Building a Standalone AI Assistant for Deep Space Exploration
This paper presents the initial experiences from the design and implementation of an AI visual and voice bot for astronaut assistance in the upcoming Indian Human Space missions. The speech pipeline of the voice bot has speech recognition, text to speech conversion, and an on-device Large Language Model (LLM) based chatbot module. This bot is equipped with a display to render an animated face and play media contents. The voice assistant is specifically designed for interaction in Indic-English and Hindi languages. The voice bot along with its display is packaged to occupy the volume of a football. The high-performance processor used is up-screened for use in space, and is size weight power and cost (swap-c) efficient compared to conventional solutions. The capability to function offline without cloud connectivity and ability to interact verbally and visually with the user makes this AI assistant unique in comparison to its contemporaries.
Authors
- Lima Agnel Tony (ORCID: https://orcid.org/0000-0002-5173-9089)
- R. Durairaj (ORCID: https://orcid.org/0000-0002-5698-2651)
- Joji J. Chaman
- A. Amritha
- E S Padmakumar
- Harikumar Ganesan (ORCID: https://orcid.org/0000-0001-7144-0562)
- G. Sangeetha (ORCID: https://orcid.org/0009-0001-6259-5172)
Publication Details
- Journal
- Space Education & Strategic Applications
- Published
- 2025-04-01
- DOI
- https://doi.org/10.18278/sesa.5.2.7
- Primary Topic
- Distributed and Parallel Computing Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00