The studio will focus on Indoor Climate and Energy Building Performance using three performance metrics: Daylight, Thermal and CFD Modelling.
Our building practice is facing an increasing demand to rethink its design methods, as it struggles to respond to new requirements for energy efficiency, sustainability, and economic and societal changes [1].
As our societies are challenged by escalating urbanisation coupled with the energy crisis, we need to develop design practices that enable smarter use of resources, higher energy conservation and better social and urban programmes, while maintaining high architectural quality.
Building Energy Simulations lie at the center of this effort, as drivers for analysing, mapping and evaluating the energy performance of our built environment [2].
Yet, our built environment today is much more than building typologies and numerical matrices and involves complex spatiotemporal structures and networks expressed through human, material and environmental interdependencies [3].
How can we design using both qualitative and quantitative data of various resolutions and complexity?
In an era marked by escalating climate concerns and the imperative for sustainable development, the utilisation of AI holds immense promise in handling complex data as well as mitigating the environmental footprint of buildings while optimising their performance.
In the ‘’AI for Building Performance’’ studio we aim to harness the potential of AI methods combined with open-source technologies as drivers for resilient building design strategies. The studio will focus on Indoor Climate and Energy Building Performance using three performance metrics: Daylight, Thermal and CFD Modelling.
The vision of the studio is to develop working methods based on real data to solve real world problems of the case study building, suggesting a transition from design models to knowledge models.
The studio aims to employ the students with fundamental conceptual and practical skills for developing data-driven AI methodologies for mapping, analysing and evaluating building energy performance . Specifically, the learning objectives of the studio are:
DF Predictor aims to revolutionize daylight prediction in architectural design by implementing the conditional generative adversarial network (cGAN) Pix2Pix to predict daylight factors, motivated by the need to improve efficiency and accuracy in daylight analysis. Daylight factor analysis, an integral part of the early design stages and mandated by building codes, is typically lengthy due … Read more
This project aims to develop a web-based application designed to optimize the construction of various architectural elements. The project requirements include several specifications from the user, such as the type of architectural element, its overall dimensions, and material selection. Additionally, the project involves specific training of a model using generated structural datasets from Karamba3D, which … Read more
Concept In a world where urban areas are predominantly developed and the heat island effect is intensifying, the construction industry significantly contributes to environmental challenges. Instead of focusing on tools that promote new construction, our goal is to provide a tool that analyzes existing buildings and suggests improvements. This approach aims to enhance sustainability and … Read more
This project aimed to develop a daylight predictor to facilitate and generate well-informed adaptive reuse projects, with a specific focus on providing sustainable design solutions for low-income housing. Los Angeles (LA) was selected as a case study due to its proactive open data initiatives and commitment to adaptive reuse. This proposal provides a snapshot of … Read more
Introduction Our project, Materializer, leverages the power of multiple self-trained machine learning models to predict material quantities based on an image uploaded by the user and the building coordinates. This innovative approach utilizes image segmentation to isolate buildings, image classification to read material pixels, and a height prediction model for buildings lacking height information in … Read more
