Uploadable Executable Specification Concept for Spacecraft Autonomy Systems

Abstract

Current spacecraft autonomy systems suffer from two main problems. First, autonomy designs cannot be adequately reviewed by system engineers, resulting in a potential loss of desired system behavior between system-level requirements and software implementation. Second, current autonomy systems cannot fully assess the systems-level impact of modifications and then quickly and safely upload those modifications to the spacecraft pre-and post-launch. These problems are addressed by the development of executable specification techniques to directly support system engineers with formalized models that translate into operational functionality. This paper describes a concept of combining a standard executable specification technique with a concept of software design using uploadable forms. This paper goes on to describe the features of this concept which include: interactive visual design and display capabilities that allow any domain expert to understand and/or perform the design; operational support capabilities that allow the on-board autonomy functionality to be modified or disabled in real-time without patching or modifying existing code; and graphical stand-alone simulation and automated verification capabilities that allow autonomy designs to proven safe prior to upload.