Copyright: Ralf Maassen / DTEuropeSource : Thales


AI x Adaptive Training

This challenge is over, applications are closed

Challenge Overview

Description du challenge

Up to now, pilots, air controllers or any operator working in a complex environment are trained following a predefined curriculum. As a result, a Gaussian curve represents the usual outcome of a given class as no differentiation is made between trainees. Consequently, the overall performance of a training scheme must improve to achieve far better results. The purpose of the project is to move as many trainees as possible to the right of the Gaussian curve.The objectives of the challenge are to:

  • Assess the progression of a pilot or air controller trainee
  • Automate the adaptation of the course or curriculum to his or her learning pace and capability
  • Provide an adaptive learning management feature
Copyright: Thales
Copyright: Thales
Copyright: Sergey Nivens
Copyright: Sergey Nivens


Expert in:

  • Machine learning
  • AI graders (bayesian...)
  • Explainability
  • Automation
  • Data science

Experience in:

  • Educational domain
  • Learning management system or alike
  • If possible, in blended learning

It would be highly appreciated if you already have access to some educational databases and material. Any educational domain is acceptable (economics, legal, science, medical, maintenance, safety...).

Projects you could be working on

You will demonstrate automated AI grader and AI adaptive learning solution on our own datasets. The solution shall evaluate acquired skills in an automated way and propose cursus modifications or evolutions together with a performance target.

This will require:

  • The use of automated AI graders
  • A Learning Management System connected to a Learning Record Store or a Learning Experience Platform.

The proposed solution should provide a complete view of the progression and the proposed corrections of the curriculum to achieve a set of given goals. The solution shall be highly dynamic and adaptive and rely on AI.