The LAMA Planner: Guiding Cost-Based Anytime Planning with Landmarks

By Silvia Richter, Matthias Westphal. Journal of Artificial Intelligence Research, 2010

790 citations recorded by the source corpus.

LAMA is a classical planning system based on heuristic forward search. Its core feature is the use of a pseudo-heuristic derived from landmarks, propositional formulas that must be true in every solution of a planning task. LAMA builds on the Fast Downward planning system, using finite-domain rather than binary state variables and multi-heuristic search. The latter is employed to combine the landmark heuristic with a variant of the well-known FF heuristic. Both heuristics are cost-sensitive, focusing on high-quality solutions in the case where actions have non-uniform cost. A weighted A* search is used with iteratively decreasing weights, so that the planner continues to search for plans of better quality until the search is terminated. LAMA showed best performance among all planners in the sequential satisficing track of the International Planning Competition 2008. In this paper we present the system in detail and investigate which features of LAMA are crucial for its performance. We present individual results for some of the domains used at the competition, demonstrating good and bad cases for the techniques implemented in LAMA. Overall, we find that using landmarks improves performance, whereas the incorporation of action costs into the heuristic estimators proves not to be beneficial. We show that in some domains a search that ignores cost solves far more problems, raising the question of how to deal with action costs more effectively in the future. The iterated weighted A* search greatly improves results, and shows synergy effects with the use of landmarks.

Subfields: Artificial Intelligence, Computer Networks and Communications, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/afc2cdd6c97cda49b829f6b8fe5da7957fb2af4e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Networks and Communications

In Computer Vision and Pattern Recognition

Citations in the corpus

Planning as satisfiability: Heuristics (2012, methodology): “gap between BLACKBOX style planners and recent state-space search planners such as LAMA [40].”
Classical Planning with Simulators: Results on the Atari Video Games (2015, methodology): “This way for combining multiple heuristics is used in the LAMA planner [Richter and Westphal, 2010], and was introduced in the planner Fast Downward [Helmert, 2006].”
Fast Detection of Unsolvable Planning Instances Using Local Consistency (2013, methodology): “FORWARD (Hoffmann and Nebel 2001), F AST DOWNWARD (Helmert 2006b), LAMA (Richter and Westphal 2010), and planners based on compilation into SAT instances (Rintanen, Heljanko, and Niemelä 2006).”
Automated Planning and Acting (2016, background): “19 The literature contains several examples of cases in which the problem representation and the computation of heuristic functions can be done more easily with state variables than with propositions [268, 510].”
"Distance"? Who Cares? Tailoring Merge-and-Shrink Heuristics to Detect Unsolvability (2014, background): “[16]), and the recent red-black plan heuristics [12], are no better at unsolvability detection than the “Methuselah heuristic” h.”
Learning Generalized Reactive Policies using Deep Neural Networks (2017, methodology): “seen, instances as well as the performance of two state-of-theart planners: LAMA (Richter and Westphal 2010) which uses the Fast Downward (FD, (Helmert 2006)) planning framework, and Fast Forward (FF, (Jörg Hoffman 2001)) planner.”