Published research

Artificial intelligence for cultural heritage

Two published studies show the data used, the role of AI and how results are evaluated.

Research case studies

AI within the project

2025

Problem: Enable contextual questions while exploring historic architecture and gardens in VR.
Data: Laser surveys, drone photogrammetry, 3D models and building-related texts.
Method and role of AI: A language model answers questions entered through a virtual keyboard within VR.
Team contribution: Design and integration of the workflow from survey to immersive interaction.
Output: A VR application with explorable buildings and natural-language interaction.
Evaluation and limitations: User tests examine realism, interaction and accessibility, with interaction issues identified for improvement.

2026

Problem: Translate Scamozzi’s architectural rules into reproducible parametric models.
Data: Descriptions, proportions and geometry from L’idea dell’architettura universale.
Method and role of AI: Language models assist Python code generation for Blender.
Team contribution: Formalisation of geometric rules and development of a Blender add-on.
Output: Parametric architectural components and inspectable, reproducible procedures.
Evaluation and limitations: Human checks of code, geometry and proportions; the study documents generation errors and limitations.

Reconstructions, digital museums and information models in the laboratory archive.

Real-time exploration of a virtual anastylosis model and interaction with 3D artefacts.

Interactive learning paths and virtual reality for the National Archaeological Museum of the Marche.

A digital museum connecting archaeological sites and evidence along the Roman road.

From laser scanning to HBIM modelling and parametric libraries.

Research and collaboration

Museums, institutions, businesses and students: find the right collaboration route on our contact page.