Radical Artificial Intelligence Plan To Speed Hunt For Housing Sites

Not much in life promises a 28,400% increase in output.

A switch to artificial intelligence in the hunt for new housing sites offers just that, according to a proptech startup that has won a contract with Birmingham City Council.

Birmingham City Council has chosen Urban Intelligence to undertake a citywide review of land for homes, following a similar successful trial in London.

London and Cardiff-based Urban Intelligence will use digital methods to help the council meet a 35% increase in the volume of new housing it is expected to provide.

The bespoke geospatial analysis software means 330,000 sites can be assessed, 284 times more than the 1,160 assessed by the council last time it completed the exercise in 2017. The exercise will take three months, down from 11 months in 2017.

Urban Intelligence’s solution is to use machine learning to appraise every land parcel across the city, maximising the chances of uncovering more sites that could respond to the pressing need for homes. It will also enable scenario testing to assess the impact of policy decisions in preparation for the new Birmingham Development Plan that will operate from 2022 to 2042. 

“As the largest local authority in Europe, the city will provide a fantastic opportunity to explore the use of geospatial methods for data-driven urban development at significant scale,” Urban Intelligence CEO Daniel Mohamed said.

Urban Intelligence’s technology has already had a huge impact in the London Borough of Hounslow by multiplying the number of sites typically assessed by 191 times. 

The team’s work in the west London district has uncovered 4,200 suitable sites of which many are council-owned and small, fulfilling the London Plan’s new requirement for Hounslow council to deliver 2,800 homes on sites under 0.25 hectares.