关键词: |
Data set, Digital data, Earth orbits, Engineering, Governments, Ground control stations, Ground stations, Low earth orbits, Simulators, Standards, Transportation, Artificial satellites, Artificial intelligence, Control systems, Machine learning, Software development, Engineers, Air force, Software design, Test and evaluation, Software engineering, Automation, Satellite, Ground station, Process mining, Transfer learning, Case study analysis |
摘要: |
US Air Force satellite ground stations require significant manpower to operate due to their fragmented legacy architectures. To improve operating efficiencies, the Air Force seeks to incorporate automation into routine satellite operations. Interaction with autonomous systems includes not only daily operations, but also the development, maintainability, and the extensibility of such systems. This thesis researches challenges to Air Force satellite automation: 1) existing architecture of legacy systems, 2) space segment diversity, and 3) unclear definition and scoping of the term, automation. Using a qualitative case study approach, we survey comparable non-satellite operation domains (Industrial Control Automation and Software Testing) that have successfully integrated automation and other satellite operation enterprises (NASA Goddard, Naval Research Laboratory, European Ground Station National Institute for Space Research in Brazil) to identify common themes and best practices. From this insight, we recommend that future satellite operation ground stations encourage the use of layered architectures, abstract satellite operation processes, and integrate simulators in future systems as concrete implementations of this common operating platform. Further elaborating on the value of these recommendations, this thesis researches the benefits of process mining in satellite operations. This research is conducted on FalconSAT-3 and PropCube 1 and 3. The process mining analysis discovered that a highly structured Concept of Operations (CONOPS) is required in order to gain significant benefits from process mining. |