Robotouille: An Asynchronous Planning Benchmark for LLM Agents

TL;DR

Robotouille introduces a new benchmark environment to evaluate large language model agents' ability to perform complex, long-horizon asynchronous planning tasks, revealing current limitations and areas for improvement.

Contribution

The paper presents Robotouille, a novel asynchronous planning benchmark for LLM agents, with datasets capturing complex scenarios beyond existing short-horizon benchmarks.

Findings

ReAct (gpt4-o) achieves 47% on synchronous tasks.

ReAct (gpt4-o) achieves 11% on asynchronous tasks.

Analysis highlights the need for better long-horizon feedback incorporation.

Abstract

Effective asynchronous planning, or the ability to efficiently reason and plan over states and actions that must happen in parallel or sequentially, is essential for agents that must account for time delays, reason over diverse long-horizon tasks, and collaborate with other agents. While large language model (LLM) agents show promise in high-level task planning, current benchmarks focus primarily on short-horizon tasks and do not evaluate such asynchronous planning capabilities. We introduce Robotouille, a challenging benchmark environment designed to test LLM agents' ability to handle long-horizon asynchronous scenarios. Our synchronous and asynchronous datasets capture increasingly complex planning challenges that go beyond existing benchmarks, requiring agents to manage overlapping tasks and interruptions. Our results show that ReAct (gpt4-o) achieves 47% on synchronous tasks but only 11% on asynchronous tasks, highlighting significant room for improvement. We further analyze failure modes, demonstrating the need for LLM agents to better incorporate long-horizon feedback and self-audit their reasoning during task execution. Code is available at https://github.com/portal-cornell/robotouille.