Hierarchical planning is a practically important approach to automated planning based on encoding abstract plans as hierarchical task networks (HTNs) [1], [2]. The network describes how compound tasks are decomposed, via decomposition methods, to sub-tasks and eventually to actions forming a plan.
Depending on the chosen HTN formalization, the decomposition methods may specify additional constraints among the sub-tasks such as partial ordering, state constraints, or causal links. The obtained sequence of actions must be executable in the classical sense, but also adhere to the constraints specified by the decomposition methods.