Autonomy Architecture for a Raven-class Telescope with Space Situational Awareness Applications

This paper investigates possible autonomy architecture designs of a Raven-class telescope as applied to the tracking and high level characterization problem in Space Situational Awareness (SSA). Various levels of autonomy are defined and existing systems and capabilities are discussed. Telescope interactions with distributed sensor networks such as the Space Surveillance Network (SSN) are reviewed, and several relationships between autonomy and scheduling of telescopes are addressed. An autonomy architecture design for a Raven-class telescope is presented and future extensions are proposed. 1 Research Motivation In 2001, the Rumsfeld Commission Report concluded that improvements in Space Situational Awareness (SSA) were needed to protect the US and its allies as well as maintain its economic and diplomatic objectives. Joint Publication 3-14, Space Operations, defines the high level activities included in SSA as the characterization and analysis of space objects (SOs) and environmental conditions interacting with space based assets. Space objects consist of active and inactive satellites, rocket bodies, and orbital debris. A key element of SSA is determining whether the orbits of SOs might bring them into close proximity, an event known as a “conjunction,” and the conditional probability of SO collision. In order to establish a robust SSA capability, obtaining regular measurements is key. The U.S. Stratetic Command (USSTRATCOM) Joint Space Operations Center (JSpOC) operates the Space Surveillance Network (SSN) and currently tracks 22,000 objects with diameters greater than 10 cm. Additionally, NASA Johnson Space Center’s (JSC) Orbital Debris Program Office (ODPO) has primary responsibility for SO population below the SSN detection limit. The need for persistent SSA has been exacerbated by both the Chinese anti-satellite test in 2007 and the Iridium/Cosmos collision in 2009, both of which have increased the low-earth orbit (LEO) SO population by more than 60%. The Space Surveillance Network (SSN) has historically been unable to collect enough raw measurements to fully characterize the SO population. Hence, scheduling limited sensor resources to collect observations of the large SO population is a complex scheduling and resource allocation problem. While many efforts are currently under way to augment the current SSN with additional sensors, a great number of additional sensors with improved capability will increase the complexity of this scheduling and planning problem. When the space fence radar becomes operational, the number of tracked objects is expected to exceed 100,000. In addition, modifying established schedules under dynamically evolving scenarios, inclement weather conditions, and hardware faults is difficult. The planning and scheduling is complicated by the fact that the current global network of SSN sensors are not exclusively controlled by the Air Force Space Command (AFSPC), but also by other entities that provide data to the command, such as foreign governments or other government agencies (OGA) such as the Missile Defense Agency. Not only has the number of SOs tracked by the SSN greatly increased, but the data products and services provided by JSpOC have a record number of customers as well. Currently more than 100 countries as well as ∗Graduate Research Assistant, School of Aerospace Engineering, Georgia Institute of Technology †Assistant Professor, School of Aerospace Engineering, Georgia Institute of Technology