The Automata project: 3D models enriched with archaeometric data

Archaeological excavations today produce considerable quantities of remains. Among them, ceramics and stone artifacts occupy a central place: shards, flint chips, and fragments of objects accumulate by the thousands, even by the tons. These items, often modest when taken individually, nevertheless constitute an essential source of information for understanding the techniques, exchanges, and lifestyles of past societies.

But this abundance poses a very real problem. Museum and conservation center storage facilities are becoming increasingly cluttered with these collections from excavations. At the same time, specialists must examine these sometimes enormous collections in order to extract as much scientific information as possible. The challenge is therefore twofold: to preserve these masses of artifacts for the long term and to analyze them effectively. 

It is in this context that the European Automata project offers a new approach to facilitate the work of archaeologists: quickly acquiring as much information as possible using advanced technologies, then consolidating all this data into comprehensive digital models available on the European Collaborative Cloud for Cultural Heritage, the “European cloud.” 

Coordinated by the University of Pisa, Automata brings together around forty scientists from a wide range of disciplines: archaeology, archaeometry, scientific imaging, experimental physics, analytical chemistry, robotics, computer vision, as well as industrial partners from various sectors. Italian, French, British, Spanish, and Israeli institutions are working closely together, including Université Bordeaux Montaigne and INRIA, Inrap France, the University of Jerusalem, and King's College London, for example.

This diversity of skills is at the heart of the project: it involves combining methods that are rarely integrated into the same analysis chain to produce an enriched digital representation of analytical data for each artifact. Automata aims to rapidly create true digital twins of archaeological objects, capable of revealing both their shape and their physical and chemical composition.

The study begins with 3D scanning of archaeological samples, whether lithic fragments or ceramics. This reconstruction provides a precise geometric model, which serves as a reference for all subsequent measurements.

Hyperspectral imaging is then performed to observe objects beyond the visible spectrum and identify variations in composition. This makes it possible to identify the most relevant areas for further study.

Targeted analyses, such as Raman spectroscopy or X-ray fluorescence spectrometry, are then carried out on these points of interest to determine the atoms present and characterize the molecular composition of the materials.

All of the acquired data is ultimately integrated into the 3D model. Each area thus combines information on shape and physicochemical properties, constituting a complete digital representation of the artifact.

In this project, the Manao project team is working closely with Archéosciences  Bordeaux, a joint research unit between Bordeaux Montaigne University and the CNRS. The goal of the scientists in this Inria team is to contribute expertise acquired through the development of La Coupole. This 3D imaging technology simultaneously records an object's geometry and reflectance—i.e., how its materials reflect light—to create digital models that are extremely faithful to reality. 

"With La Coupole, we at Manao have expertise in a system that combines robotics and optics. Even though we are not looking for the high-definition performance we have with La Coupole for the Automata project, it remains an almost ideal experimental setting. We can test different parameters to obtain accurate reconstructions that can be used by other project partners," explains Clément Joubert, a postdoctoral researcher in the Manao project team. 

To validate these methods, the Manao project team has already begun working on around twenty archaeological samples, mainly ceramic fragments measuring between 1 and 12 centimeters, sent by the University of Pisa. Each piece is meticulously photographed by La Coupole and then reconstructed in 3D. “This project presents us with a real scientific and technological challenge,” continues the scientist. "At the moment, La Coupole does not yet have a method that allows us to acquire the artifact in a single scan: we have to digitize the top and bottom separately, then merge the data without losing accuracy. Our experiments are enabling us to gradually refine the protocols and identify the best strategies for automating the acquisition process while maintaining the quality of the reconstructions."

For the next phase of the project, Italian partners specializing in robotics will take over. The goal is to design a device that is much more compact than La Coupole, simpler and less expensive to deploy, while directly integrating analytical instruments such as spectrometers into the robotic system itself. This development will make it possible to automate the entire acquisition chain, from 3D scanning to physicochemical characterization, in a single technical environment.

“The challenge now is to miniaturize and integrate these different technological components into a single platform,” emphasizes Rémy Chapoulie, researcher at the Archéosciences Bordeaux laboratory and head of the Archeovision 3D platform. We want to move from a high-performance but complex experimental device to a lighter, more mobile, and more accessible system capable of combining 3D imaging and archaeometric analysis.  This integration will enable us to take a decisive step forward in the conservation and analysis of these archaeological artifacts, benefiting the entire archaeological science community and, more generally, cultural heritage via the European cloud. "