DASSAULT AVIATION ET L’ISAE-SUPAERO RENOUVELLENT LEUR COLLABORATION SUR LA CHAIRE DE RECHERCHE « CONCEPTION ET ARCHITECTURE DE SYSTÈMES AÉRIENS COGNITIFS »
DASSAULT AVIATION AND ISAE-SUPAERO RENEW THEIR COLLABORATION ON THE "COGNITIVE AIR SYSTEMS DESIGN AND ARCHITECTURE" RESEARCH CHAIR
Initiated in 2016, the "CASAC" research and training chair, signed by Dassault Aviation, ISAE-SUPAERO and its Foundation, aims to rethink the relationship between crews and the systems used in aviation. Following promising initial results, the two aeronautics players are renewing their partnership for a further three years.
Initiated in 2016, the "CASAC" research and training chair, signed by Dassault Aviation, ISAE-SUPAERO and its Foundation, aims to rethink the relationship between crews and the systems used in aviation. Following promising initial results, the two aeronautics players have renewed their partnership for a further three years.
WORK TO OPTIMISE MAN-MACHINE INTERACTION
This chair, whose main research areas are neuroergonomics, decision-making autonomy in automated systems and systems engineering, aims to study various aspects of collaboration between man and machine. The challenge is to make civil and military air operations safer, more robust and more efficient, while guaranteeing complete control to crews. The systems under consideration are very often operated in complex situations; they therefore have advanced automatisms to carry out their missions in a more autonomous manner, still under human control, based on decision-making algorithms from the field of artificial intelligence.
At ISAE-SUPAERO, the Aerospace Vehicle Design and Control Department (DCAS) has expertise in neuroergonomics and artificial intelligence for system control. Neuroergonomics is the discipline that analyses the functioning of the brain and the work behaviour of users through the prism of neuroscience. In concrete terms, it involves evaluating the mental states of the user with regard to his or her ability to perform the tasks assigned to him or her. Artificial intelligence for system control is the discipline that develops automated decision-making algorithms. The decisions concern, for example, the set of tasks to be proposed to the operator or to be carried out automatically in order to reduce the operator's workload and improve his performance in relation to the operational context.
"Dassault Aviation is particularly concerned by the problems of human-machine interaction because military aviation is very demanding due to the diversity and unpredictability of missions, which require complex tactical management. The challenge is to provide the human being with all the services that will enable him or her to assume responsibility for this management. This is why we are working with ISAE-SUPAERO to identify the phenomena that will affect the performance of collaboration between crews and their machines," says Jean-Louis Gueneau, coordinator of the scientific aspects of the Chair at Dassault Aviation.
INITIAL CONCLUSIVE RESULTS
The Chair's research work on Human-Machine Interaction, carried out from 2016 to 2021, has enabled the development of various physiological measurement tools, as well as machine learning and automated action planning techniques. In particular, the teams worked on the development of active or passive assistance functions to help pilots and operators improve their performance.
For this purpose, 'pilot monitoring' was the initial focus of work in order to better understand the activity of the crew. Experiments on simulators using behavioural and physiological measurement tools have been carried out to determine metrics capable of evaluating operator performance, and their level of commitment or stress.
Dassault Aviation plans to integrate such functions into its civil and military aircraft in the next decade.
THE MACHINE, THE OPERATORS' TEAMMATE
"The main thrust of this chair is the development of innovative technologies that help, on the one hand, to qualify the interaction between the human and the machine to determine whether cooperation is effective and, on the other hand, to automatically decide what should be maintained, suggested or changed to promote team performance," explains Caroline Chanel, head of the chair at ISAE-SUPAERO.
To do this, quantitative behavioural and physiological metrics will be merged with more qualitative metrics in order to assess the effectiveness of human-machine cooperation. This efficiency measurement will then be exploited by algorithms from the field of artificial intelligence to adapt and reinforce this cooperation.