Asteria – A Spaceflight Artificial Intelligence Assistant

As humanity ventures further into deep space, works toward a sustainable human presence on the Moon and Mars, and expands its commercial endeavors with concurrent missions, effective astronaut support becomes increasingly vital and more complex. This paper introduces Asteria, a next-generation spaceflight assistant designed to enhance astronaut performance and situational awareness during missions. Traditional ground-based support systems rely heavily on audio communication from Mission Control Centers (MCC), which will become increasingly unreliable due to latency and blackout windows. Moreover, this mode of support is inefficient, often leaving astronauts idle while awaiting critical information. Existing methods also foster siloed knowledge and require extensive training, leaving astronauts susceptible to cognitive overload and diminished situational awareness. Asteria addresses these challenges by employing a hybrid artificial intelligence (AI) architecture designed to operate on low-power, low-compute platforms, such as next-generation spacesuits, to deliver real-time assistance independent of MCC availability. Its deployment promises to revolutionize astronaut operations by reducing reliance on ground support and ensuring continuous access to mission-critical knowledge. Key benefits include heightened situational awareness, improved decision-making, and the ability to perform autonomous, Earth-independent operations. This paper outlines the current domain challenges, introduces Asteria and its core capabilities, describes the underlying AI architecture, and details its integration with existing systems such as the Collins Aerospace, an RTX business, Leto™ Spaceflight Intelligence System. Finally, we discuss current limitations and outline the next steps for future development and deployment.