Abstract
Oltre gli strumenti: integrazione AI-driven multisensore per l'archeometria
La diffusione di strumenti portatili e non invasivi (e.g. HSI, pXRF, Raman) ha trasformato lo studio dei materiali archeologici, ma ha anche prodotto grandi quantità di dati eterogenei, difficili da integrare e riutilizzare. Questo contributo presenta un workflow, sviluppato nell’ambito del progetto AUTOMATA, volto a supportare l’integrazione multisensore, la validazione automatizzata e la gestione dei dati. I dati provengono da un insieme di manufatti ceramici analizzati mediante tecniche complementari. Essi vengono armonizzati e arricchiti tramite procedure supportate dall’intelligenza artificiale, finalizzate al controllo della qualità delle acquisizioni e alla selezione dei metodi di analisi più idonei. Il sistema opera secondo un’architettura human-in-the-loop e si integra con una struttura dati progettata secondo i principi FAIR. I risultati preliminari mostrano un miglioramento in termini di qualità, tracciabilità e comparabilità dei dati, sottolineando il ruolo delle infrastrutture digitali nello sviluppo di pratiche archeometriche più riproducibili e sostenibili.
Presentazioni
- DOCUMENTO
Video
Bio
An archaeologist specialising in the application of archaeometric techniques to the study of archaeological materials and ancient production processes. After completing a degree in Archaeology at the University of Padua, she obtained a joint Master’s degree in Archaeometry in collaboration with Université Bordeaux Montaigne. She has taken part in several archaeological excavations and carried out laboratory work using a range of analytical techniques (optical and confocal microscopy, SEM-EDS, Raman, pXRF, HSI). She is currently a PhD candidate at the University of Pisa, where she develops integrated approaches combining archaeometry and digitalisation, with a focus on the integration and interoperability of multisensor data to enhance the analysis of archaeological materials, particularly ceramics.
Abstract
Beyond instruments: ai-driven multisensor integration for archaeometry
The widespread adoption of portable and non-invasive instruments (e.g. HSI, pXRF, Raman) has transformed the study of archaeological materials, but has also generated large volumes of heterogeneous data that are difficult to integrate and reuse. This contribution presents a workflow, developed within the AUTOMATA project, designed to support multisensor integration, automated validation, and data management. The data derive from a set of ceramic artefacts analysed using complementary techniques. They are harmonised and enriched through procedures supported by artificial intelligence, aimed at quality control of data acquisition and the selection of the most appropriate analytical methods. The system operates according to a human-in-the-loop architecture and integrates with a data structure designed in line with FAIR principles. Preliminary results show improvements in data quality, traceability, and comparability, highlighting the role of digital infrastructures in advancing more reproducible and sustainable archaeometric practices.
Attribuzione 4.0 Internazionale