Councils across Britain are employing artificial intelligence to analyze local data and identify potential sites for new housing developments. These efforts are part of a broader push to streamline land identification and planning processes.
A pilot program in 2024, with results recently published by the Ministry of Housing, Communities and Local Government, demonstrated the efficacy of AI in this area. Lewisham council in south London, for example, utilized an AI tool trained on Ordnance Survey data to identify 3,000 parcels of land suitable for development, estimating that 9,747 homes could be built on these sites. The tool, named the Small Sites AI Finder, was developed in collaboration with RCKa, an architecture firm.
The use of AI in planning departments is expanding. The Ministry of Housing, Communities and Local Government has announced a new pilot program involving 11 partnerships between councils and technology companies. This initiative, funded by £2.4 million and lasting eight months under the Proptech Innovation Fund, will explore various AI applications.
Participating authorities include the Liverpool City Region Combined Authority and the West Midlands Combined Authority, which will test an AI platform called Blocktype for housing capacity assessments. Other councils will explore AI's role in Section 106 negotiations, the legal agreements that require developers to fund local infrastructure. Birmingham City Council and Reading Borough Council will test AI Gizmo, while Uttlesford District Council and Harlow Council will evaluate Gpeto AI. The London Boroughs of Tower Hamlets and Barking and Dagenham will use software named PlanningHub.
AI in Planning Decisions
In parallel, two AI tools are under evaluation for their potential to assist in approving or denying homeowners' planning applications for projects like extensions and loft conversions. Barnet, Camden, and Dorset councils are testing a prototype AI that 'triages' planning applications, providing summaries and initial assessments to planning officers. Developed by government in conjunction with Google DeepMind, Google Cloud, Faculty, and local planning authorities, this tool aims to reduce processing times from eight weeks to four. Another tool, Extract, is now available to all English councils and uses AI to digitize historical planning documents and maps.
Addressing Green Belt Development
Many councils involved in the 2024 pilot operate in areas with significant Green Belt land, where development is strictly regulated. Sevenoaks council in Kent, with 93 percent of its land in the Green Belt and 60 percent designated as an area of outstanding natural beauty, used the Landclan tool to identify brownfield sites for development. Reigate and Banstead council in Surrey employed Vestega to assess site suitability and viability, focusing on council-owned land. Previously, their housing team could only assess one development site per year, but the pilot enabled them to identify 19 suitable parcels for council-led housing, potentially leading to 400 homes. The council's principal development manager, Catherine Radziwonik, noted that this technology has accelerated their site assessment and housing delivery pipeline.
West Oxfordshire District Council and Cotswold District Council used Go Vocal and Landclan to allow public submissions of potential development sites, reducing assessment time from approximately ten weeks to one. In the Wirral, Placemaker was used to assess over 800 sites, increasing assessment speed by 50 percent. Basildon council identified 42 potential housing sites using Landclan, though only two have since received planning permission. Bedfordshire council utilized Placemaker to assess all land parcels, including greenfield sites, based on local service access and infrastructure capacity. Durham county council also reported a halving of site assessment time with Placemaker. East Herts District Council used a website developed by Cadcorp to gather around 290 site submissions from various stakeholders.
Councils plan to use the data generated from these AI tools to inform their local plans, which guide future area development.