Cities are increasingly the place where interesting work is happening within the broader open data community.
Cities, of any size, have a well defined area of influence, a ready made community and are becoming instrumented with sensors. The latter is either explicit, through the installation of devices by local government or implicit via the data automatically collected by and about citizens through mobile apps and devices.
Smart cities
I dislike the term “smart city” for many reasons. The ODI reframing of “open city” is more accessible but the idea still needs some exploration.
The original narrative around smart cities is what I’ve referred to as “the bat cave vision“.
Somewhere in the city, accessible only to a few, is a shadowy control centre. Inside it is a set of wonderful toys that are used to observe and then take action in the city.
In the Bat Cave vision the city is instrumented and the data is pressed into service to protect and help the citizens. There’s a great article called a History of the Urban Dashboard that digs into the evolution of this view of cities, which I’d recommend reading.
I think when people talk about failure of the smart city vision, they’re referring to this particular view. Given that much of that narrative seems to have been around selling hardware and infrastructure to cities it’s perhaps not surprising.
Unless you have a clear view of what problems you’re solving, then no amount of hardware and Ipads are going to help. Bat man had a first mission and then built the cave. Not the other way around.
And that’s setting aside whether dashboards themselves are really useful. To quote from the historical review:
Given that much of what we perceive on our urban dashboards is sanitized, decontextualized, and necessarily partial, we have to wonder, too, about the political and ethical implications of this framing: what ideals of “openness” and “accountability” and “participation” are represented by the sterilized quasi-transparency of the dashboard?
Data portals
If smart cities were primarily about collecting data for an administration to use, then data portals propose the opposite.
There are many, many open data portals across Europe and globally many of which focus on specific cities or localities. To date, these portals have primarily focused on publishing data from government for use by citizens, businesses and other organisations.
The target audience for using data portals is varied. And, depending on where you sit in that audience, you may find that portals are either getting in your way or not providing enough support.
Based on my experience with Bath: Hacked and from talking to those involved in other local initiatives, the critical success factor for these initiatives is finding and engaging with that audience.
Local authorities rarely have experience of working with a local technical or start-up community except as suppliers (and perhaps not even then due to procurement issues). And while authorities usually have great engagement with local civic groups, data isn’t typically a part of that conversation or used as a basis for collaboration.
City marketplaces?
It’s reasonable to consider whether there’s a middle ground between these two views. One in which there is sharing and reuse of data on both sides.
In his recent blog post, Eddie Copeland has suggested one approach: the city data marketplace. The post piqued my interest as I’ve got form (and scars) in this area.
Copeland suggests that data portals should evolve from just publishing platforms into a forum in which data can be more easily exchanged between a variety of publishers and consumers. This is something I absolutely agree with.
But I’m not convinced by the re-framing around a marketplace vision. And particularly one which proposing paying for access to data.
I think city data portals should be multi-tenant, allowing anyone in a city to be publish data into it, for use by others. And I believe that this also means that they should be owned and operated as a shared piece of data infrastructure.
Looking at each of the eight benefits that Copeland proposes:
- Increase availability of previously inaccessible datasets – a shared, openly available platform could also deliver this same benefit. By spreading maintenance costs across many different organisations, a platform can have lower overheads and be more inclusive: letting even small volunteer organisations with limited resources share their data. For commercial organisations there is scope to benefit from collaborative maintenance around local datasets, or publishing data in exchange for others doing the same.
- Increase innovation – a collaboration platform could drive similar supply-and-demand around data, indeed for citizen data collection and open, freely available dataset may be more of a draw and easier to deliver than one commercially licensed via a marketplace
- Competitive pricing – ability to compete on prices suggests that there will be several potential sources for the same dataset. This overlooks the fact that even after many years of growth, the majority of datasets in the open data ecosystem are single sourced. It’s only around a few foundational datasets, such as mapping where we see a mix of commercial and open alternatives. And open seems to be winning.
- Review and feedback – we built this type of feature into Kasabi and it exists in many data portals too. But if there is only a single source for the data then reviews and feedback don’t provide much help. While a low quality option might be quickly supplanted by an alternative version, the network effects around data usage mean that a single dataset will quickly dominate. This is something I hadn’t originally appreciated.
- Potential new revenue stream for city authorities – cost recovery for APIs and services offered is one way to ensure sustainability. But it works in an open platform too.
- Help spread best practice – slow spread of best practice is definitely an issue, but there are ways to achieve that outside of a marketplace. For example in the South West we’ve started sharing experience around publishing parking data which I hope will lead to convergence on standards and also a set of open source & free apps that can easily consume that data.
- Highlight which open data sets cities should provide – again, agree with the need to identify and find value in open data. But there’s some excellent thinking emerging around this already.
- Policy based on hard data in place of modelling – data-driven policy making, ftw!
My general intent here is to start a wider dialogue around the future direction of city data portals, rather than critique Eddie specifically. I’ve itemised the points to identify areas of agreement as much as disagreement.
I’m just concerned that we might waste time exploring a paid future rather than embracing a commons based approach. I don’t think we’ve fully explored or experimented with the potential benefits of shared, collaborative open data platforms.
Let’s talk about zoning
You can hear me talking about digital graffiti and data infrastructure for cities in this talk from a few years ago.
The graffiti metaphor I used in my talk was deliberately intended to be challenging. A city data infrastructure shouldn’t reflect a single world-view or be a clean, sealed environment. It should allow itself to be annotated and re-purposed by its community whilst continuing to deliver value to everyone.
One of the points I was trying to make in the talk is that local government are already really good at helping to divide up physical cities to allow communities to grow, meet and do business. We call it zoning.
What is the equivalent of zoning for digital cities?
I’d argue that it will involve:
- defining a space (or spaces) within which anyone can share data, either openly or with additional controls.
- creating touch points, identifiers, that connect datasets and our physical and digital spaces
- describing the rules of engagement for using that space, such as open licensing, transparency around data collection, and anonymisation practices
- support the collection of data from the physical city, by providing access to infrastructure and perhaps permission to instrument the city
- providing an environment that supports both public and private spaces and enterprises
Zoning might not sound exciting. It’s unlikely to sell hardware and may not directly inspire business models. But its reusing a successful, proven pattern. It’s also focused on managing access to, and use of (public) spaces, which feels more inclusive to me.
Lost Boy 