As this ever-evolving feedback loop of planning data is used and developed, each module can be replaced with incrementally smarter, automated versions that drive the capacity and capability growth of the system as a whole.
We think there are five critical elements to making this planning system a working reality, each with core challenges, solutions and future opportunities. Over the coming months, we’ll write a more detailed blog post about each.
You can view our vision for a 21st-century planning system here.
1. Create a digital standard for planning applications
Over 450,000 planning applications – each rich with data-rich drawings, tables and analysis – are made each year in the UK, but in formats (usually PDF or even scans of paper documents) that are non-searchable or readable by computers.
Developing a single digital planning standard would mean planning applications could be submitted directly into an accessible-to-all digital database, enabling real-time and cross-referenceable access to all applications at any stage of the process. Government could start with this by updating the National Validation Requirements, so applicants can provide data files rather than scale drawings, for example; then it could then create a digital standard for Planning Application Boundaries, allowing for easy geo-location and some basic automated screening of the application against policy, such as housing density. Over time we can start structuring ‘development descriptions’ on application forms, standardizing drawings, 3D models and evidence data; meaning that the full breadth of all planning applications and all conceivable associated data could in time be contained in a single Planning Application file.
If this standard was developed as an open source planning data dictionary, it would allow both repeatability and scalability across the UK, making is easy to monitor the performance of the system and allow us to build tools to automate validation and screen proposals against policy before case officers spend time assessing them.
Work has already been started on this by Hackney Council, PlanX and the GLA, and with the right kind of centralised or co-ordinated leadership, investment and future-proofing provided by a systems-based approach, their work could be quickly and effectively scaled nationwide.
2. Create machine readable planning policies
Planning policies, because of their judicial nature, can all too often be wordy and ambiguous, contradicting themselves and lacking clarity on what is negotiable and what is required. This increases the prevalence of litigation and consultancy costs where decisions need to be, often needlessly, challenged, creating uncertainty for both developers and local communities. To add to this the analogue way in which policies are stored make them hard to find, read or understand for either human or machine.
We need to demonstrate and accommodate an understanding that whilst planning is by necessity a discretionary quasi-judicial system, more clearly defined and accessible policies and expectations do in fact help that negotiation – for public, planners and developers – not limit it.
Government could kick start this process by making spatial policies more accessible. In particular, policies which impact General Permitted Development Orders, such as Conservation Areas and Article 4 Directions. These could be easily searchable and queried, and would drastically cut down the amount of work to decipher whether planning permission is needed or not, as PlanX is attempting to do.
We can then make accessible and standardise spatial policies on Proposals Maps by putting making data currently on local GIS systems open and downloadable. This will allow us to understand the total amount of land allocated for development, industry, green belt etc, putting us in a much better position when writing national policy.
Next, we need to tackle the rest of the policies in Local Plans, making them machine-readable and tagged with appropriate metadata. This allows us to start creating nuanced rule-based policies which can be automatically queried and used to quickly screen development proposal before they are handed over to case officers for assessment. There is no need to have highly qualified planners checking spreadsheets for compliance with home sizes, when this can be done by computers, for example.
3. Link planning policies to their evidence base
In the time it takes for Local Plans to go from drafting to adoption, economic and technological change makes many policies out-of-date. Policies related to housing affordability, social and physical infrastructure demand are particularly time-sensitive. All of this leads to a plethora of supplementary planning documents being created to patch up local plans as a result.
If policies were written in a machine-readable format, they could link directly to the evidence documents and the data they are based on. We could start this with policies based on Housing Need, to which we already have a standardised methodology, and which could be automatically updated as demographics or affordability changes.
We could then look across the whole policy landscape to identify other places where we can create standardised methodologies, identify the data they are based on and build the algorithms to allow them to adapt in real-time.
4. Automate development management processes
Every LPA is burdened with low-value, resource-intensive tasks, from validating applications to collating consultation comments and filing decision notices. Highly qualified planners spend too much time on these tasks, and many can be designed out or fully automated.
Many of these processes are the same across all Planning Authorities and are guided by national legislation, such as national validation checklists. We need to identify the most resource intensive of these and build consortiums to design and build tools that can automate them.
These tools could be provided by both the public sector or private companies as long as they meet a range of specific requirements that allows for easy interoperability. Over time we envisage there to be a growing and vibrant industry of digital planning tools. This would free planners to spend more time on proactive planning and monitoring the success and impact of existing plans, where vision and creativity are integral to their role.
5. Validate planning assumptions and measure impact
Once planning applications are approved, keeping track of what starts on site, what is completed and how buildings perform is done poorly, if at all. Understanding the impact of that individual development in economic, social or environmental terms is even more scarce.
A Digital Planning Application Standard as set out in step 1 needs to be designed to accommodate data about how the building performs once it’s built and occupied over its whole lifetime. This should include existing data such as it’s water usage and waste production, for example, to begin with, but then we need to track how its inhabitants use social and transport infrastructures, such as schools and buses. Over time, as we collect more and more data, we eventually move to measuring the impact the building has on outcomes, such as productivity and wellbeing of its inhabitants.
The new London Plan already proposes the use of sensors to monitor water and energy consumption, air quality, noise and congestion for new developments to monitor planning agreements and impact assessments. Critically this data needs to be linked to the Planning Application Standard, which becomes richer and more useful over time, can be used to validate the assumptions made during the planning stage, over time becoming the equivalent of a digital building passport.
What happens next?
To build an end-to-end 21st-century digital planning system, we have to address each of these areas. In doing so we can quickly start evolving our slow, alienating and overly politicised planning system into one that is more agile, transparent and certain, and, most importantly, delivers better homes and infrastructure where most needed.
If you’d like to help us deliver this vision, please get in touch.