ANCORELOG Analytical Core Logging System & T-REX (timegated RAMAN sensor)


ANCORELOG project consists in the development of an analytical core logging system. The project is funded by the EIT Raw Materials program and coordinated by DMT group.

Mining exploration and extraction involve a huge amount of rock sample analysis, especially drill cores samples. An important amount of time and money is associated with the acquisition and characterization of those samples as they are critical in the decision-making process of both long-term exploitation plans and daily production optimization.

The characterization of those samples suffers from several limitations that today technology has the potential to tackle:

  • Manual logging is time-consuming and prone to subjectivity (inconsistencies)
  • Chemical assays on samples are costly and greatly increase the turnaround time before the availability of results for interpretation
  • Demand for information is increasing, and identifying several properties of interest involve more and more specialization (e.g., geological, geotechnical and geo-metallurgical domains)


ANCORELOG analytical core logging system is meant to address this challenging characterization task by using the combination of complementary innovative sensors:

  • 3D range imaging camera for textural analysis and localization of core pieces
  • Hyperspectral SWIR camera for lithological / mineralogical characterization
  • XRF probe / LIBS scanning system for semi-quantitative chemical analysis
  • Timegated RAMAN probe for mineralogical characterization (T-REX project)
Ancorelog prototype

GeMMe group (ULiège) is responsible of the data fusion task, and the integration of artificial intelligence tools leading to the interpreted results provided to the end-user. GeMMe is also responsible of the implementation of the hyperspectral SWIR camera in the prototype.

A complete software framework has been developed during this research project in order to display the multi-sensor data, and to assign labels to a fraction of measured samples that will be used to train supervised classification models. Several case-studies have been studies in the context of the project.

During the project, a fully innovative approach has been defined, where large neural networks are used to combine image-type data with punctual data and, finally, output automated logs for the various properties of interest that will serve the end-user in his further analysis (lithology class, alteration minerals, etc.). Besides, chemical analysis is also provided along depth for all the scanned samples.


RAMAN sensor is currently being integrated in the context of T-REX project and its added value is being evaluted on active case-studies.