Copyright : BertrandSource : Thales

Challenge

AI x Future Combat Air System

Pioneering the future of combat systems with next-gen AI solutions

ApplyAppliquer
This challenge is over, applications are closed.

Challenge Overview

Description du challenge

The Future Combat Air System (FCAS) is a collaborative effort between Dassault Aviation, Airbus, and Indra Sistemas to develop a combat system of systems that includes a Next-Generation Weapon System (NGWS) comprising remote carrier vehicles and a New Generation Fighter (NGF). The NGWS requires optimization for sensor suite usage, learning with few examples or guided learning, massive synthetic image production, incremental database enrichment with novelty integration, change analysis, and abnormal behavior detection.

Copyright : ulyssejdv
Copyright : ulyssejdv
Copyright : Jarek
Copyright : Jarek

Wanted

Expertise in

  • Multi-constraint optimization for dynamic use of heterogeneous resources
  • Learning with few examples or guided learning
  • Synthetic image production
  • Incremental database enrichment
  • Change analysis
  • Abnormal behavior detection

Knowledge of

  • Future Combat Air System (FCAS)
  • Next-Generation Weapon System (NGWS)
  • New Generation Fighter (NGF)
  • Aerospace technology
  • Military systems
  • Machine learning and artificial intelligence
  • Data analysis and processing
  • Image processing and computer vision

Projects you could be working on

  • Multi-constraint optimization for dynamic use of heterogeneous resources

On the basis of several use cases describing:

  1. a theater with actors capable of reacting and interacting
  2. drivable resources, with variable capacities and their own constraints
  3. a "mission" objective aiming at multi-criteria performances (accuracy, delay, coverage...) in the respect of implementation rules (constraints)

The solution should propose and evaluate planning and optimization techniques under constraints. It should be CPU efficient, improve the success rate of the mission, be modular and scalable, and if possible explainable, and/or capable of proposing alternative plans, with possible human intervention.

  • Creating solutions for learning with few examples or guided learning

  • Synthetic image production

The solution concerns the SAR/OPTRO cooperation in aerial to ground imagery, to realign the images, and improve the detection of objects of interest.
To do this, it is necessary to massively simulate optronic and SAR images corresponding to the same scenes and with the same objects in a large variability of contexts, so as to be able to train machine learning algorithms.

  • Developing solutions for incremental database enrichment with novelty integration
  • Creating algorithms for change analysis
  • Developing solutions for abnormal behavior detection