published : 2023-11-16
Do you read me, HAL? Space agencies weigh pairing astronauts in deep space with AI companions
NASA is researching whether an 'AI social support tool' for astronauts could help on long journeys
Space agencies around the world are developing AI companions to help astronauts stave off loneliness, combat space-induced mental illness and assist with work on multi-year trips.
Alexandra Whitmire, a scientist with NASA's Human Factors and Behavioral Performance team, explained that deep space travel presents unique challenges for astronauts, different from those experienced in orbit.
Deep space missions, such as the journey to Mars, can last around 2.5 years, and the small size of the vehicle means that the crew will be confined to a small habitat.
To address these challenges, an AI social support tool is being researched as part of a toolkit of countermeasures for future missions.
Both NASA and the European Space Agency (ESA) are exploring the use of AI companions to support astronauts' mental health and assist with their workflows.
ESA introduced the Crew Interactive Mobile Companion (CIMON), a volleyball-like computer, that could aid astronauts with experiments on the International Space Station (ISS).
Later iterations of CIMON aimed to connect emotionally with the crew, answer questions, and record interactions.
Real-world AI systems, however, require further research to understand the extent and methods of support, potential pitfalls, and their effectiveness as behavioral health countermeasures.
While AI companions may offer a safe sounding board for some astronauts, the ability to connect with family and maintain team cohesion are also considered crucial methods to support their behavioral health.
Other space agencies, such as JAXA, the French Space Agency, the UK Space Agency, and the Italian Space Agency, have also funded AI projects.
Nevertheless, the focus remains primarily on AI tools that can assist and augment the mental health support for astronauts, with a human-centric and human-driven approach.