The lunar ambitions of NASA's Artemis program spell out multiple crewed missions to the Moon and later to Mars. The ESA Terrae Novae 2030+ exploration strategy roadmap aims to enable Europe's participation in the first crewed exploration mission to Mars. Yet, missions to deep space still pose significant challenges in terms of crew autonomy. As missions will become longer and more distant, the operations and sustainability of such structures in stressful isolation conditions constitute a high level technological and human challenge. The objective of Spaceship FR team at CNES (Centre National d'Études Spatiales - FRANCE) is to develop innovative technological bricks for the future Moon and Mars bases, thanks to technologies and skills of its network of partners: academia, laboratories and industries. AI4U is a system based on Artificial Intelligence made to interact with the astronauts with an intuitive, natural and multimodal interface and with the different systems in order to increase the autonomy level of operations. Its three main tasks are supporting the crew during the operations, doing simple tasks instead of the crew, being a companion to decrease stress and mental load. This paper aims to present how AI4U is interfaced with the different systems in order to increase the autonomy level of operation and the last developments. Our main focus is to offer the best user experience using its capacity to detect human emotions and cognitive load to simulate emotions and to adapt its prosody, improving the verbal and non-verbal communication skills including face and gestural recognition and using Large Language Model. On the other hand, we work on the operational concept and the hardware and software architecture to be as close as the final operational uses of AI4U. The initial results were tested during analog mission with students.
Navarro et al. (Sun,) studied this question.