Autonomously controlling flexible timelines: From domain-independent planning to robust execution
Tiago Nogueira, Simone Fratini, Klaus Schilling · 2017
The introduction of more onboard autonomy in future single and multi-satellite missions is both a question of limited onboard resources and of how far can we actually thrust the autonomous functionalities deployed on board. In-flight experience with nasa's Deep Space 1 and Earth Observing 1 has shown how difficult it is to design, build and test reliable software for autonomy. The degree to which system-level onboard autonomy will be deployed in the single and multi satellite systems of tomorrow will depend, among other things, on the progress made in two key software technologies: autonomous onboard planning and robust execution. Parallel to the developments in these two areas, the actual integration of planning and execution engines is still nowadays a crucial issue in practical application. This paper presents an onboard autonomous model-based executive for execution of time-flexible plans. It describes its interface with an apsi-based timeline-based planner, its control approaches, architecture and its modelling language as an extension of apsl's ddl. In addition, it introduces a modified version of the classical blocks world toy planning problem which has been extended in scope and with a runtime environment for evaluation of integrated planning and executive engines.