GIS, supporting the delivery of more digital projects in Hackney

Hackney’s corporate GIS team sits within our Data and Insight team within ICT. Our mission is to support projects, colleagues and residents to make the most of Hackney’s spatial information. However, we have a blocker: most of the work we do comes to us via colleagues who already use GIS, or know someone who uses GIS, or at least have an idea about how spatial information can help them in their jobs. This represents a fair amount of people, and their number has increased a lot since we launched our new Intranet GIS (Statmap Earthlight) 12 months ago. However we think we could do better and we want to find new ways of supporting our digital delivery teams with spatial analysis, particularly those who don’t yet know the power of it as a tool.

So last week we gathered our colleagues from across our Digital and Delivery teams, and explained to them how we are delivering our mission across Hackney service areas. A possibility for our colleagues to mentally match our offer with the various situations they encounter when they are out. A chance for us to discover use cases and ‘GIS gaps’ we wouldn’t have suspected otherwise.

We detailed the 3 work streams we are following to realise our mission, as follows (every local authority’s GIS team will probably recognise itself).

Infrastructure and data governance: we administer and develop the corporate spatial data warehouse, which has been providing a single view of Hackney’s spatial information since 2008 and integrates with other council systems through web services and APIs. Most importantly, we’re supporting our community of GIS users, and working with our data owners to introduce new ways of managing data quality. The first step of this journey was the creation of our metadata model and data catalogue.

Data Analysis and Insight: we are using spatial analysis to answer questions and support decision making. We use different supports to develop and share results, like Qlik Business Intelligence dashboards and Jupyter notebooks.

Data sharing and web mapping: we are providing technical support for service areas to share their data when they have to. We developed and maintain Hackney’s open mapping portal. We are, however, trying to move away from the monolithic portal approach and are practising our ‘Agile delivery’ of web maps focusing on a specific sharing need, using Leaflet.

Our presentation had two immediate benefits: firstly, our user research team requested access to the Intranet GIS with the view that it could help in discovery work that they do with services to understand user needs. Secondly, a colleague from Delivery asked how non GIS-users can realise the richness of our spatial data repository if the only way to browse this repository is… to be a GIS user. Do we need a channel for non-GIS users to discover GIS? What should it look like? This will guide our thinking for the preparation of the ‘Annual GIS day’ in November. Any suggestions welcome.

You can see a subset of the examples we showcased in this slide deck. We are keen to hear about how our approach compares to how other GIS teams are operating and think of ways we can learn from others to improve our offer in Hackney.

An open data standard for planning applications?

We’re working to find out what a digital planning application service would look like if it were “so good, people prefer to use it”. However, one of the early things we learnt was that high quality data is the key enabler of providing a better digital service. 

We need to ensure our work on planning applications considers the opportunity for an open data standard, so that multiple authorities can benefit. Such a standard could support working across multiple planning authorities, e.g. for applications on borough boundaries. It could support interoperability within a future digital planning ecosystem, including front and back-end systems, centralised portals and registers.

To that end Hackney convened a session with colleagues from MHCLG, iStand UK, Future Cities Catapult and the GLA, to help us explore the idea of such a standard.

We (SB and TP) attended as members of the Data & Insight Team at Hackney, with an interest in the development and use of open data standards. Here’s an outline of what we learned…

Scope of the problem

Building a domain model is complex because of the interrelationship of policy, application types, land types, and different stages of decision-making.

Different planning authorities have different processes and timeframes, using different terminology, as well as different technology (for some insight on the fragmentation among local planning authorities, see Molly Strauss’ account of the GLA’s ongoing efforts to improve London-wide planning data). Some issues are only relevant in certain areas (e.g. radon gas).

In addition, much planning application information is contained within unstructured documents. In the digital planning portal, used by a large proportion of applicants, data is often entered into free text fields without validation against definitive registers. It is hard to tell how much of this content could become formalised and we must expect that some of it will remain unstructured. Similarly, much of what goes into planning decisions isn’t exposed as data yet (for example, local plan policies are not always rendered into spatial data).

Previous related efforts

We heard about efforts to consolidate data across local planning authorities. This seems to have been focussed on publishing unified planning application registers for counties including Hampshire and Surrey, via planning “hubs”. Among the challenges mentioned were:

  • Authorities using the same planning management software, but in different ways.
  • Difficulties agreeing on nomenclature.
  • Licensing restrictions, e.g. Surrey provides an API licensed only “for non-commercial and personal use” apparently due to use of Ordnance Survey “derived data”.

See the LGA schema produced as a result of this work.

Methodology

We were reminded that a sustainable open data standard is more likely to emerge if we follow a few general rules:  

  • Start small. We should aim for a compact standard that could be extended later, following the ‘plugins’ approach of Open Source projects.  
  • Allow a lot of time. The standard should emerge from the model, not the opposite. It is also dangerous to pursue a definitive standard too early: when it starts being adopted, it becomes harder to change it.
  • Involve the right stakeholders. A standard is more ‘believable’ if it has been shaped by many diverse stakeholders, including from government entities. If not, it is likely to be reflecting the needs and culture of a limited group.
  • Plan for governance: in order to last and evolve, a standard needs a proper governance structure, led by a credible entity and offering a channel for stakeholders to propose changes.

What’s next?

Work on a minimum viable product for a digital planning service is about to start, involving the partner boroughs and Snook and Hactar. Using the advice outlined above, the aim will be to keep the data model as lightweight, high-level and open as possible, so it can be a good candidate for evolving into a standard.

This will only happen with input from a range of people with different perspectives. Hackney Digital Team would love to hear your ideas and learn about similar initiatives.

By Sandrine Balley and Tapan Perkins