Conditional Multi-Stage Failure Recovery for Embodied Agents

TL;DR

This paper presents a novel conditional multistage failure recovery framework for embodied agents using zero-shot chain prompting, significantly improving error handling and task success rates in complex environments.

Contribution

The work introduces a four-stage error recovery framework leveraging LLM reasoning, achieving state-of-the-art results on the TfD benchmark with zero-shot capabilities.

Findings

Outperforms baseline by 11.5% in failure recovery

Surpasses existing models by 19% on TfD benchmark

Demonstrates effective multi-stage error handling in embodied agents

Abstract

Embodied agents performing complex tasks are susceptible to execution failures, motivating the need for effective failure recovery mechanisms. In this work, we introduce a conditional multistage failure recovery framework that employs zero-shot chain prompting. The framework is structured into four error-handling stages, with three operating during task execution and one functioning as a post-execution reflection phase. Our approach utilises the reasoning capabilities of LLMs to analyse execution challenges within their environmental context and devise strategic solutions. We evaluate our method on the TfD benchmark of the TEACH dataset and achieve state-of-the-art performance, outperforming a baseline without error recovery by 11.5% and surpassing the strongest existing model by 19%.