio antennas. And sometimes the cistern itself has rotted away so it’s just the antennas. What are the implications of that? Should we zone things differently? Should the city take an active role in making it beautiful?
IG: In a lot of cities, Chicago included, old structures that were designed to hold heavy load such as post offices, banks, or printing plants are now chosen to be used to as data centers. The buildings that became obsolete now can be easily re-used for these new technologies.
JT: I think old cities have an advantage because the cost is already embedded. But you have to be creative about thinking about how to reuse that infrastructure. One of the great things about the “L” [Chicago’s elevated train system] is that if you want to run fiber, you don’t have to dig in the street. You don’t have to worry about network repeaters, because you are out in the open as opposed to the NYC subways. I think that’s why you don’t hear a lot of good stories about smart urbanism from cities that have been built from nothing. There is something about the layering. It has to do with the human engagement with the city. You don’t engage with an amusement park or mall because they are fake. They are deliberately fake. The beauty of cities is that there is some friction, they can be dirty, they can be hard to quantify. That’s why people like living in them though. And then you overlay this clean, digital interface over the city. The designers of these technologies applied in cities have to be comfortable with them being used in ways that they didn’t intend. I guess it is just not the way that computer developers think.
IG: Are there cities that are good examples of data use? What are the factors that make them successful?
JT: Glasgow is doing interesting things and so is Barcelona. In some ways they are not the cities you think. You would think that San Francisco, for instance, would be leading things. They really are not. Most of the tech talent is in the valley, which is very automobile centric. You can argue that there is a lot of smart technology out there, but there is not a lot of urbanism. Everybody drives.
There is a reason that the garage is mythic in the valley. In terms of factors, leadership does matter, but it’s not even primarily from the government. For example this last weekend [May 31− June1] was the National Day of Civic Hacking and hundreds of cities participated. It turns out that Detroit has tons of local data, it just wasn’t collected by the government. There were people outside of the government that needed this data for whatever their thing is, and they built it up themselves on the edges of city government. That is probably the most important thing: leaders in the community.
IG: Definitely a key group to identify. Thank you very much for sharing your insight about data and cities. A complex and fascinating topic.
JT: My pleasure.