文章背景与核心概要

本篇技术报告聚焦于实验室自动化的开源移动操作原型系统，旨在解决如何将自然语言指令与传感器观测可靠地对齐为安全机器人动作这一长期挑战。文章介绍了一种集成双臂OpenArm、移动底盘、垂直滑轨、RGB-D感知、激光雷达建图以及ROS2/MoveIt执行栈的综合系统。

该研究的核心创新在于引入了“表征交接（Representation Handoffs）”机制。通过在语言、感知、规划和安全性之间建立结构化的中间接口，系统能够系统性地桥接各模块，并在实际部署中暴露出如标定误差和物体资产缺失等关键阻碍。这为具身智能系统中多模态技术与底层机器人安全的融合提供了一种极具实践价值的调试接口方案。

基于OpenArm的实验室移动操作之表征交接 (Representation Handoffs for OpenArm-Based Laboratory Mobile Manipulation)

arXiv: [2608.07154 [cs.RO]]
提交时间: 2026年8月7日
发表会议: 机器人学：科学与系统（RSS）2026研讨会
作者: Yang Shen, Chonghao Cheng, Ziyi Zhao, Jialuo Zhu, Zhenyi Yi, Qi Zhao, Jian Yang, Yuhui Shi, Chin-Teng Lin

📌 摘要与总结

This field report introduces an open-source mobile manipulation prototype designed for laboratory automation, addressing the ongoing challenge of reliably aligning natural language instructions and sensor observations into safe robotic actions.

By integrating dual OpenArm manipulators with a mobile base, vertical slide, RGB-D sensing, lidar-based mapping, and ROS2/MoveIt execution, the authors implement a system anchored around representation handoffs. These handoffs systematically bridge language, perception, planning, and safety through structured intermediate interfaces—proving crucial for identifying deployment blockers like calibration gaps and missing object assets.

本篇现场报告介绍了一种专为实验室自动化设计的开源移动操作原型，旨在解决将自然语言指令与传感器观测可靠对齐为安全机器人动作这一持续存在的挑战。

通过将双OpenArm机械臂与移动底盘、垂直滑轨、RGB-D感知、基于激光雷达的建图以及ROS2/MoveIt执行框架集成，作者实现了一个以表征交接（representation handoffs）为核心的系统。这些交接通过结构化的中间接口系统性地连接了语言、感知、规划和安全，这对于识别诸如标定间隙和缺失物体资产等部署障碍至关重要。

📑 摘要

Open-source robotics and foundation models have lowered the barrier to embodied AI, yet language-guided laboratory automation still requires reliable alignment from instructions and observations to safe actions. This field report presents an OpenArm-based mobile manipulation prototype for laboratory-style tasks, built by integrating dual OpenArm manipulators with a mobile base, vertical slide, RGB-D sensing, lidar-based mapping, ROS2/MoveIt execution, and profile-defined skill interfaces. The system is organized around representation handoffs: natural language requests are constrained into registered skill calls, sensor observations are grounded into maps and object priors provide role and skill constraints, and runtime bindings compile validated skills into executable motion goals. We use dry-run traces and startup checks to evaluate this integration path, showing how the prototype exposes missing calibration, incomplete object assets, and unfinished real-scene visual grounding as explicit deployment blockers. These intermediate representations serve as practical debugging interfaces for integrating language, perception, planning, and robot safety in embodied systems.

开源机器人技术和基础模型降低了具身智能的门槛，然而语言引导的实验室自动化仍然需要将指令和观测可靠地对齐为安全的操作。本现场报告介绍了一种基于OpenArm的移动操作原型，用于执行实验室风格的任务。该原型通过集成双OpenArm机械臂与移动底盘、垂直滑轨、RGB-D感知、基于激光雷达的建图、ROS2/MoveIt执行模块以及配置文件定义的技能接口构建而成。系统围绕表征交接进行组织：自然语言请求被约束为已注册的技能调用，传感器观测被扎根（grounding）到地图和物体位姿中，物体先验提供角色和技能约束，运行时绑定则将验证后的技能编译为可执行的运动目标。我们利用演练轨迹（dry-run traces）和启动检查来评估这一集成路径，展示了该原型如何将缺失的标定、不完整的物体资产以及未完成的真实场景视觉接地暴露出作为显式的部署障碍。这些中间表征可作为在具身系统中集成语言、感知、规划和机器人安全的实用调试接口。

🛠️ 系统架构与集成

The system architecture is organized around the following core concepts: * Hardware Stack: Dual OpenArm manipulators, mobile base, vertical slide, RGB-D cameras, and lidar sensors. * Software Stack: ROS2 and MoveIt execution frameworks alongside profile-defined skill interfaces. * Core Representation Handoffs: 1. Language Mapping: Constraining natural language requests into registered skill calls. 2. Perception Grounding: Translating raw sensor observations into actionable maps and object poses. 3. Constraint Management: Utilizing object priors to assign functional roles and skill limitations. 4. Runtime Compilation: Binding validated skills into safe, executable motion goals.

系统架构围绕以下核心概念构建： * 硬件栈： 双OpenArm机械臂、移动底盘、垂直滑轨、RGB-D相机以及激光雷达传感器。 * 软件栈： ROS2和MoveIt执行框架，以及由配置文件定义的技能接口。 * 核心表征交接： 1. 语言映射： 将自然语言请求约束为已注册的技能调用。 2. 感知扎根： 将原始传感器观测转化为可操作的地图和物体位姿。 3. 约束管理： 利用物体先验来分配功能角色和技能限制。 4. 运行时编译： 将验证后的技能绑定为安全、可执行的运动目标。

🔗 链接与资源

• Full-Text Access: View PDF | HTML Version | TeX Source
• Digital Object Identifier (DOI): 10.48550/arXiv.2608.07154

• External Citations: Google Scholar | Semantic Scholar | NASA ADS

• 全文访问： 查看 PDF | HTML 版本 | TeX 源码

• 数字对象唯一标识符 (DOI)： 10.48550/arXiv.2608.07154
• 外部引用： Google Scholar | Semantic Scholar | NASA ADS

(Note: License icon preserved below as required by source markup) (注：根据源标记要求，下方保留了许可证图标)