Behind the case study: How open capture of data will help improve buildings in line with the SDGs
The Colouring Cities Research Programme (CCRP) is a UK-based program that develops and operates open-source databases to capture building data for collaborative urban problem solving. Run by The Alan Turing Institute – the UK’s national institute for data science and AI – the CCRP recently submitted a case study to our InfraTech library.
Today we are joined by the CCRP's Project Lead, Polly Hudson, who explains how CCRP works, the implementation challenges, and how InfraTech will help the sector to address common challenges.
Can you tell us about the CCRP and the stakeholders involved?
The CCRP works across countries with academic and other institutions specialising in building stock[1] research to co-develop and test open platform code that gathers and shares information about the building stock in these locations. Colouring Cities platforms also act as free tools to support engagement from government, industry, academia, the third sector, and citizens in data collection and verification
We currently have partners in Australia, Bahrain, Britain, Canada, Colombia, Germany, Greece, Indonesia, Lebanon, and Sweden, and have created a global network of connected but independent platforms to share knowledge and data at national and local levels.
What challenges does the CCRP solve?
Our aims are to:
- Facilitate collaborative urban problem-solving by creating affordable, sustainable, open, spatial databases for this purpose
- Effect a step-change in the amount and type of building attribute data available to people who are involved in creating, managing, using, and researching into buildings
- Rapidly advance our collective scientific understanding of global and national building stocks as complex dynamic systems
- Help increase the quality, sustainability, efficiency, and resilience of buildings in line with the UN Sustainable Development Goals, and engage stakeholders in this process
- Advance discussion on data ethics issues relating to the release of building attribute data.
What solution has the CCRP has developed, and how does it work?
The CCRP has developed and is testing open code for affordable, sustainable, spatial databases on building stocks. Platforms capture, collect, and visualise spatial data on the composition, performance, and dynamic behaviour of buildings – or combine and use the data already captured to generate new large-scale datasets ready for verification at building level.
The databases collect 12 types of building-level spatial data:
- Location
- Land use
- Typology
- Size
- Construction
- Age and history
- Street context
- Team
- Planning controls
- Energy performance
- Resilience
- How well the community thinks the building works.
Special features currently being explored include colour-coding of planning applications and streaming into building footprints, and capturing data on emergency scenarios.
The system can also tackle problems that commonly arise with building attribute data. These include fragmentation, incompleteness, low quality, poor standardisation, low geographic coverage, and low granularity – as well as security and privacy considerations associated with data capture.
What were the main challenges in implementing this?
It has been complicated to retain continuity of high-quality research software engineers for long periods, to maintain open code quality, operate efficiently, and minimise knowledge loss. In response, we moved to a model where a percentage of expertise from each partner is pooled, and long-term relationships with academic engineering teams are built at international and national level. This has taken time and has involved prioritising data ethics and the wellbeing, security, and privacy of contributors, partners, and building occupants / owners.
To get the prototype to an operational stage has required funders and administrating departments with vision – the long initial period of funded research (from 2016), was essential. Grants that focus on instant deliverables and immediate impact were unsuitable for the early stages of this type of work.
Feedback loops and algorithms supporting automated attribute classification still require considerable work. Ethical challenges have also arisen in relation to collecting certain types of data that are useful to sustainable development research.
However, the incremental approach has allowed solid and sustainable foundations to be built. It has enabled us to begin rapidly expanding research collaborations and developing a network of permanent public databases on building stocks that are affordable, accessible, trustworthy, of relevance to multiple stakeholders, and able to be slowly improved and adjusted over time.
How can the sector better support the adoption of InfraTech to address environmental, economic, and social challenges?
Ongoing investment is needed in data sharing architectures that maximise the open exchange of infrastructure data across sectors, and at the same time afford sufficient safeguards for contributors and asset owners / occupiers.
Greater awareness of the power of visualisation of spatial infrastructure data to support stakeholder knowledge exchange, and of investing in the next generation of visualisers is also required.
Stricter guidelines will be increasingly important for organisations releasing attribute data at the property / building level, particularly relating to the interiors and performance / operation of domestic buildings, as a new generation of open access 3D city models begins to emerge.
Notes
[1] |
‘Building stock’ is defined by the Designing Buildings Wiki as the total number of buildings in a country, region, municipal area, or estate. |