Abstract
In artificial intelligence planning, having an explanation of a plan given by a planner is often desirable. The ability to explain various aspects of a synthesized plan to an end user not only brings in trust on the planner but also reveals insights of the planning domain and the planning process. Contrastive questions such as “Why action A instead of action B?” can be answered with a contrastive explanation that compares properties of the original plan containing A against the contrastive plan containing B. In this article, we explore a set of contrastive questions that a user of a planning tool may raise and propose a re-model and re-plan framework to provide explanations to such questions. Earlier work has reported this framework on planning instances for discrete problem domains described in the Planning Domain Definition Language (PDDL) and its variants. In this article, we propose an extension for planning instances described by PDDL+ for hybrid systems that portray a mix of discrete-continuous dynamics. Specifically, given a mixed discrete-continuous system model in PDDL+ and a plan describing the set of desirable actions on the same to achieve a destined goal, we present a framework that can integrate contrastive questions in PDDL+ and synthesize alternate plans. We present a detailed case study on our approach and propose a comparison metric to compare the original plan with the alternate ones.
- [1] GitHub. [n.d.]. KCL-Planning/SMTPlan. Retrieved September 22, 2022 from https://github.com/KCL-Planning/SMTPlan/tree/master/benchmarks.Google Scholar
- [2] . 1998. PDDL—The Planning Domain Definition Language.
Technical Report . Yale Center for Computational Vision and Control.Google Scholar - [3] . 1995. The algorithmic analysis of hybrid systems. Theoretical Computer Science 138, 1 (1995), 3–34. Google Scholar
Digital Library
- [4] . 1993. Hybrid automata: An algorithmic approach to the specification and verification of hybrid systems. In Hybrid Systems, , , , and (Eds.). Springer, Berlin, Germany, 209–229. Google Scholar
Cross Ref
- [5] . 1999. Planning as heuristic search: New results. In Recent Advances in AI Planning. Lecture Notes in Computer Science, Vol. 1089. Springer, 360–372. Google Scholar
Cross Ref
- [6] . 2019. Towards explainable AI planning as a service. In Proceedings of the 2nd ICAPS Workshop on Explainable Planning. https://strathprints.strath.ac.uk/69987/.Google Scholar
- [7] . 2020. Planning for hybrid systems via satisfiability modulo theories. Journal of Artificial Intelligence Research 67 (2020), 235–283. Google Scholar
Cross Ref
- [8] . 2018. Explicability? Legibility? Predictability? Transparency? Privacy? Security? The emerging landscape of interpretable agent behavior. arXiv E-prints, arXiv:1811.09722 (2018).Google Scholar
- [9] . 2017. Plan explanations as model reconciliation: Moving beyond explanation as soliloquy. In Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI’17). 156–163.Google Scholar
Cross Ref
- [10] . 2018. Explanations based on the missing: Towards contrastive explanations with pertinent negatives. In Advances in Neural Information Processing Systems 31, , , , , , and (Eds.). Curran Associates, 592–603. http://papers.nips.cc/paper/7340-explanations-based-on-the-missing-towards-contrastive-explanations-with-pertinent-negatives.pdf.Google Scholar
- [11] . 2019. Model agnostic contrastive explanations for structured data. arXiv E-prints, arXiv:1906.00117 (2019).Google Scholar
- [12] . 2020. A new approach to plan-space explanation: Analyzing plan-property dependencies in oversubscription planning. In Proceedings of the 34th AAAI Conference on Artificial Intelligence (AAAI’20), the 32nd Innovative Applications of Artificial Intelligence Conference (IAAI’20), and the 10th AAAI Symposium on Educational Advances in Artificial Intelligence (EAAI’20).9818–9826. https://ojs.aaai.org/index.php/AAAI/article/view/6534.Google Scholar
- [13] . 2020. Certified unsolvability for SAT planning with property directed reachability. In Proceedings of the 30th International Conference on Automated Planning and Scheduling. 90–100. https://ojs.aaai.org/index.php/ICAPS/article/view/6649.Google Scholar
- [14] . 2017. Unsolvability certificates for classical planning. In Proceedings of the 27th International Conference on Automated Planning and Scheduling (ICAPS’17). 88–97. https://aaai.org/ocs/index.php/ICAPS/ICAPS17/paper/view/15734.Google Scholar
- [15] . 1971. Strips: A new approach to the application of theorem proving to problem solving. Artificial Intelligence 2, 3 (1971), 189–208. Google Scholar
Cross Ref
- [16] . 2006. Modelling mixed discrete-continuous domains for planning. Journal of Artificial Intelligence Research 27 (2006), 235–297. Google Scholar
Digital Library
- [17] . 2017. Explainable planning. arXiv E-prints, arXiv:1709.10256 (
Sept. 2017).Google Scholar - [18] . 2012. \(\delta\)-complete decision procedures for satisfiability over the reals. In Automated Reasoning, , , and (Eds.). Springer, Berlin, Germany, 286–300. Google Scholar
Digital Library
- [19] . 2012. Delta-decidability over the reals. In Proceedings of the 27th Annual IEEE Symposium on Logic in Computer Science (LICS’12). IEEE, Los Alamitos, CA, 305–314. Google Scholar
Digital Library
- [20] . 2014. Delta-complete analysis for bounded reachability of hybrid systems. CoRR abs/1404.7171 (2014).Google Scholar
- [21] . 2002. LPG: A planner based on local search for planning graphs with action costs. In Proceedings of the 6th International Conference on Artificial Intelligence Planning Systems. 13–22. http://www.aaai.org/Library/AIPS/2002/aips02-002.php.Google Scholar
- [22] . 2010. Coming up with good excuses: What to do when no plan can be found. In Proceedings of the 20th International Conference on Automated Planning and Scheduling (ICAPS’20). 81–88. http://www.aaai.org/ocs/index.php/ICAPS/ICAPS10/paper/view/1453.Google Scholar
- [23] . 2006. The fast downward planning system. Journal of Artificial Intelligence Research 26 (2006), 191–246. Google Scholar
Cross Ref
- [24] . 2001. FF: The fast-forward planning system. AI Magazine 22, 3 (2001), 57–62. Google Scholar
Digital Library
- [25] . 2019. Explainable AI planning (XAIP): Overview and the case of contrastive explanation. In Reasoning Web: Explainable Artificial Intelligence. Lecture Notes in Computer Science, Vol. 11810. Springer, 277–282.Google Scholar
Digital Library
- [26] . 2001. The FF planning system: Fast plan generation through heuristic search. Journal of Artificial Intelligence Research 14 (2001), 253–302. Google Scholar
Cross Ref
- [27] . 2006. SatPlan: Planning as Satisfiability. Department of Computer Science and Engineering, University of Washington, Seattle, WA.Google Scholar
- [28] . 2019. Model-based contrastive explanations for explainable planning. In Proceedings of the ICAPS 2019 Workshop on Explainable AI Planning (XAIP’19). https://strathprints.strath.ac.uk/69957/.Google Scholar
- [29] . 2021. Contrastive explanations of plans through model restrictions. CoRR abs/2103.15575 (2021).Google Scholar
- [30] . 2009. Why and why not explanations improve the intelligibility of context-aware intelligent systems. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (CHI’09). ACM, New York, NY, 2119–2128. Google Scholar
Digital Library
- [31] . 2018. Contrastive explanation: A structural-model approach. arXiv E-prints, arXiv:1811.03163 (2018).Google Scholar
- [32] . 1989. ADL: Exploring the middle ground between STRIPS and the situation calculus. In Proceedings of the 1st International Conference on Principles of Knowledge Representation and Reasoning. 324–332. Google Scholar
- [33] . 2012. A universal planning system for hybrid domains. Applied Intelligence 36, 4 (2012), 932–959. Google Scholar
Digital Library
- [34] . 2009. UPMurphi: A tool for universal planning on PDDL+ problems. In Proceedings of the 19th International Conference on Automated Planning and Scheduling (ICAPS’09). 106–113. http://aaai.org/ocs/index.php/ICAPS/ICAPS09/paper/view/707.Google Scholar
- [35] . 2021. Translations from discretised PDDL+ to numeric planning. In Proceedings of the 31st International Conference on Automated Planning and Scheduling (ICAPS’21). 252–261. https://ojs.aaai.org/index.php/ICAPS/article/view/15969.Google Scholar
- [36] . 2016. Heuristic planning for hybrid systems. In Proceedings of the 30th AAAI Conference on Artificial Intelligence. 4254–4255. http://www.aaai.org/ocs/index.php/AAAI/AAAI16/paper/view/12394.Google Scholar
- [37] . 2020. A contrastive plan explanation framework for hybrid system models. In Proceedings of the 2020 18th ACM/IEEE International Conference on Formal Methods and Models for System Design (MEMOCODE’20). 1–11. Google Scholar
Cross Ref
- [38] . 2016. Interval-based relaxation for general numeric planning. In Proceedings of the 22nd European Conference on Artificial Intelligence (ECAI’16). 655–663. Google Scholar
Digital Library
- [39] . 2005. Processes and continuous change in a SAT-based planner. Artificial Intelligence 166, 1-2 (2005), 194–253. Google Scholar
Cross Ref
- [40] . 2012. Planning as an iterative process. In Proceedings of the 26th AAAI Conference on Artificial Intelligence (AAAI’12). 2180–2185.Google Scholar
- [41] . 1998. Conformant graphplan. In Proceedings of the 15th National/10th Conference on Artificial Intelligence/Innovative Applications of Artificial Intelligence (AAAI’98/IAAI’98). 889–896. Google Scholar
- [42] . 2019. Why can’t you do that HAL? Explaining unsolvability of planning tasks. In Proceedings of the 28th International Joint Conference on Artificial Intelligence (IJCAI’19). 1422–1430. Google Scholar
Cross Ref
Index Terms
A Contrastive Plan Explanation Framework for Hybrid System Models
Recommendations
HTN Plan Repair via Model Transformation
KI 2020: Advances in Artificial IntelligenceAbstractTo make planning feasible, planning models abstract from many details of the modeled system. When executing plans in the actual system, the model might be inaccurate in a critical point, and plan execution may fail. There are two options to handle ...
A Logic-Based Explanation Generation Framework for Classical and Hybrid Planning Problems
In human-aware planning systems, a planning agent might need to explain its plan to a human user when that plan appears to be non-feasible or sub-optimal. A popular approach, called model reconciliation, has been proposed as a way to bring the model of ...






Comments