João Matias
DOI Number: XXX-YYY-ZZZ
Conference number: HiSST 2024-00145
In the framework of the Horizon Europe project called “European iNitiative for Low cost, Innovative & Green High Thrust Engine”, new technologies are being developed to be applied on future launch vehicles. One focal technology explored is the incorporation of on-board Health Monitoring Systems (HMS) with the inclusion of artificial intelligence (AI) at a propulsion and launcher level for failure detection and identification (FDI) during its entire mission profile. This paper follows the ongoing trade study process performed for the preliminary design of such a rocket engine health monitoring system (EHMS), including the key design considerations and decisions made. Firstly, different aspects of the high-level architecture are discussed, resulting in establishing a distributed heterogenous system where each EHMS unit monitors a single rocket engine. A brief component trade-off analysis concludes that combinations of “commercial off-the-shelf” (COTS) components could be promising candidates for high performance edge AI inferencing within such an EHMS unit. Finally, the challenges associated with performing the study were raised, namely the difficulty in assessing the validity of COTS components for space applications and the lack of test datasets for evaluating space AI applications.