What kinds of buildings make up London? What are their size and shape? How are they made and what of? Who built them and which are the most energy efficient? Where can you find different types? And which do you like, and think contributes to the success and well-being of the city.
We are collecting twelve types of data to answer these and many other questions about London. To do this we are testing four approaches: existing bulk dataset collation; computational generation; crowdsourcing and livestreaming.
Click on any box to find out more about our different data categories.
Owing to the complexity of the stock as a system, we have tried to group our datasets in as simple a way as possible. The 12-category-grid above also doubles as our logo, and as Colouring London's control panel, allowing you to easily navigate across our maps.
The grid contain over 50 subcategories of data. These have been selected through consultation with over fifty organisations combined with a detailed analysis of data needs of researchers working in sustainability science and energy analysis, and urban science, based on the assessment of academic papers. The list not exhaustive and new subcategories can be added at any point. Our work also involves investigating and selecting suitable data formats.
The grid is divided into two sections. The first includes data on Location, Use, Type, Age, Size, Construction and Context contain information on the form and use of the building. These are relevant to multiple applications ranging from calculating and predicting volume of energy emissions, to assessing housing quality and supply, to predicting structural failure or informing the development of local plans. Using these data, open 3D rule-based city models can also begin to be be built. The remaining five categories on the grid, Team, Planning, Sustainability, Dynamics and Community all have multiple functions, and are designed not only capture and contain data to support analysis are used to help improve quality and sustainability in stocks through greater engagement and knowledge exchange from all those involved in.
Location data is the first type of data in the grid as all other data types rely on coordinates. addresses and building footprints for every building, to allow data be collected, mapped and spatially analysed. The greater the number of location categories filled on our map the darker the building colour is, allowing you to quickly see where more data are required.
Building footprints are a very important type of data essential for Colouring Cities platforms to work, as they act as mini filing cabinets, storing, locating and visualising contributed data. Use of UK building footprints (OS MasterMap/ OSMM) for Colouring London has permitted by Ordnance Survey, the UK's national mapping agency, (under the Greater London Authority's licence) though government copyright restrictions mean that though we can use these footprints to capture and visualise data we cannot release these as open data. We instead release coordinates for data (following government release in 2020, of Unique Property References Numbers (UPRNs) meaning our data can be mapped as point not polygons unless to an OSMM licence (held for example by universities and local authorities) is available.
Colouring Cities collaboration with the Ordnance Survey and the GLA helps demonstrate why the open release of comprehensive, updated national mapping agency data at building level is essential if UK cities are meet ever more stringent emissions and sustainability goals. It also shows how the release of national mapping data, including UK address data, has the capacity to unlock vast repositories of currently untapped knowledge about the city's buildings, held by diverse sectors and disciplines, and by citizens, and able to be used, and donated, for the public good.
Though steps have been made by government to opening up OSMM full release of this high quality, constantly updated dataset, and of UK address data still appears some way off. This is needed to effect a whole-of-society approach to improving the quality and sustainability of the UK's building stock. It will also bring the UK in line with many other countries openly releasing location data on stocks, and prevent it falling behind in many areas of urban/sustainability research. In the meantime we will try to collect and release as much information on Location as we can. This includes building name, number and address, lat/long coordinate, UPRN and OS TOID (geometry/polygon) number.
The most significant current source of open location data is OpenStreetMap (OSM), which has been at the forefront of driving release of open data on cities since 2004. Its collaborative maintenance model, along with Wikipedia's has also strongly influenced Colouring London's approach. Originally set up to crowdsource data on streets, OSM also now focuses on providing open data on buildings. All data generated by Colouring Cities platforms are therefore designed to link seamlessly to OSM, with OSM IDs, and attribute data integrated wherever possible.
In order to analyse cities in a more scientific way, their basic composition first needs to be understood. One of the most commonly used types of data in planning, urban analysis and energy assessment is land use data. This provides information of the kind of buildings that exist and how many examples of each different kind are there and how much floor space is available for different types of activity. It also helps answer questions such as where do specific types of land use cluster, is this the same in all cities, and how does this affect the way the city operates?
Despite widespread demand for land use data, at property and building level, for cities as a whole, comprehensive open land use data are still not available in the UK. Such data are however held by Her Majesty's Revenue and Customs' Valuation Office Agency, for approximately 27 million taxable UK properties. Ordnance Survey also holds detailed land use data within its Addressbase product range. Once these datasets have been made available Colouring London will be able to transfer efforts capturing, collating and generating data, to integrating, verifying and enriching thesese datasets instead.
Fortunately information on building use is quite easy to collect, and can usually be worked out simply by viewing the front of the building from the street or using a streetview image. Over 90% of taxable properties in the UK are residential. Recognising houses is quite straightforward. Blocks of flats can be differentiated from offices by the presence of curtains. Identifying activities within non-residential buildings is slightly harder but activities can often be determined by the shape and size of buildings, their window layout as well as signage to the front, or at entrances. We're particularly keen to encourage schools to add data for this section, as schools were instrumental in capturing data for Britain's first ever National Land Use Survey in the 1930s. We will be combining crowdsourced data with whatever open land use data we can find as well as using historical network data to predict the geolocation of commercial buildings
If you would like to add data just select a building and decide which land use 'Group' the activity belongs to. These are: Agriculture; Fisheries; Outdoor amenity & open space; Amusement & show places; Libraries, museums & galleries; Sports facilities & grounds; Holiday parks & camps; Allotments and city farms; Transport tracks and ways; Transport terminals and interchanges; Car parks; Vehicle storage; Good and freight terminals; Waterways, Energy production and distribution; Water storage & treatment; Refuse disposal; Cemeteries & Crematoria; Post & telecommunications; Dwellings; Hotels, boarding & guesthouses; Residential institutions; Medical & healthcare services; Places of worship; Education; Community services; Shops; Financial and professional services; Restaurants & cafes. Public houses bars & nightclubs; Manufacturing; Offices; Storage; Wholesale distribution; Vacant; Derelict; Defence; Minerals & quarries. When saved this will also automatically classify the building into one of nine building 'Orders':
We've chosen to test the National Land Use Database (NLUD) classification here. There is currently no single land use classification system currently in the UK. Look up tables will also be generated to allow data be compared with VOA primary and special category codes , OS AddressBase classification codes and Planning Use Classes. If the building has several different uses you can pick our 'Mixed Use' option. We're also planning to collect information on vacant and derelict buildings.
The kind of activities and number of people a building was originally designed to hold, as well as the period in which it was built, will affect a building's form, including its size, shape, decorative features and layout. Such characteristics are also used to group buildings into specific types or typologies, where copies or versions also exist.
Understanding the location of different building typologies is important in areas such retrofit of buildings to improve energy efficiency, allowing retrofit methods and budgets to be more accurately targeted and understanding how buildings and packed together. Understanding survival rates for different typologies, and identifying and retaining adaptable ones, is also necessary to reduce unnecessary waste and energy in construction, and to learn from the past to build more long-lasting buildings for the future.
When combined with footprint, size, height and age data, typology information also helps build a picture of a building's 3D geometric form. This is increasingly important in energy and urban heat analysis, and in the development of 3D rule-based digital city models able to simulate future
planning and energy scenarios.
Some of the data on our 'Type' category will be inferred automatically, using footprints and other attribute data. This includes subcategories for basic type classification, dynamic tissue classification (indicating susceptibility to change), and building adjacency (e.g is it detached, semi-detached). These datasets are being produced in discussion/collaboration with research collaborators. Links to 3D open typology models will also be provided in future. We are hoping to crowdsource information on the remaining subcategories i.e. original use (for which historical information is required); the historical period in which the building was built; and the kind of additions or mutations made to the original type, e.g a side, rear or roof extension. Here contributors historical knowledge of local areas is particularly valuable.
Building age is also commonly used, and recorded, by architectural historians, building conservationists, heritage specialists and urban morphologists.
Information on building age, generated from date of construction, is extremely important for geolocating building types. More recently building age data have also become increasingly used in energy and urban sustainability research, particularly in emissions analysis and urban heat assessments. Here construction date is often combined with other attribute data to help describe the building's form, particularly its geometry and volume. Age data, combined with historical construction and demolition data (captured in our 'Dynamics ' category) are also needed to produce actual building lifespan data. This is required to forecast when and where specific areas of the stock may have to be replaced, and to plan lifespan extension. (Age data will also in future be use, within our 'Sustainability' category, to provide an indication of potential building lifespans as well.
Building age data, are held by the UK government's Valuation Office Agency, for all taxable properties for use in property tax assessment, but are not openly released even for academic research. Colouring Cities develops interdisciplinary approaches to assist the capture of of the most accurate building age data possible. Our 'Age' and 'Dynamics' categories have been designed, in consultation with historians and heritage experts, to encourage input from architectural historians, civic societies, building conservationists and others with expert knowledge of building history. We are also experimenting with automated approaches to age data generation which allow faster coverage of the city. Here the age of buildings is inferred by the age of the road. Local historians are invited to verify the data or alter it where applicable. For contributors new to dating a good place to start by using local history publications, old maps and local authority conservation to start t date your street.
Data on the size and geometry of a city's buildings have many applications ranging from use in 3D digital city models, to understanding implications to changes to the height of a city's buildings, to analysing and predicting energy use, and greenhouse gas emissions, and the build-up of urban heat.
Data on the dimensions of buildings are also relevant to many other areas of urban research, from analysing housing capacity and identifying areas suitable for densification, to observing (within urban science and urban morphology) long-term patterns of change within urban form.
In the 'Size' category, data on building height, number of storeys, and floor area are all collected.
Team captures data on developers, designers and builders. For most buildings this requires expert input from professional and amateur historians, as over 85% of London's stock is estimated to have been built before 2000. For annual new builds (which represent under 1% of the stock) we encourage construction firms and professional bodies to add and update data. Awards and quality marks are also included here, to celebrate construction firms' commitment to sustainability, and industry skills and expertise. The section is also designed to drive up new build quality by enabling the longevity, energy performance and workability/attractiveness of buildings (as viewed by users/citizens) to be more easily tracked over time, and used to improve building design in future.
Planning captures data on whether a building is protected from demolition or change, and also includes a link to the UK planning portal, where recent historical planning application information is kept. Here the idea is to test a live stream traffic light system which allows planning applications to be visualised, with the colour of a building to change automatically depending on its planning status (i.e. orange for submission, green for approval, red for refusal and pink for appeal). Buildings will also colour based on, largely crowdsourced, information on whether construction has commenced , and on whether the building has been completed and occupied. This is particularly important in understanding the spatial distribution of available housing. Users will also have the option of colouring a building if they think an application is considered likely to be submitted. This helps communities understand in advance if they buildings they consider to be of local importance are being threatened, and allows them to use the 'Community category' to highlight buildings they have found to contribute to the area/city. Both 'Planning' and 'Community' categories are also designed to help planners and developers gauge/anticipate, at the earliest possible stage of the planning process, community feeling in relation to demolition/new build in relation to an area or site, and gain deeper insights into the value of local buildings in the operation of the local area as a whole.
Sustainability is a broad category designed not only to capture data on energy performance but also to stimulate discussion on inclusion of new types of data designed maximise building longevity and minimise emissions. Data categories of future interest include for example the repairability, adaptability and potential lifespan of buildings/or building components, as well as the toxicity of materials.
Dynamics captures data on the evolution of the city, on incremental development within plots over long periods of time, and on building lifespans. These are needed to track rates of change, assess typology survival rate, predict lifespans and anticipate vulnerability to demolition and system failure. They are also required in urban metabolism studies to assess and predict the volume of energy and waste flows occurring in the city, generated from changes to the stock, and differences in the amount of energy used in new material extraction/new construction/old building disposal compared with building updating/extension/reuse. For this category we are beginning to test a range of data capture methods, from the crowdsourcing of statistical data on historical constructions and demolitions on sites, to the use of machine learning to extract big data from historical maps. We currently provide a map on which historians can co-work to colour in, extremely difficult to access, historical data which can then be used to inform analysis, forecasting models and dynamic simulations.
Categories are not set in stone. Where there is general consensus that a category or subcategory should be adjusted or added, relevant changes can be made. You can add your suggestions to our discussion threads here.
The plan in 2022 is to also provide open georeferenced raster maps relating to the physical landscape including those illustrating historical development, greenspaces and waterways, geology and flood risk. We will also be working on integrating open 3D models and interactive data analytics features.
We are gradually releasing subcategories for testing. If you are able to help us by adding or checking the accuracy of data, just go to Edit Maps, click on a building, choose any category of interest and fill in any information you can. Every entry and/or verification helps. Some categories are easier to fill than others. Our 'Community' category is perhaps the best place to start. Here you can simply colour any building you think contributes to the city.
Examples of subcategories within our 12 main categories are shown below. We're trying to keep collection as efficient and simple as possible.
Building Name & Number
OS Open ID (UPRN, TOID)
3D form (generic)
Dynamic tissue type
Open3D procedural typology link
What year was building begun?
Front the same date as core?
Web links to historical information
Height & number of storeys
Plot to footprint ratio
Green context (gardens)
Green context (street)
Team & awards
Where applicable : Client, Developer, Designer
or design source, Engineer & Surveyor, Builder/s.
Awards & Quality marks
Planning portal link
Is it protected? Listed building/conservation area?
What is its planning status?
Energy performance rating
Last major retrofit
Repairability rating for system type
Lifespan data - Constructions and demolition date pairs for history of site
Links to historical information sites