Optimizing Deep-Space Operations: The Impact of LuzIA on Crew Communication and Productivity
- Paper ID
93164
- author
- company
Instituto Tecnológico Autónomo de México
- country
Mexico
- year
2025
- abstract
Long-duration space missions present unique challenges, particularly in maintaining effective communication between astronauts and mission control, especially when signal latency and limited communication windows become significant obstacles. This paper introduces LuzIA, a multi-layer decoder-only transformer-based large language model designed to serve as an advanced virtual assistant that enhances asynchronous communication and workflow management in such circumstances. LuzIA is inspired by sci-fi visions from the past that envisioned intelligent, responsive systems capable of assisting humans in the exploration of space. Acting as an intermediary, LuzIA efficiently receives, analyzes, and prioritizes messages from ground control while offering automated responses for routine issues, minimizing the need for direct crew interaction. Furthermore, LuzIA is programmed to manage a pre-scheduled plan established by the astronauts, ensuring timely execution of tasks and experiments while adapting to evolving mission demands. By monitoring and revising schedules, LuzIA alleviates cognitive load on the crew, enabling smoother daily operations. The system also facilitates real-time data recording and analysis, supporting informed decision-making for both the crew and mission control through adaptive learning algorithms that enhance its processing capabilities over the course of the mission. Preliminary findings indicate that LuzIA can significantly reduce response times and increase operational efficiency, thereby contributing to the safety and success of long-duration space missions. This paper discusses the implementation, functionality, and potential benefits of LuzIA in supporting human crews engaged in deep-space explorations, bringing to fruition the imaginative ideas of past science fiction into practical application.