Bladerunner or Babylon? APIs Are Shaping the Cities of our Future

Cities are an important platform on which tech solutions are being created. Uber, Airbnb and Yelp for example, all are tech giants that grew because they were able to leverage the city as the common denominator for their service. They each use the city as the real life platform on which they are transacting value between two partners, whether that be drivers and passengers, home owners/renters and travelers, or restaurants/shops and local customers.

There are also a growing number of civic tech startups that are solving city problems or working to provide services and tools that enable city authorities to function better, make better use of resources, and encourage citizen engagement. If you can create a solution for one city, it may well be replicable in other cities and be able to scale to a global, or regional, audience.

And while much more slow moving, city authorities are also recognizing that their world is changing and that citizens want to engage with them via social media, mobile phone and online, and that they can carry out their work more effectively using mobile apps and cloud-based services.

We live in an area of digital transformation that is pervading every industry including governments, because APIs.

Now APIs are at the center of the idea of smart cities: they are enabling an automated, responsive data architecture that reimagines city management and government services as a platform.

But if the automated, technocratic potential of APIs are taken to their logical conclusion, we risk ending up in the acid rain-soaked, data-controlled dystopian megalopolis of Blade Runner.

Don’t believe it? Well, it has already happened.

Swiss architect, urban planner and one-time Paris resident, Le Corbusier spelled had a stunning architectural vision. As an urban planner he became fixated on the idea of The Radiant City, a mindset that sounds very similar to the thinking behind smart cities. Academic Teresa Almeida said of Le Corbusier’s vision that:

“The main principles of Le Corbusier’s theories include scientific rationalism, efficiency, and social improvement through design.”

Le Corbusier even developed a smart city-like template for Paris which, luckily, never eventuated.

Le Corbusier’s model for Paris (Photo by Amber Case, licensed under CC BY-NC 2.0)

But his ideas were taken up wholesale in the ’50s in America, and we see them today in the blight of the high rise, urban housing ‘Projects’ that created racially profiled, soulless poverty traps particularly in the U.S. but were also used as the main design theory for public housing estates in Australia, the UK, and around the world and have since been responsible for ghettoizing poverty and reinforcing long-term low-socioeconomic area disadvantage.

But the mindset that Le Corbusier started with is very similar to much of the thinking we hear around how to leverage APIs for smart cities.

We still see the goals of scientific rationalism, efficiency and social improvement through design being suggested by city planners as the way to create a smart city.

Do a Google image search on smart city models and you see a gorgeous page of colorful infographics and matrices that describe how urban planners, mesmerized by the potential of automated APIs, think their place-making systems will operate. But check the comments page of recent articles like TechRepublic’s piece on ‘What IoT and smart governments will mean for you’ and you see what the community thinks with feedback like “1984”, “Real life isn’t about technology” and “More Big Brother. No Thanks.”

Last year, at Barcelona’s annual Smart City Expo, several leading thinkers each presented models of how a smart city should operate.

Some of the smart city models presented at Smart City Expo and World Conference in 2014

For example, city-nation-state Singapore shared their smart nation vision, others talked of a matrix that merges functional areas of city government policy with technology drivers, while smart cities advocate Boyd Cohen proposed a series of 62 indicators that measure a smart cities success (I’ll come back to this last one in a bit).

The problem here is in taking a policy-driven approach first to city planning. What is really needed is a user-centric, service-oriented approach, similar to what last year’s APIdays Global speaker Ivo Jans from Buildozer presented when he advocated for scenario-driven API design.

The shortcomings of a policy-driven top-down approach can be seen in how open data initiatives have been implemented around the world. There is some inspiring work being done by civic tech startups. OpenDataSoft, for example, recently published a global list of 1,600 open data portals — many of them city-based — and made the research collection available on GitHub so that anyone can contribute to the open source-like project and help maintain information about the portals.

But look at just three example cities — Paris (France), Brisbane (Australia) and Seattle (USA) — and we see that when instituted by a top-down approach, the data that is provided does not reflect citizen needs.

In none of these three cities is it possible to produce a map of all supermarket locations and then overlay that with demographics data on socioeconomic disadvantage.
In other words, you cannot easily use open data to create a map showing which neighborhoods in each city are at risk of food security because residents do not live within walking distance of a food and groceries outlet.

Yelp has better data on supermarkets than these cities. In fact, you could probably use the Yelp API to at least map the supermarket locations faster than you could by using city open data portals.

The three steps that need to be taken to create citizen-focused API automation

1. Cadastral and Demographic Data via API

The starting point for opening data that can be API-enabled for use in a city-as-a-platform agenda should be to first digitize our real world assets. In database parlance, this is what we would call data modeling. Start with the decision-making structures (a complete list of all city departments and institutions is often not available), then all the physical assets being managed by a city: the bus stop locations, the libraries and childcare centers, the community resources, the parks, the plazas, the unused tracts of land that cities own or manage. And then add comprehensive data on business and land use (a cadastre). Alongside this, census data (and, ideally, population movements) help give a place engagement understanding of our cities.

Together, these are crucial for enabling sensor systems and IoT via API at a later date.

Once these core datasets are opened up via API, they instantly become a resource that can enable businesses to begin identifying new commercial opportunities, and allow community groups to activate citizen participation.

This is what Jerry Paffendorf’s team at LOVELAND Technologies is working towards. They are on a mission to put all of America’s land parcels onto maps so that the information is useful for citizens, communities, startups and city governments not just “a few professionals and decision makers”.

LOVELAND is creating a cadastre by crowdsourcing data and uses APIs to make data available for planning and resident consultation projects.

It is a similar model to how the U.S.-based Trailhead Labs has worked with city authorities, government agencies and community groups to map parks walking trails. Now they are able to encourage others to create interactive tools and are looking at partnerships to leverage this data in ways that re-engage citizens, sport and outdoor industries, environment groups and travelers with America’s beautiful parklands.

Trailhead Labs worked with community agencies and government agencies to build a mapping platform for U.S. parklands and uses APIs to turn this data into a mapping platform

2. Take a User Centric Approach

Partly what is making LOVELAND Technologies and Trailhead Labs successful is a bottom-up approach. In startup parlance, this is what we would call a user-centric approach. In successful civic tech projects, citizens are starting the data collection and participating in improving the robustness of the data. They are not being given outdated data and marketing-based contests to see what they can make do with the data in a hackathon-style competition that often comes as the ubiquitous promotional strategy after a city has released an open data portal.

Instead, cities need to look at ways to encourage and facilitate citizens to contribute data, and to make use of data in their existing community work. This is how The Guardian managed to activate the UK population around government spending data where local residents would help plug in data on spending about their local representatives in order to build up a nationwide databank that now anyone can explore to see levels of expenditure for their local politicians, or the model that Open Corporates has used to encourage a comprehensive business registry accessible across all jurisdictions.

This is the exact opposite of Boyd Cohen’s 62 smart city indicators. Last year, he sent a survey to 120 cities around the world. Only 11 responded. The whole model was based on a top-down approach: the indicators were selected by representatives from IBM, a consortium of infrastructure vendors, the Smart City Expo organizer, and the director of a business school. There was no community representative, and the indicators were not tested with citizen groups to test fitness for purpose in community work.

It seems like a missed opportunity not to encourage that, rather than the city authorities to respond, citizens help contribute to the data and the collection and reporting. Perhaps it also would have been a way to engage with citizens around how they could use that indicator data to see how their city is performing or to start looking at what data lends itself to enabling citizen-focused smart city infrastructure.

This is how SeeClickFix got traction: they went straight to the community, and gave them the tools to alert cities of potholes, rubbish dumps, ailing infrastructure. It is how VIMOC Technologies’ Landscape-computing began: they worked with a business association to understand current needs of downtown businesses and then placed sensors in parking bays as a way to collect data on retail district activation. The business association then lobbied for that infrastructure to be rolled out and enhanced across the city.

3. Apply APIs and Data to Use Case Challenges

The third layer after open data building blocks and engaging with citizens is to focus on key urban challenges. In microservices parlance, this is what we would call domain driven design.

Everyone loves a dashboard, but a city dashboard is meaningless. Beyond helping the technocrats measure their top-down policy models, they are fairly important at helping citizens get access to the health and social resources that make living in a city vibrant, dynamic, and supportive.

Personally, I propose that any city needs to solve the following five challenges. APIs can help automate data collection and measure progress, map current city systems and help citizens identify opportunities for policy intervention. Each is a horizontal problem, running across city government vertical policy drivers like land use planning and waste management, and each requires collaborative action and the use of APIs, sensors, open data, and civic tech input to create connected solutions:

  1. Multimodal transport and housing affordability
  2. Food sustainability
  3. Managing the night time economy
  4. Addressing the risk of urban heat island effect (or another locational climate change priority such as sea rise in some districts or extreme weather incidents in others)
  5. Micro-enterprise development.
Not this: A city dashboard created for the city of London with broad data sets that don’t really enable action

Some stakeholders seem to think that the ability to create a dashboard is what makes a smart city smart. Luckily, there are emergent signs that other decision-makers are thinking in terms of a more solution-focused use of data and APIs. Not the London City Dashboard that, I admit, is mesmerizing in the way that a love-of-dashboards can be, but it cannot tell us anything that helps us act. Instead, examples from the US Census Bureau’s CitySDK initiative (inspired by the European CitySDK project) offer more meaningful data examples of how APIs are helping build new insights into cities that can then help us address inequity and urban risks.

But this: Austin Park Equity dashboard map showing socieoeconomic advantage of city districts and their equitable access to city resources like parkland

Building Babylon

A strong relationship between cities, civic tech and community groups is the pivot point for addressing these challenge scenarios, and in doing so it turns the smart cities risk of being top-down driven to being bottom-up by focusing and engaging on the end user — the citizen — experience.

Tech philosopher Bernard Stiegler imagines a future where, after APIs take our jobs and cause the collapse of consumer capitalism, it will be knowledge that generates new potential and can be leveraged by social organizations to enhance our lives and our cultures. He argues for the same solution that equity advocate Amartya Sen proposes: a citizen’s contributory income “granted to everyone, in order to cultivate their abilities”.

Solving those five urban challenges using API data blocks and an user-centric citizen approach represents an ideal experimental ground where we can imagine and test new economic and social systems that will make sense in post-automated city life. But living in a city designed for citizens and local businesses is not a given. In many places, we are further down the technocratic vision towards Blade Runner than we are towards the vibrant fertile cityscape of Babylon. Where is API automation taking your city?

APIdays Global in Paris on 8 & 9 December will explore the impacts of API automation on business, IT systems and broader society. Join us.

Written by Mark Boyd. Based off a presentation delivered to APIdays Global in Paris, 2014 and updated with recent examples and further thinking.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.