Description
Behind the C40’s ambition that all new buildings be net zero by 2030 is a second, larger ambition: that all buildings be net zero by 2050. The realisation that existing suburban homes must play a key role in delivering on sustainability targets is reflected in mandated improvements in energy efficiency, such as Directive 2010/31/EU. Over the next 10 years, it is estimated that adapting existing buildings to increase their energy efficiency will account for 70% of total construction work in Denmark.
This adaptation is summed up in the idea that ‘the home of the future is the one that you already live in’. But how to enable it? While architectural tools and methods are advanced for new-build, these are not directly transferable to the adaptation of existing buildings, where solutions must respond to a different set of complexities: a large number of buildings that require small scale works, undertaken on smaller budgets, and where local climate and household level decision-making is the central driver of change.
This project prototypes a tool that gives non-architects the ability to visualise climate-responsive adaptations for their homes to motivate change. The tool takes the form of an AI ‘digital twin’ that, in real time, proposes possible adaptations to an existing home that increase its resilience relative to flooding, overheating and water scarcity. The tool inputs a photo of the home (a danish ‘typhuse’) and local weather data, and outputs an adapted image of the home and a set of ‘what-if’ analytics (savings, costs of making the change). By challenging normative understandings of architecture as constant, unchanging and eternal, this tool helps homeowners perceive their own home as adaptable and contingent, and possessing a continuous potential for adaptation. In doing so, the project explores the idea of the architect as a generator of big data, questions how architects might use machine learning to bridge design into occupancy, and points to the opportunity for new & different kinds of architectural service based in machine learning.
