CRAFTED - Constructing Restoration Archives with Future Technologies for European Documentation
Durata:
Responsabile scientifico:
Tipo di progetto:
Ente finanziatore:
Codice identificativo progetto:
Ruolo PoliTo:
Abstract
Historic buildings throughout Europe preserve centuries of architectural, material, and artisanal knowledge. However, the technical information generated during their restoration—encompassing construction techniques, material provenance, craft practices, and specialized terminology—is generally dispersed, inconsistently recorded, and fragmented across linguistic and disciplinary boundaries. This fragmentation significantly hinders comparative research, sustainable conservation strategies, and broader public understanding of Europe’s built heritage. CRAFTED (Constructing Restoration Archives with Future Technologies for European Documentation) tackles this challenge by examining how generative artificial intelligence (AI) can systematically reconstruct, structure, and contextualize restoration knowledge derived from historical building sites across Europe. The project advances the hypothesis that integrating cutting-edge generative AI with domain-specific expertise in architectural history and conservation facilitates the extraction and modeling of structured, multilingual technical knowledge from diverse and often non-standardized historical sources.
CRAFTED proposes prototyping an AI-enhanced digital infrastructure capable of processing, analyzing, and interrelating data related to construction techniques, materials, the geographical origins of components, historical and contemporary site actors (e.g., artisans, architects, engineers), and terminological practices. As an initial pilot, the project focuses on creating a dynamic glossary of historical construction techniques and technical terms used in Early Modern restoration projects. This glossary is compiled in English to ensure semantic coherence and international usability but is made accessible in multiple European languages to reflect regional variations and support inclusivity. This multilingual accessibility is central to the project’s goal of integrating diverse national restoration histories within a shared European knowledge framework. The glossary serves not only as a linguistic reference but also as a semantic core for a broader knowledge graph that maps interrelations among terms, materials, processes, and historical contexts. Artificial Intelligence, including Generative AI, supports terminology extraction and the inference of semantic relationships, enabling the glossary to scale over time and adapt to new sources, linguistic inputs, and user expertise. CRAFTED takes a consciously interdisciplinary approach, using methods from architectural history, heritage studies, digital humanities, computational linguistics, and AI. The project seeks to generate both theoretical insights and technological tools: it proposes a new digital epistemology for architectural and restoration knowledge, broadens the applicability of generative AI to historical and domain-specific language contexts, and introduces an innovative methodological framework for semantic modeling of heritage data. The project is inherently high-risk due to the variability, scarcity, and multilingual aspects of its source materials, along with the challenge of adapting generative AI to low-resource, semantically unstable domains. Nevertheless, its high-gain potential lies in its ability to create a transformative infrastructure for restoration knowledge—one that informs conservation practice, advances digital heritage research, and broadens public access to historical construction cultures. CRAFTED rethinks how it collects, interprets, and uses digitized restoration data. By creating an intelligent, multilingual archive rooted in historical expertise and computational innovation, the project establishes a new foundation for interdisciplinary research and sustainable cultural heritage preservation in Europe aligning with current research trajectories that emphasize the strategic role of digital tools, historical knowledge integration, and adaptive reuse in building resilient, sustainable frameworks for cultural heritage preservation.
Persone coinvolte
- Valentina Burgassi (Responsabile Scientifico)
- Alberto Cannavo' (Responsabile Scientifico di Struttura)
- Laura Sardone (Responsabile Scientifico di Struttura)
Strutture coinvolte
Parole chiave
Settori ERC
Obiettivi di Sviluppo Sostenibile (Sustainable Development Goals)
Budget
| Costo totale progetto: | € 90.000,00 |
|---|---|
| Contributo totale progetto: | € 90.000,00 |
| Costo totale PoliTo: | € 90.000,00 |
| Contributo PoliTo: | € 90.000,00 |