1. In Bettencourt & West’s work (discussed in Allometry), we see another reminder that data-driven science, from shotgun sequencing to culturomics, is adept at mining associations, correlations and sequences, but not at uncovering their meaning in aid of a theory rooted in the physical world. The conversion of data through knowledge to understanding is a wider task, but one that starts with data.
2. Historically, urban data has been confined to a costly, state-run census — often not practical to produce more frequently than once a decade — combined with land registry records (land-use data), small sample surveys and observational fieldwork (gatecounts etc). For almost two hundred years this has been the scope of urban data available for research, the information void allowing for overly normative theories of the city to gain influence.
3. We find ourselves increasingly deluged with real-time data from transit networks, infrastructure (1) (2), and 3G mobile phones. We are also able to gather large-scale citizen generated data through mass participation. The challenge is to find new analytical techniques responsive to this data. These techniques need to scale, and generally revolve around automation (data mining) and more participation (informed citizens, collaborative filtering).
4. The more freely accessible urban data is, the more usable our cities will become.
5. Availability of data creates informed citizens (e.g. UK Crime Map) who will both adapt their behaviour and want to contribute to the planning process. At the same time, it creates feedbacks to public services (see motivational mapping).
6. Planning in the age of data deluge is a large-scale, participative group activity. A software mediated citizen planning.
7. Data is neither neutral nor objective. It is acquired, processed and published by institutions with particular aims, and then interpreted by a range of organisations. The opening up of crime data will affect house prices for example, and many stakeholders will attempt to manage the interpretation of said data. Journalism and research both have important roles to play here.
8. Visualisations, in aid of narrative, are critical in dealing with large-scale data. Data needs spokespeople. This is in part a journalistic activity (see Bay Citizen’s Bike Accidents).
9. Civic data belongs to citizens and not institutions.
10. The inter-operability of metropolitan infrastructure rests on open data based on standards.
11. We are seeing the evolution of the City API (see the exemplary Rennes API). Real-time civic data now needs read/write web services. These can be collaboratively designed (Open311). Read-only feeds (e.g. TfL) are only the first step in a transition towards an opendata city.