The house of the future II

Generative artificial intelligence and expert database exploration for housing design. (DIN.IAG)

Project information

Project title: La casa del futuro II. Generative artificial intelligence and exploration of expert databases for residential design / House of the future II. Generative artificial intelligence and exploration of expert databases for residential design.

Code: INC-UDIT-2025- PRO15

PI Project: Phd. Eva Hurtado Torán (eva.hurtado@udit.es)

Eva Hurtado, researcher at UDIT and PI of the VerSus Group, explains in this video the purpose and objectives of the DIN.IAG Project.

In addition to Eva Hurtado, this project is being carried out by: Enrique Villamuelas - Predoctoral Researcher, Rosa Barrero - Predoctoral Researcher, Alberto Calderón - Predoctoral Researcher and Diana Albares - Research Assistant.

Objectives

Can architects and designers be involved in the creation of expert databases for Artificial Intelligence?

  1. Work with Generative Artificial Intelligence to test algorithms in expert spatial design.

  1. Study the protocol and structure of a database specialised in residential architecture for use by IAG.

  1. To study in depth how generative models interpret and represent architectural spaces.

  1. To test the automation of increasingly complex aspects of urban collective housing projects.

METHODOLOGY

Una moderna sala de estar con diseño minimalista y grandes ventanales que ofrecen vistas panorámicas.

Practical Applications

Elaboration of a reference corpus to compare IAG algorithms applied to Spatial Design. Establish a standard protocol of reference encodings for AI developers in the architectural field.

Una presentación en una clase con un expositor y una conferencista ante una pantalla con imágenes.

Innovation and competitiveness

This research will expand on previous work on algorithmic transparency, with specific emphasis on architectural applications.

We aim to participate in the registration and classification of IAG trainings based on expert and accessible data categories.