Robot-R1: Reinforcement Learning for Enhanced Embodied Reasoning in Robotics

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kim, Dongyoung, Park, Sumin, Jang, Huiwon, Shin, Jinwoo, Kim, Jaehyung, Seo, Younggyo
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909992348549120
author Kim, Dongyoung
Park, Sumin
Jang, Huiwon
Shin, Jinwoo
Kim, Jaehyung
Seo, Younggyo
author_facet Kim, Dongyoung
Park, Sumin
Jang, Huiwon
Shin, Jinwoo
Kim, Jaehyung
Seo, Younggyo
contents Large Vision-Language Models (LVLMs) have recently shown great promise in advancing robotics by combining embodied reasoning with robot control. A common approach involves training on embodied reasoning tasks related to robot control using Supervised Fine-Tuning (SFT). However, SFT datasets are often heuristically constructed and not explicitly optimized for improving robot control. Furthermore, SFT often leads to issues such as catastrophic forgetting and reduced generalization performance. To address these limitations, we introduce Robot-R1, a novel framework that leverages reinforcement learning to enhance embodied reasoning specifically for robot control. Robot-R1 learns to predict the next keypoint state required for task completion, conditioned on the current scene image and environment metadata derived from expert demonstrations. Inspired by the DeepSeek-R1 learning approach, Robot-R1 samples reasoning-based responses and reinforces those that lead to more accurate predictions. To rigorously evaluate Robot-R1, we also introduce a new benchmark that demands the diverse embodied reasoning capabilities for the task. Our experiments show that models trained with Robot-R1 outperform SFT methods on embodied reasoning tasks. Despite having only 7B parameters, Robot-R1 even surpasses GPT-4o on reasoning tasks related to low-level action control, such as spatial and movement reasoning.
format Preprint
id arxiv_https___arxiv_org_abs_2506_00070
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Robot-R1: Reinforcement Learning for Enhanced Embodied Reasoning in Robotics
Kim, Dongyoung
Park, Sumin
Jang, Huiwon
Shin, Jinwoo
Kim, Jaehyung
Seo, Younggyo
Robotics
Artificial Intelligence
Large Vision-Language Models (LVLMs) have recently shown great promise in advancing robotics by combining embodied reasoning with robot control. A common approach involves training on embodied reasoning tasks related to robot control using Supervised Fine-Tuning (SFT). However, SFT datasets are often heuristically constructed and not explicitly optimized for improving robot control. Furthermore, SFT often leads to issues such as catastrophic forgetting and reduced generalization performance. To address these limitations, we introduce Robot-R1, a novel framework that leverages reinforcement learning to enhance embodied reasoning specifically for robot control. Robot-R1 learns to predict the next keypoint state required for task completion, conditioned on the current scene image and environment metadata derived from expert demonstrations. Inspired by the DeepSeek-R1 learning approach, Robot-R1 samples reasoning-based responses and reinforces those that lead to more accurate predictions. To rigorously evaluate Robot-R1, we also introduce a new benchmark that demands the diverse embodied reasoning capabilities for the task. Our experiments show that models trained with Robot-R1 outperform SFT methods on embodied reasoning tasks. Despite having only 7B parameters, Robot-R1 even surpasses GPT-4o on reasoning tasks related to low-level action control, such as spatial and movement reasoning.
title Robot-R1: Reinforcement Learning for Enhanced Embodied Reasoning in Robotics
topic Robotics
Artificial Intelligence
url https://arxiv.org/abs/2506.00070