Emma Warneford, Senior Specialist in Planning at Newcastle City Council, and Prof. Dani Arribas-Bel, Deputy Programme Director in Urban Analytics at the Alan Turing Institute, tell us more about the progress of the pilot to date.
Figure 1: Newcastle City Council is helping explore how scenario modelling can be used in complex land use decision making.
Delivering for a growing city
Newcastle City Council (NCC) are looking to build their evidence base to optimise the use of land in the city to ensure that it meets the needs and aspirations of current and future generations. NCC are undertaking strategic spatial planning analysis to balance the delivery of various policy priorities. These include supporting inclusive economic prosperity, moving towards net zero emissions, improving environmental outcomes, and delivering accessible housing to sustainably meet the needs of local communities and a growing population.
NCC faces challenges in identifying sites for future housing development in order to meet housing supply targets, due to the land use constraints they face. Development options are limited by the river Tyne running through the city, and the tightly drawn surrounding Green Belt area.
Identifying sites for housing development therefore requires the consideration of different approaches and options including brownfield redevelopment and urban densification. Understanding where these options are viable can be challenging and considerations need to be taken into account about how these different options impact on NCC’s wider policy priorities.
To support NCC in the identification of sustainable development sites, we are developing a modelling system that leverages data science and AI to cut through inherent complications in land use planning. This will support the planning of strategic interventions by:
-
- Evaluating the impact of high level development options against policy priority indicators
- Using machine learning and AI to suggest interventions to achieve policy outcomes
Leveraging data science and AI
The first stage of this pilot was to gain an understanding of the priority competing aspects of land use in Newcastle that could be used as indicators in the modelling of scenarios and establish a baseline. These would help NCC assess the impacts of different options for housing developments.
Through a number of collaborative workshops, the following indicators were determined:
-
- Environment: Air quality
- Economy: House prices & job accessibility
- Society & Health: Accessibility to green space
Figure 2: Publically available datasets have been used to quantify development indicators.
Publicly accessible datasets have been identified to quantify these indicators, alongside the Urban Grammar and Synthetic Population Catalyst land use data products previously developed by the Alan Turing Institute.
The existing baseline will be modified through planning scenarios which are currently being co-produced with NCC, such as:
-
- Consideration of low-density or mid-density residential development
- Densification of inner city areas
- Redeveloping brownfields into dense neighbourhoods
- Combinations of the above
The modelling system will then be able to illustrate the trade-offs between competing objectives through an evaluation of the key indicators.
An Ensemble Engine will also be developed which can suggest scenarios that can lead to desired outcomes. This will allow planners to input target growth goals, such as additional housing, and a number of interventions would be suggested, using machine learning and artificial intelligence, that are capable of delivering this goal.
What’s next – developing a visualisation tool
An interactive tool to wrap all the analytical work in a friendly accessible package is under development to support this work.
NCC will be able to use this exploratory tool to take a systems wide view of land use and comprehensively explore the trade-offs between indicators to support land use decision making, and simulate the impacts of proposed land use change to mitigate the risk of disbenefits and unintended consequences.
The Alan Turing Institute is an advocate for the benefits of open AI and data science, and will make the code and outputs open-source where licensing permits on conclusion of the pilot, which is expected in June.
The findings from this pilot will be incorporated into the National Land Data Programme report which will be published in summer 2023. For more information about the programme and to get in touch, contact us at NLDP@cabinetoffice.gov.uk.