non-linear planning

Terms from Artificial Intelligence: humans at the heart of algorithms

Page numbers are for draft copy at present; they will be replaced with correct numbers when final book is formatted. Chapter numbers are correct and will not change now.

Non-linear planning determines a partially ordered set of operations or actions for a robot or software agent to perform in order to achieve a goal. For example, in order to make a mug of tea (goal) you have four actions (A) boil water, (B) put a teabag in a mug, (C) pour water into the mug, and finally (D) remove the teabag when fully brewed. Action C must precede D and both A and B must precede C, but A and B could be performed in either order (or indeed concurrently). It is in contrast to linear planning, which creates a fixed sequence of actions. Non-linear planning may be easier to create, however when the agent needs to act, the partial order must be resoved to give a fixed order (one or other of A or B needs to be performed first).

Used in Chap. 15: page 242