Correcting Hierarchical Plans by Action Deletion

Abstract

Hierarchical task network (HTN) planning is a model-based approach to planning. The HTN domain model consists of tasks and methods to decompose them into subtasks until obtaining primitive tasks (actions). There are recent methods for verifying if a given action sequence is a valid HTN plan. However, if the plan is invalid, all existing verification methods only say so without explaining why the plan is invalid. In the paper, we propose a method that corrects a given action sequence to form a valid HTN plan by deleting the minimal number of actions. This plan correction explains what is wrong with a given action sequence concerning the HTN domain model.

Hierarchical planning and reasoning about partially ordered plans—From theory to practice

“…A closely related question was investigated recently (Barták et al. 2021), where we are again given a plan as an input. This time, however, it is not a solution already—it is just supposed to be a solution, but plan verification tells that it is not, for example, because the plan cannot be produced by the task hierarchy alone.…”

On Heuristics for Parsing-based Verification of Hierarchical Plans with a Goal Task

“…As explained in paper (Barták et al 2021b) the parsing approach uses the STRIPS model of actions (Fikes and Nilsson 1971). Let P be a set of propositions that describe properties of a world state.…”

“…The second one translates the problem into a planning problem (Höller et al 2022) (planning approach). The third one uses parsing (Barták et al 2018;2021b) (parsing approach).…”

“…It can find any task that decomposes into the given sequence of actions. Therefore it can also be used in plan recognition (Vilain 1990) or plan correction (Barták et al 2021a). However, that is not the standard definition of plan verification.…”

Was Fixing This Really That Hard? On the Complexity of Correcting HTN Domains

“…So, given a flawed planning domain, a set of white list plans, and a set of black list plans, we want to output a sequence of changes to the domain such that all plans in the white list are solutions in the modified domain and all plans in the black list are not. Barták et al (2021) studied a similar scenario where they corrected a plan instead of a domain.…”