GRAPPA: Generalizing and Adapting Robot Policies via Online Agentic Guidance

TL;DR

This paper introduces GRAPPA, an agentic framework utilizing specialized conversational agents to enable robots to self-guide and adapt policies online, improving manipulation success in unseen environments without extra demonstrations.

Contribution

It presents a novel multi-agent framework for robot self-guidance that grounds policies to environmental objects and adapts actions online, enhancing generalization and robustness.

Findings

Achieves higher success rates in manipulation tasks

Effective in both simulation and real-world environments

Reduces reliance on additional human demonstrations

Abstract

Robot learning approaches such as behavior cloning and reinforcement learning have shown great promise in synthesizing robot skills from human demonstrations in specific environments. However, these approaches often require task-specific demonstrations or designing complex simulation environments, which limits the development of generalizable and robust policies for unseen real-world settings. Recent advances in the use of foundation models for robotics (e.g., LLMs, VLMs) have shown great potential in enabling systems to understand the semantics in the world from large-scale internet data. However, it remains an open challenge to use this knowledge to enable robotic systems to understand the underlying dynamics of the world, to generalize policies across different tasks, and to adapt policies to new environments. To alleviate these limitations, we propose an agentic framework for robot self-guidance and self-improvement, which consists of a set of role-specialized conversational agents, such as a high-level advisor, a grounding agent, a monitoring agent, and a robotic agent. Our framework iteratively grounds a base robot policy to relevant objects in the environment and uses visuomotor cues to shift the action distribution of the policy to more desirable states, online, while remaining agnostic to the subjective configuration of a given robot hardware platform. We demonstrate that our approach can effectively guide manipulation policies to achieve significantly higher success rates, both in simulation and in real-world experiments, without the need for additional human demonstrations or extensive exploration. Code and videos available at: https://agenticrobots.github.io