Complex natural materials (sediments & food products), or from industry (waste) need to be sorted for their analysis and recovery. CEREGE has developed with ATG Technologies, a SME specialized in robotics, a microparticle sorting automaton, capable of sorting microorganisms thanks to image recognition based on convolutional neural networks (patent pending, Marchant & al, 2020). This automaton, MicroFossil Sorter (MiSo), the first in the world, is routinely used at CEREGE to image and sort microfossils and waste residues. This automaton is the basis of a second version of the sorter including X-ray detectors [Equipex+ IMAGINE-2, financed]. Nevertheless, the increase in power of microparticle sorting also requires the development of new treatment processes in addition to these new sensors. The objective of the MicroTrEE project is to develop with ATG a new generation of intelligent sorting automatons that will (i) maintain AMU & ATG’s lead in the field by incorporating 3D, parallelizing the processing chains, incorporating ML algorithms to manage each chain and robotize the cleaning of the automaton, but also (ii) develop new applications in environment, paleoenvironments and mining exploration.

Funding : A*MIDEX Foundation

Partners : CEREGE & ATG Technologies

Dates : 2024-2026

PI : Thibault de Garidel-Thoron, CEREGE