This project introduces a graph-based machine learning system designed to automate furniture placement in apartment layouts. The workflow was developed as part of an architectural design studio and played a key role in generating comfort-driven interior configurations. At its core, the system encodes spatial and functional rules such as placing a sofa near a window … Read more
A STORY OF SYNTHETIC LIFE FORM HAVE YOU EVER DREAMT OF A BETTER VERSION OF YOURSELF? The Substance, directed by Coralie Fargeat (Universal Pictures, 2024) This project simulates a linear evolutionary process in which a synthetic organism progressively transforms to become more human-like. The evolution is guided entirely by language models, with no human intervention … Read more
“Spatial Cognition in Robotics: Optimizing Construction Workflow with AI”Shu Xiao – MRAC02 Thesis, IAAC 2024–2025 Despite advances in manufacturing automation, the construction industry continues to rely heavily on manual labor due to its unpredictable environment, safety concerns, and fragmented workflows. This project explores how vision-guided robotics and real-time AI agents can close that gap by … Read more
Abstract In this project, we explore how graph-based thinking can reshape construction planning by bridging design data and scheduling logic. Drawing inspiration from modular architecture and network theory, we investigate new ways to visualize, analyze, and optimize the sequencing of building elements. By combining insights from BIM, parametric modeling, and graph analysis, the work aims … Read more
1. Research Statement AIM OF RESEARCH: To create a biocomposite material primarily out of Cacau Seed Shell and utilizing PLA as its matrix and use it for robotic additive manufacturing process to produce interior design pieces, following evaluation parameters from circular economy 2. How to add the maximum amount of CSS to PLA Matrix 3. … Read more
Our Solution: The Co-creator App We’ve developed a Copilot App – a local AI system that lives in your building and speaks your language. Literally. You make natural language requests, and it translates them into real architectural actions by understanding the complex relationships between spaces and residents through graph databases and GDS algorithms. Here’s how … Read more
AQUA.NEST: An AI-Powered Decision Support Toolbox for Adaptive Urban Design Using Nature-Based Solutions in Flood-Exposed Coastal Cities Coastal urban flooding is emerging as one of the most urgent climate challenges of our time. Without strong adaptation, annual flood damages could reach €930 billion globally by 2050, even with improved infrastructure (OECD, 2020). Rising sea levels and … Read more
Thermal Comfort Index Prediction // Definition OBJECTIVE: predicting thermal comfort in air-conditioned residential buildings using machine learning algorithms. What is indoor thermal comfort? Thermal comfort is “That condition of mind that expresses satisfaction with the thermal environment” (ISO 7730) People may feel that their surroundings are warm, cold or simply comfortable depending on the thermal state … Read more
GWP-data enriched graphs Imagine being able to see the carbon footprint of a building, not after it’s built, but while it’s still a sketch. What if architects and engineers could get real-time sustainability feedback the moment they decide on a material or tweak a wall layout? That’s the vision behind Visual GWP. We combined Graph … Read more
A GRADIO tool that generates architectural line drawings from Rhino3D Viewport screenshots by relating on a trained LoRA of architectural line drawings that takes the viewport input prompt option setting and generate a ready to use line drawing The User even have the option to edit the Hue and Saturation of the line drawing and … Read more
Sketches in Space with AI and LoRA What if architects could doodle inside their designs—live, in 360°? ARchitect is a mixed-reality interface I developed for stylizing architectural spaces using expressive, cartoon-like overlays. Powered by LoRA fine-tuning and a custom-trained Flux, ARchitect reimagines how we interact with our built environment—by making architectural visualization playful, intuitive, and … Read more
Intro Designers know the pain: a perfect image—ruined by one wrong material. A distracting object. A missing texture. An awkward composition. You could retouch it manually, but how long will that take? And will it ever really blend? What if you could just circle the problem area and ask for something better? Something that fits … Read more
Why “Faster” Matters A single photorealistic render can quickly be very expensive and take days to produce, locking design changes behind hefty bills and long waits for architects. Generative AI, through models like Stable Diffusion and custom LoRAs, slashes both cost and time, turning sketches into near-final visuals in seconds. By democratizing high-fidelity imagery, it … Read more
Introduction In this blog, we walk through the complete data science process applied to a unique challenge: predicting the acoustic comfort index of apartment units based on architectural and environmental features. This includes data preprocessing, feature engineering, model training, interpretation, and tuning—with tools like XGBoost, Neural Networks, and SHAP for explainability. The workflow is split … Read more
Bridging Carbon estimations in early design stages The built environment accounts for nearly 40% of global carbon emissions, the urgency to integrate sustainability into architectural and engineering processes has never been greater. Yet, decisions made in the earliest stages of building design often lock in the majority of a structure’s lifetime carbon footprint, long before … Read more
1 — Project OverviewOur project is based on the building geometry model that is provided with three apartment types and spaces in between them. We chose to focus on the shared spaces: the courtyards, terraces, and balconies. The building is located in Barcelona, so we rely on fixed environmental conditions and don’t re-run simulations. We … Read more
