Autonomous interactive agents for global telescope network orchestration

The expansion of scientific, economic, and military activity in space has driven concomitant growth in the diversity, dynamism, and size of the anthropogic space object population. The safety of spacecraft operations depends on the detection and monitoring of transient events among this population, such as satellite maneuvers, proximity operations, and component articulation, as well as the routine maintenance of high-accuracy ephemerides. Optical sensors are effective tools for this task when applied in large numbers. However, responsiveness to transient events necessitates geographic diversity, and cost-effectiveness motivates heterogeneity of aperture and instrumentation. Orchestrating a globally distributed, diverse collection of sensors to satisfy space-object tracking objectives remains an open challenge. In this work, we adopt the open-source MACHINA agentic software framework to embody an autonomous space domain awareness agent that addresses this challenge.