Advances in automation and artificial intelligence (AI) in recent years hold promise for support of long duration crewed exploration missions to Mars and beyond. Tools developed from these advances could assist in the maintenance of critical systems such as life support systems. Current AI tools are able to synthesise large amounts of text data and can interact with a user, based on the trained data, which could serve in supporting, diagnosing and providing relevant actions to help in maintaining the system. The use of AI tools could therefore be a way of supporting crews with such tasks and could be valuable in terms of saved crew time and increase productivity.

The “Staying Alive!” experiment aimed to investigate the role of AI in crew performance and their perception, applied to critical systems in the context of the AMADEE-24 mission. The experiment sought to understand how human crews work with the help of AI support to maintain complex systems in crewed space exploration. The experiment featured a simulated photobioreactor over which the crew performed various tasks. A graphical user interface guided them through the various maintenance or troubleshooting tasks with the help of procedure manuals. Four different support scenarios were provided to the analog astronauts, i.e. no support, interaction with an AI chatbot, communication with the Mission Support Center (MSC) with a 10-minutes communication delay, or both AI and MSC support, in combination.

Upon the completion of each experiment run, the analog astronauts were surveyed about their experience in performing the task in the above-mentioned scenarios, in order to assess their performance as well as their perception of the different support types. The presentation will discuss the experiment as featured during the mission, the data collection and survey methodology, findings of the experiment, alongside general lessons learnt for the experiment and the next steps for the Staying Alive experiment. Special emphasis will be set on the results of the performance comparison between the different support scenarios as well as on the outcomes of the surveys on the perceptions of the crew performing the experiment.

Source: Sahil Bhatia¹, Katrin Schirm¹, Malinda Mecit¹, Prof. Vera Hagemann¹, Dr. Christiane Heinicke¹, “Staying Alive: Investigating The Role of Artificial Intelligence in Supporting Crewed Operations During AMADEE – 24 Analog Mission”, AMADEE-24 Science Workshop Booklet 2024

¹ University of Bremen, Germany