The building stock is a city’s most significant socio-cultural and economic resource and its largest capital asset. Buildings are also where we spend most of our lives and most of our money, and where enormous potential for energy and waste reduction lies.
In order to improve the quality, sustainability and resilience of building stocks, and to help reduce emissions from them, comprehensive information on their composition, operation and dynamic behaviour are now required. However in many countries relevant data are extremely difficult to obtain, often highly fragmented, restricted, missing or only available in aggregated form.
Colouring Cities sets out to address this issue. The initiative develops open code designed to facilitate the construction and management of low cost, open data platforms, providing twelve types of data on buildings, at building level. These are needed to answer questions such as: How many buildings do we have? Which building types, uses, construction systems, ages, styles and sizes are located where? How repairable, adaptable and extendable are they? How long could they last if properly maintained? How energy efficient are they? Can they easily be retrofitted? How do they relate to the plot, street? What is their green context? Who built them and what is their ownership type, and how well do local communities think they work?
Colouring Cities also looks to advance a more efficient, whole-of-society approach to knowledge sharing on buildings and cities, allowing for permanent open databases to be collaboratively maintained and enriched, year-on-year, by citizens, academia, government, industry and the voluntary sector. Colouring London, our live working prototype, has been built and tested over the past five years using a step-by-step collaborative approach. This has involved over sixty UK partners and two hundred consultees from academia, government, industry, the voluntary sector and the community (working across science, the humanities and the arts) as well as our platform users.
In 2020 the Colouring Cities Research Programme was set up at The Alan Turing Institute to support international research institutions wishing to reproduce and co-work on Colouring Cities code at city or country level. We currently collaborate with academic partners in Lebanon, Bahrain, Australia, Germany and Greece.
Explaining the science
The Colouring London prototype is a web-based open data platform and database that contains building footprints for every building in the city. These are used as mini filing cabinets to store, collate, collect, generate, verify and visualise data, and are provided (for London) through a partnership with Ordnance Survey and the Greater London Authority.
To capture, collate and promote the sharing information, currently held by diverse stakeholders, the platform presents users with a series of maps. Each acts as a giant coloured jigsaw puzzle, relating to a specific aspect of the city’s stock, able to be gradually pieced together by colouring buildings in. Colouring can be done by collating existing open datasets, by generating new large-scale datasets (quickly and cheaply) using a variety of computational approaches, and by crowdsourcing knowledge at local level from citizens and other experts. Live streaming of data, where applicable, is also planned.
Processes can also be combined. For example computational approaches can be used to infer one building characteristic from another, based on expert knowledge of the way parts of a building relate; e.g. the size and shape of a building’s footprint plus the building’s age can be used to produce information on its size, materials, construction system, roof shape. However to check the reliability of information and methods and to maximise accuracy, features also need to be incorporated to encourage local experts and building specialists to correct and enrich datasets at local level. This results in feedback loops able to improve the quality of the database, and enhance computational methods.
Colouring Cities aims to help facilitate the free exchange of data and knowledge on building stocks across cities and countries, to improve their quality, efficiency, sustainability and resilience. It has been developed to support United Nations Sustainable Development Goals.
Colouring Cities sets out to demonstrate that permanent, high quality, low cost, open databases on stocks can be easily and cheaply built and maintained for cities and countries, providing these encourage collaborative working and incorporate diverse data generation methods.
Colouring Cities also looks to provide basic data necessary to facilitate rapid advances in the scientific analysis of cities, and to highlight the importance of studying the stock as a dynamic system and of capturing information on change over long periods of time. This is necessary to identify and unpick negative locked-in patterns that may be impeding sustainable development, and to improve accuracy in forecasting models.
Success will be measured by the extent of reproduction of Colouring Cities code, our ability to develop an efficient collaborative maintenance prototype that works for all stakeholders, and the range of applications, and research papers, in which data are used.
Colouring Cities platforms are designed to provide open data to support academic research into sustainable urban development and the scientific analysis of cities. They are also relevant to those working in planning, building design, energy analysis, building construction, maintenance and management, and to citizens and schools.
Colouring Cities platforms are also currently being considered as potential tools for disaster management, housing stock auditing, material flow tracking, and for tracking cultural asset loss. Platforms also demonstrate the power of our collective knowledge.
UK partners include Ordnance Survey, Historic England, the Greater London Authority, UCL (including the Centre for Advanced Spatial Analysis, the Survey of London and UCL Energy Institute), the Building Research Establishment Trust, the Centre for Digital Built Britain, Loughborough University, the Royal Society of Arts, the Royal Institution of Chartered Surveyors (RICS), The Royal Institute of British Architects, the Open Data Institute, and the National Library of Scotland.
Colouring London was set up at the Centre for Advanced Spatial Analysis, UCL, in 2016 and launched as a live, working prototype in 2018. In 2019 it won GISRUK’s Best Paper award, and was named by RICS in its 2019 ‘Big Data 20’, as one of the twenty most exciting innovations in proptech. It was also selected by the Open Data Institute, alongside Wikipedia and OpenStreetMap, as an example of a successful collaborative maintenance project. In 2020, the platform was used by the Centre for Digital Built Britain at the University of Cambridge, as an example of its ‘Gemini principles’, which promote openness, quality, and good management of built environment data. Since 2018, Colouring London has had 4.7 million edits, of which 200,000 have been made by human contributors and the remainder by automated processes.
Our current Colouring Cities Research Programme partners are: The American University of Beirut (Colouring Beirut), The University of Bahrain/Bahrain Authority for Culture and Antiquities (Colouring Bahrain), the University of New South Wales (Colouring Australia), The National Technical University of Athens (Colouring Athens), and The Leibniz Institute for Ecological Urban and Regional Development (Colouring Dresden).