Data-Centric AI in Multi Domain Awareness

In the realm of space operations, effective data strategies are essential for enhancing our understanding of space activities and ensuring spacecraft safety within geosynchronous orbit (GEO) and beyond. These strategies involve data fusion, ML algorithms for anomaly detection, and big data analytics to process and extract insights from the vast volumes of data generated during space operations. By integrating diverse data sources, we gain a comprehensive view of space activities, enabling real-time monitoring and immediate alerts for abnormal events. Spacecraft anomalies are identified through various data types, including EO/IR imagery, astrometric and photometric attributes, and RF payload data. This project aims to construct an ontology for automated space-related alerts, classify and identify alerts using data-driven methods, and develop mathematical solutions to interpret space data effectively. The project also addresses the consequences of insufficient training in AI systems, ensuring robust training and preventing data cascades. Ultimately, this project enhances space situational awareness for government and commercial applications by automating the analysis of Space Domain Awareness (SDA) data, leading to valuable insights into spacecraft operations.