[论文解读] An Open-Source Robotics Research Platform for Autonomous Laparoscopic Surgery

Ariel Rodriguez, Lorenzo Mazza+5 more|arXiv (Cornell University)|Mar 9, 2026

Surgical Simulation and Training被引用 0

一句话总结

论文提出一个开源、机器人不可知的RCM控制器，具有闭式形式的解析速度求解器，以及一个基于ROS的平台，使用UR5e和Panda手臂实现远程操作、数据采集和自主策略部署用于腹腔镜任务，已在幻象、离体和在体猪模型中进行验证。

ABSTRACT

Autonomous robot-assisted surgery demands reliable, high-precision platforms that strictly adhere to the safety and kinematic constraints of minimally invasive procedures. Existing research platforms, primarily based on the da Vinci Research Kit, suffer from cable-driven mechanical limitations that degrade state-space consistency and hinder the downstream training of reliable autonomous policies. We present an open-source, robot-agnostic Remote Center of Motion (RCM) controller based on a closed-form analytical velocity solver that enforces the trocar constraint deterministically without iterative optimization. The controller operates in Cartesian space, enabling any industrial manipulator to function as a surgical robot. We provide implementations for the UR5e and Franka Emika Panda manipulators, and integrate stereoscopic 3D perception. We integrate the robot control into a full-stack ROS-based surgical robotics platform supporting teleoperation, demonstration recording, and deployment of learned policies via a decoupled server-client architecture. We validate the system on a bowel grasping and retraction task across phantom, ex vivo, and in vivo porcine laparoscopic procedures. RCM deviations remain sub-millimeter across all conditions, and trajectory smoothness metrics (SPARC, LDLJ) are comparable to expert demonstrations from the JIGSAWS benchmark recorded on the da Vinci system. These results demonstrate that the platform provides the precision and robustness required for teleoperation, data collection and autonomous policy deployment in realistic surgical scenarios.

研究动机与目标

• 提供一个开源、机器人不可知的平台，强制执行腹腔镜手术的手术通道RCM约束。
• 交付一个完整栈的ROS系统，实现远程操作、数据采集和自主策略部署。
• 在幻象、离体和在体猪模型中验证RCM精度和轨迹光滑性。
• 通过收集的数据集进行模仿学习，展示数据收集和策略学习就绪性。

提出的方法

• 开发一个闭式解析速度控制器，能够在笛卡尔空间对任何合适的机械臂确定性地强制执行RCM。
• 在ROS软件栈中集成两种硬件配置（UR5e和Franka Panda），用于远程操作和策略部署。
• 实现基于ZMQ的服务器–客户端架构，将策略推断与ROS解耦，并加入用于动作验证的安全控制器。
• 在远程操作中记录原始传感器数据和内窥镜视频，创建用于学习的结构化数据集。
• 应用五阶多项式速度规划器，按照冲击和加速度限制生成朝向目标的平滑运动。

Figure 2: The platform is built with modularity and scalability at its core, with inter-component communication abstracted through ROS topics to remain hardware-agnostic. This architecture allows input devices and robot actuators to be readily exchanged, and new modalities can be incorporated by sim

实验结果

研究问题

• RQ1 一个开源、机器人不可知的RCM控制器是否能够在幻象、离体和在体设置中实现亚毫米级的RCM偏差？
• RQ2 该平台是否能够在实现自主策略部署的同时，产生与既有基准相当的专家级轨迹光滑性？
• RQ3 基于ROS的全栈架构是否在硬件和环境之间提供稳健的远程操作、数据采集和策略落地能力？
• RQ4 在收集数据上训练的模仿学习策略是否能够在腹腔镜任务（如肠道抓握和牵拉）中实现较高的成功率？

主要发现

• RCM偏差在幻象、离体和在体试验中保持亚毫米级。
• 轨迹光滑性指标（SPARC、LDLJ）与在da Vinci系统的JIGSAWS基准中的专家演示相当。
• 在该平台数据集上训练的模仿学习策略达到85%的成功率（17/20次.roll-outs）。
• 在真实手术情景下，平台在远程操作、示范记录和自主策略部署中保持稳健表现。
• 系统具有模块化、硬件无关性，支持UR5e和Panda手臂，配备立体内窥镜和沉浸式可视化。
• RCM控制器在无需迭代优化的情况下工作，使用闭式解析速度解以确定性地强制执行约束。

Figure 4: Immersive stereoscopic visualization through the Meta Quest headset running Endomersion [ 11 ] showing the surgical workspace captured by the endoscope. Only the left eye view is shown; the right eye receives the corresponding stereo pair to enable depth perception.