CAPE: Corrective Actions from Precondition Errors using Large Language Models
TL;DR – In this paper, we introduce CAPE: an approach to correct errors encountered during robot plan execution. We exploit the ability of large language models to generate high-level plans and to reason about causes of errors.