Commonsense-Grounded Path Planning from Abstract Instructions

Masafumi Endo, Kohei Honda, Ryo Yonetani

We present \emph{commonsense ranked search} (CoRS), a novel path planner that turns an abstract instruction into a route that follows commonsense. While existing methods respect the considerations written down in advance, a robot working among people must follow those left unstated too, as with a wet floor that a worker avoids without being told. CoRS leverages large language models (LLMs) and vision-language models (VLMs) as commonsense knowledge to reason about these latent considerations in its planning. Given an abstract instruction (\emph{e.g.}, ``move carefully'') and visual observations of each region in the environment, CoRS derives a consideration for each region, as in ``this wet floor is slippery and worth a detour.'' It then compares the considerations between regions to see which of the two the robot should avoid more, as in ``the crowd is worse than the wet floor.'' These judgments sort the regions into a commonsense ranking, whose costs drive a conventional search that always returns a valid route. We build a benchmark for planning under latent considerations, with three environments, 1350 problems, and five instructions at three levels of abstraction. Experiments…