There is much ongoing research into how to plan sequences of actions. Geffner and Bonet [2013] and Yang [1997] present overviews of automated planning.

The STRIPS representation was developed by Fikes and Nilsson [1971]. The Planning Domain Definition Language (PDDL) [Haslum et al., 2019] is a language that extends STRIPS in various ways, including conditional effects (when the effect only occurs when some condition holds), constraints on intermediate states, and explicit time durations. PDDL and its extensions have been used for international planning competitions.

Forward planning with good heuristics [Bacchus and Kabanza, 1996] is the basis for the most efficient algorithms [Geffner and Bonet, 2013]. (See Exercise 6.4.)

Regression planning was pioneered by Waldinger [1977]. The use of weakest preconditions is based on the work of Dijkstra [1976], where it was used to define the semantics of imperative programming languages. This should not be too surprising because the commands of an imperative language are actions that change the state of the computer.

Planning as CSP is based on Graphplan [Blum and Furst, 1997] and Satplan [Kautz and Selman, 1996] and is also investigated by Lopez and Bacchus [2003] and van Beek and Chen [1999]. Bryce and Kambhampati [2007] survey the field.

Partial-order planning was introduced in Sacerdoti’s [1975] NOAH and followed up in Tate’s [1977] NONLIN system, Chapman’s [1987] TWEAK algorithm, and McAllester and Rosenblitt’s [1991] systematic nonlinear planning (SNLP) algorithm. See Weld [1994] for an overview of partial-order planning and Kambhampati et al. [1995] for a comparison of the algorithms. The version presented here is based on SNLP.

Wilkins [1988] discusses practical issues in planning. Weld [1999], McDermott and Hendler [1995], and Nau [2007] and associated papers provide overviews.

Neufeld et al. [2019] provide a comprehensive survey of the use of planners in video games. Orkin [2006] describes the GOAP model and its use in F.E.A.R. The HTN approach is described by Ghallab et al. [2004]. Yannakakis and Togelius [2018] provide an overview of the roles of AI in game development, including planning. Colledanchise and Ögren [2018] describe behavior trees and their relationship to the subsumption architecture developed by Brooks [1986].

Planning research is published at general AI journals and conferences and specialized conferences, most notably the International Conference on Planning and Scheduling (ICAPS); https://www.icaps-conference.org.