MBSE models are useful knowledge engineering models that are widely use in System Engineering and Space Mission Analysis. In our case, we worked with AI planning and specifically HDDL. We therefore implemented HIST (HDDL fIles SysML Translation) to be able to extract our MBSE model information and directly have it ready to test for our systems. The tool has been implemented as a part of the IGLUNA-CoRoDro analogue mission, however, it is still under development to add capabilities and test its robusteness in different modelling context. You are free re-use the code and change it. However, if you are using it for your project, please cite:
@article{rimani2021applicationMBSE,
title={Application of MBSE to model Hierarchical AI Planning problems in HDDL},
author={Rimani, Jasmine and Lesire, Charles and Lizy-Destrez, Stéphanie and Viola, Nicole},
publisher={ International Conference on Automated Planning and Scheduling (ICAPS) 2021, KEPS Workshop},
year={2021}
}
This work is distributed under the MIT License.
The code is still under development. We try to polish it, test it and add modules as we encounter new challenges in the domain of autonomous systems.
- The "NewDevelopmentPath" branch is a cleaner version of the "PreviousVersion", code-wise, and that can manage uml's "buffer nodes" in SysML:
- The domain definition is stable and shows better performance than the previous version. It still needs to be re-validated.
- The problem definition is stable. It encodes objects and initial conditions for a specific scenario derived from the MBSE model.
This code extracts from a model in Papyrus (https://www.eclipse.org/papyrus/) the information to automatically generate an HDDL domain file.
In addition, the code can partially generate an HDDL problem files. Because AI planning problem files are domain dependent, they are difficult to generalize. We are still trying to find a way to do so.
The code provides a feedback file:
- with the possible problem that may be arised from the translation of the domain file from SysML to HDDL.