Application of MBSE to model Hierarchical AI Planning problems in HDDL

This paper, presented at the KEPS Workshop at ICAPS 2021, proposes a methodology to bridge the gap between MBSE and AI planning; specifically, to generate HDDL (Hierarchical Domain Definition Language) domain and problem files directly from SysML models. The core idea is that HDDL problem and domain files are built up from SysML formalism, which is already widely used in MBSE to design and study space mission architectures. The motivation is twofold: HDDL files are hard to share and modify without losing traceability, and AI planning languages like HDDL are rarely taught to space systems engineers, who are the primary architects of space concepts of operations.