The CAVES and PANGAEA team specialises in training programmes that equip astronauts and mission developers with scientific, expeditionary and behavioural skills. The group’s primary output is focused on two training programmes, CAVES, a course that uses natural cave systems for expeditionary and human behavioural and performance training, and PANGAEA, a course for geological and astrobiological field training. Complementary to their training goals, these programmes are used as research and development platforms to advance several of ESA’s technological, scientific and operational areas.

One such tool developed for PANGAEA and future missions is the PANGAEA Analytical Toolset, consisting of the Mineralogical Database (MinDB), a curated collection of information on all the minerals found on the Moon, Mars and in meteorites, and the Machine Learning (ML) software that combines deep learning multi-class and multi-label classification algorithms together with data fusion from multi-method spectroscopy, allowing a drastic increase in the accuracy of automatic mineral and rock recognition.

Further information on PANGAEA and CAVES programmes can be found on the following websites:

ESA - What is PANGAEA?

ESA - What is CAVES?

Candidates interested are encouraged to visit the ESA website: http://www.esa.int

You are sought to contribute to the ongoing development of Machine Learning algorithms for recognition of planetary materials from multispectral datasets. This project combines deep learning multi-class and multi-label classification algorithms together with data fusion from multi-method spectroscopy, allowing real-time mineral and rock recognition.

For detailed information on this internship position, please click here: ESA - Space training team – Planetary Mineral Database Development and Validation of Spectra Classification Methods.

Result Orientation
Operational Efficiency
Fostering Cooperation
Relationship Management
Continuous Improvement
Forward Thinking

For more information, please refer to ESA Core Behavioural Competencies guidebook

You must be a university student, preferably studying at master’s level. In addition, you must be able to prove that you will be enrolled at your University for the entire duration of the internship.

The working languages of the Agency are English and French. A good knowledge of one of these is required. Knowledge of another ESA Member State language is an asset.

During the interview, your motivation for applying to this role will be explored.

You should possess practical experience with Machine Learning-based classification and regression methods, including Convolutional Neural Network (CNNs) and/or Kolmogorov-Arnold Networks (KANs).
Strongly desirable: familiarity with advanced ML methodologies such as multi-modal data fusion, signal decomposition/unmixing, incremental/few-shot learning, and conformal prediction. Academic or professional proficiency in Python and its core scientific/ML frameworks: TensorFlow, JAX, Keras, and Scikit-learn and NumPy. Hands-on experience processing, analyzing, and visualizing scientific data obtained from analytical instrumentation, and working with datasets and databases are also valuable. Familiarity with MLOps software concepts (model/dataset integration, pipeline automation). A valuable asset: basic understanding of physical/chemical data metrics (e.g., spectral analysis, stoichiometric consistency) or interpretable and explainable (physics informed) ML concepts and tools.