>>  MACAD is the world’s first online accredited master programme in the field of AEC (Architecture, Engineering and Construction), with a focus on Computational Design and Integration of new digital technologies in the Architectural design and construction process <<<<

In the third and last module of the MaCAD – Master in Advanced Computation for Architecture and Design, students explored the use of the most advanced tools for designing through machine learning and artificial intelligence.

The module posed new learning challenges and research questions to the students, who were connected online from 25 different countries:
– How will Artificial Intelligence (AI) change the AEC sector?
– How can designers train, apply, and evaluate AI models?
– How can they source or develop appropriate datasets of design intelligence?
– Can we use machine learning to analyse, classify and optimize geometry?
– How can Generative Adversarial Neural Networks (GANS) models inform generative Building Design? 

 

The impressive results of the module proved once again the success of the MaCAD educational model, connecting students from 25 different countries with an outstanding network of professionals and researchers in the field of Architecture, Engineering and Construction (AEC): between April and June students worked with Senior Faculty Stanislas Chaillou (former Data Scientist & Architect at Spacemaker.ai), Angelos Chronis (Head of the City Intelligence Lab at the AIT Wien), Gabriella Rossi (PhD Researcher at CITA), Zeynep Aksöz (researcher and practitioners at TU Vienna, Karamba3D, Open Fields) and MaCAD Director David Andrès León

Applications for 2021-22 edition are open!

Want to know more? You can schedule a video or text chat with the MaCAD Coordinator or Apply using the buttons below.

HIGHLIGHTS

AI – Urban Voids
MaCAD Students: leksander Mastalski, Amal Algamdey, Amar Gurung, German Bodenbender and Felipe Romero.
MaCAD Senior Faculty: Angelos Chronis w/ Lea Khairallah
Course: Artificial Intelligence in Architecture Studio

Big urban data now being easily available online, there is an opportunity to utilise this information to generate new relationships between various features within the urban fabric. This new information will be useful not only to architects and developers but also to any individual or institutions looking to understand the urban features whether it is to set up a new cafe, particular shop, housing, school or a clinic. Urban Voids is a data-driven approach to analyse and predict potential locations for the addition/intervention of amenities within the city. The predictions and scores are based on a series of urban analyses, simulations and the use of KMeans clustering. The aim is to create a tool that will work on a feedback loop system where the information is constantly being updated. At the back end, there are the various analysis, simulations and clusterings, the results from this are then being visualised in a web-based platform (Mapbox) and to complete the loop, the user inputs a new location and amenity type to generate a new prediction and scoring for the new information.

 

See the full project in the IAAC Blog at this link

Context Decoder
MaCAD Students: Keshava Narayan Karthikeyan, Laura Marsilo, Nataliya Voinova, Nawapan Suntorachai
MaCAD Senior Faculty: Angelos Chronis w/ Lea Khairallah
Course: Artificial Intelligence in Architecture Studio

Based on the hypothesis that it is possible to define a correlation between urban fabric and social behavior patterns, the project “Context decoder” draws connections between the urban fabric and society, analyzing the similarities and correlations between input features to give designers a deeper comprehension of the place. The objective is to achieve holistic context interpretation for the selected location by using spatial or non-spatial factors.
Context Decoder combines several components, which have existed separately for quite a long time already, into an integral instrument with new possibilities. The project was developed in 3 processes including: Definition of Methodology and Data Selection, Analysis and Machine Learning for clustering, Integration to Application

See the full project in the IAAC Blog at this link

In-between Spaces

MaCAD Students:Hesham Shawqy, João Silva, Polina Hadjimitova, and Varun Mehta
MaCAD Senior Faculty: Angelos Chronis w/ Lea Khairallah
Course: Artificial Intelligence in Architecture Studio 

The project revolves around using artificial intelligence to support, document, and resolve issues of the unplanned portions in urban areas. The scope of IN BETWEEN SPACES expands way beyond documenting informal settlements, It involves mapping empty spaces within informal settlements through a participatory approach. Which not only aids in improving accessibility by generating new street networks. But also locates potential nodes for various infrastructure development within these organic urban systems.

The process documentation consists of resources to create data sets and train a Pix2Pix model to map unmapped portions. Followed by the use of web-based resources such as web flow and map box; focussing on their integration for building an interface to collect data through a participatory approach. It further shares how collected data can be used to inform a different Machine Learning algorithm through a Live Google sheet and how the results can be used to analyze and generate street networks within the settlement, improving the accessibility within these settlements
IN BETWEEN SPACES intends to enable anyone to replicate and grow on any/all parts of the project.

See the full project in the IAAC Blog at this link

URBAN LUX // PREDICTING RADIATION USING MACHINE LEARNING
MaCAD Students: Hesham Shawqy and German Otto Bodenbender
MaCAD Senior Faculty: Stanislas Chaillou  w/ Oana Taut
Course: Digital Tools for Generative Building Design

To best accommodate rapid urbanization while making cities more sustainable, livable, and equitable, designers must utilize quantitative and qualitative tools that allow them to have real-time results to make informed decisions at the design stage. On the other hand, the current urban design process is often a long and time-consuming workflow that typically involves a team of architects and urban planners who conceive a handful of schemes based on zoning requirements with the help of CAD systems. Due to the scale of the problem as well as its complexity, designers have really limited time and capacity to try to design and test as many iterations as possible. This often results in just a few different options that do not resolve the problem in its full spectrum.

Based on this challenging scenario, the project researches and develops a new methodology using machine learning that can help us to predict urban radiation in real-time without the need for time-consuming computational models.
The model was trained using multiple configurations and achieved a good generator loss, and was tested using a specific test dataset. These are a series of tiles that the model saw before, and were used  to understand and compare how well the model predicted. Students tested on three different sets of tiles. The first set presented a small variance in the parameter space, having almost square and most typical urban block layouts. The second set had more diverse urban blocks with irregular dimensions that resemble the rectangular blocks found in the city of new york. In the third dataset, students broke almost completely the urban block to then test the limits of this prediction model.

See the full project in the IAAC Blog at this link

Renderman
MaCAD Students: Felipe Romeo, Keshava Narayan Karthikeyan, Nawapan Suntorachai
MaCAD Senior Faculty: Stanislas Chaillou  w/ Oana Taut
Course: Digital Tools for Generative Building Design

The rapid development of urban centers and the evolution of architecture, machine learning, artificial intelligence and generative design have brought the opportunity to rethink traditional design workflows and provide a broader design vision. This case study intends to use machine learning to generate a tool that can help reduce time visualizing new designs; the intention is to integrate machine learning for conceptual rendering and open the opportunity to further investigate in this field. Due to the creation of new digital technologies and the relation of computation within architecture and design, designers face new opportunities to investigate and incorporate new tools that inform better design solutions. Visualizations and rendering processes have become a significant component of the overall design workflow for existing and new projects; however, this task is considered one of the most time-consuming within the overall workflow. The evolution of GAN’’s and the use of ANN within the architectural paradigm opens the opportunity to explore and recreate existing rendering datasets created for multiple projects allowing the creation of a new tool able to predict and generate conceptual imagery for early design phases. For now, design goals have been identified and created by architects/designers; however, artificial intelligence and machine-learning algorithms can improve the visualization of multiple unseen options to optimize and inform initial visualizations. This case study evolves with the initial idea of using different segmentation processes for real-time visualizations to generate an initial data set used to train the machine and provide initial testing within this field. The segmentation has been done using two different methods: 1- Segmentation using K-Means clustering within Grasshopper (as per 1st iteration below) and 2- creating a manual dataset to improve accuracy for the predictions (as per 2nd and 3rd iteration). The initial dataset used to train the machine consists of 300 images for each segmentation process (total 600 images); as expected, results vary based on the architecture of the ML model and the input data. The overall process has been done using Google Collaboratory as the computational platform and a Pix2Pix notebook; this notebook demonstrates image to image translation using conditional GAN’s (1). This technique allows the option to colorize and convert the segmented images into a predicted image based on colors and simple geometry.

See the full project in the IAAC Blog at this link

Blaze – Growing Light unit
MaCAD Students: Hesham Shawqy and German Otto Bodenbender
MaCAD Senior Faculty: David Andres Leon w/ Dai Kandeel.
Course: Digital Tools for Algorithmic Geometrical Optimization

Blaze is a ceiling lighting unit generated using L systems and branching processes. In this project, students explored different techniques for topological optimization and digital fabrication using machine learning. They used a set of tools that include the NetworkX Python library, Visual studio, K_means clustering, and Hops.
First, they generated simple quad mesh faces using Grasshopper, and then they used Kangaroo for mesh relaxation to keep a clean mesh typology with even valence vertices.
They then used Anemone to generate L systems geometry where they could have control over the geometry of the lighting unit, how many lighting branches, and branches length. That allowed them to have different iterations of the same lighting unit for different uses.
In order to optimize the geometry for digital fabrication, they explored different methods using NetworkX python library and K_means clustering. These methods were written in visual studio as a python code, and they used Hops component to bring this code inside Grasshopper3D to strip the mesh.

 

See the full project in the IAAC Blog at this link

Applications for 2021-22 edition are open!

Want to know more? You can schedule a video or text chat with the MaCAD Coordinator or Apply using the buttons below.