Consolidating LAMA with Best-First Width Search

TL;DR

The paper tackles fast, goal-less planning (agile planning) by investigating how to combine two strong planners, LAMA and BFWS. A naive integration, adding BFWS($f_6$) as an extra open-list to LAMA, harms performance, prompting an extensive ablation study to isolate effective components. The result is a lean, new planner called NOvelty, Landmarks, and AlternatioN (NOLAN), which uses an alternation over a compact set of open-lists $[h^{FF}, h^{FF}_+, h^{LM}_+, f_2^{h^{LM}}]$ and yields state-of-the-art agile performance across 2426 IPC tasks, outperforming all baselines by about 13% in agile score. This work shows that careful pruning and consolidation of complementary search strategies can surpass naive hybrids, and identifies practical memory considerations in large domains like VisitAll for future improvement.

Abstract

One key decision for heuristic search algorithms is how to balance exploration and exploitation. In classical planning, novelty search has come out as the most successful approach in this respect. The idea is to favor states that contain previously unseen facts when searching for a plan. This is done by maintaining a record of the tuples of facts observed in previous states. Then the novelty of a state is the size of the smallest previously unseen tuple. The most successful version of novelty search is best-first width search (BFWS), which combines novelty measures with heuristic estimates. An orthogonal approach to balance exploration-exploitation is to use several open-lists. These open-lists are ordered using different heuristic estimates, which diversify the information used in the search. The search algorithm then alternates between these open-lists, trying to exploit these different estimates. This is the approach used by LAMA, a classical planner that, a decade after its release, is still considered state-of-the-art in agile planning. In this paper, we study how to combine LAMA and BFWS. We show that simply adding the strongest open-list used in BFWS to LAMA harms performance. However, we show that combining only parts of each planner leads to a new state-of-the-art agile planner.

Figures (2)

• Figure 1: Coverage over time for LAMA, BFWS($f_6$), and LAMA-W($f_2^{h^\text{LM}}$).
• Figure 2: Coverage over time for each state-of-the-art planner and NOLAN.