Assurance Guidance for Space Mission use of Data-Driven Machine Learning

Martin S. Feather, Steven Guerrini, Philip C. Slingerland, Max Spolaor · 2023

We report on our experience developing guidance for assurance personnel on how to assure data-driven Machine Learning (ML) systems to be used in the space domain. The impetus for this work stems from the rapid advancement of ML techniques, the positive results of which are inspiring even relatively risk-averse space missions to seek their use. However, assurance personnel, charged with assessing the adequacy of processes and practices followed during space mission developments, are often unfamiliar with these techniques, and lack matured standards to guide their assessments.

Read the paper · More papers on PaperTik