Grounding of HTN Planning Domain

Abstract

Many Artificial Intelligence techniques have been developed for intelligent and autonomous systems to act and make rational decisions based on perceptions of the world state. Among these techniques, HTN (Hierarchical Task Network) planning is one of the most used in practice. HTN planning is based on expressive languages allowing to specify complex expert knowledge for real world domains. At the same time, many preprocessing techniques for classical planning were proposed to speed up the search. One of these technique, named grounding, consists in enumerating and instantiating all the possible actions from the planning problem descriptions. This technique has proven its effectiveness. Therefore, combining the expressiveness of HTN planning with the efficiency of the grounding preprocessing techniques used in classical planning is a very challenging issue. In this paper, we propose a generic algorithm to ground the domain representation for HTN planning. We show experimentally that grounding process improves the performances of state of the art HTN planners on a range of planning problems from the International Planning Competition (IPC).

The PANDA Framework for Hierarchical Planning

“…There has been little work on grounding hierarchical problems [39]. In classical planning there are techniques for state-based reachability analysis that can be reused in the hierarchical setting.…”

“…During the last years, much progress has been made, not only regarding solving techniques [1, 2, 8-11, 15, 19, 33, 34, 36, 42, 44], but also preprocessing (e.g. grounding and reachability analysis [12,39]), and related tasks like plan and goal recognition or plan verification [3,7,29].…”

HDDL: An Extension to PDDL for Expressing Hierarchical Planning Problems

“…For details on the grounding process, we refer the reader to Alford, Bercher, and Aha (2015a). There are currently only two publications devoted to grounding in more detail (see Ramoul et al, 2017, andBehnke et al, 2020). We now give the following definitions based on a ground problem.…”

“…Some description languages for HTN problems define abstract tasks only in an implicit way by their use in methods. This includes the language used by SHOP and SHOP2 (Nau et al 2003), PDDL1.2 (McDermott et al 1998, HATP, as well as GTOHP (Ramoul et al 2017). Implicit definition of the compound task set has also been chosen in some formal definitions of hierarchical problem classes (Alford, Bercher, and Aha 2015a;2015b).…”

“…We first introduce a lifted HTN formalism from the literature. Then we introduce the new language, give its syntax and meaning, discuss design choices and the differences and similarities to approaches from the literature, namely PDDL1.2 (McDermott et al 1998), SHOP(2) (Nau et al 2003), ANML (Smith, Frank, and Cushing 2008), HPDDL (Alford et al 2016a), GTOHP (Ramoul et al 2017), HTN-PDDL (González-Ferrer, Fernández-Olivares, and Castillo 2009), and HATP (de Silva, Lallement, and Alami 2015). Afterwards we give a full EBNF syntax definition based on the definition of PDDL2.1 and discuss every extension and change.…”