RI: Medium: Robust Intelligent Manipulation and Apprenticeship Learning for Robotic Surgical Assistants

RI：中：机器人手术助手的鲁棒智能操作和学徒学习

• 批准号: 0905344
• 负责人: Cenk Cavusoglu
• 金额: $ 135.99万
• 依托单位: Case Western Reserve University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-08-01 至 2013-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0905344&HistoricalAwards=false
• 关键词: RI; Medium; Robust; Intelligent; Manipulation

"This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5)."This project aims to develop models, algorithms, and testbeds for robust intelligent manipulation that will enable supervisory control for robotic surgical assistants (RSAs). Drawing on the combined expertise of the investigators, the research is focusing on both analytic and empirical approaches.Using analytic approaches, the investigators are defining mathematical models of system state, deformable tissue dynamics, and stochastic uncertainty. These models are used to develop robotic motion planning and control algorithms to plan and perform grasping and manipulation of deformable objects under uncertainty. Using empirical approaches, the investigators are developing machine learning techniques to efficiently find dynamic control policies and characterize performance metrics based on expert demonstrations. Simulation and hardware testbeds for a common set of benchmark problems are being developed to evaluate these algorithms and methods.This project will advance basic scientific understanding by developing new analytic methods for robust grasping and manipulation of deformable objects. This project will also advance empirical approaches by developing controllers that learn deformable object manipulation skills by observing expert demonstrations. The project will extend, compare, and evaluate analytic and empirical methods and seek to develop new hybrid methods within the focused context of providing robust intelligence for RSAs.Robust intelligent manipulation for RSAs will improve patient health and reduce costs by enhancing surgeon performance, reducing tedium, and decreasing operation time. Outreach to local girls' high schools and predominantly minority Cleveland high schools is pursued as part of the project to encourage participation of underrepresented groups in engineering and computer science.

“该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。“该项目旨在开发用于强大智能操作的模型，算法和测试平台，从而实现机器人手术助理（RSA）的监督控制。 利用研究人员的综合专业知识，研究集中在分析和经验的方法。使用分析方法，研究人员正在定义系统状态，可变形组织动力学和随机不确定性的数学模型。 这些模型用于开发机器人运动规划和控制算法，以规划和执行不确定性下的可变形物体的抓取和操纵。 使用经验方法，研究人员正在开发机器学习技术，以有效地找到动态控制策略，并根据专家演示来表征性能指标。 目前正在为一组常见的基准问题开发模拟和硬件测试平台，以评估这些算法和方法。该项目将通过开发新的分析方法来加强对可变形物体的稳健抓取和操纵，从而促进基本的科学认识。 该项目还将通过开发控制器来推进经验方法，通过观察专家演示来学习可变形物体操作技能。 该项目将扩展、比较和评估分析和经验方法，并寻求在为RSAs提供强大智能的重点背景下开发新的混合方法。RSAs的强大智能操作将通过提高外科医生的表现、减少乏味和减少手术时间来改善患者健康并降低成本。 作为鼓励代表性不足的群体参与工程和计算机科学的项目的一部分，继续向当地女子高中和克利夫兰少数民族占多数的高中进行宣传。

项目成果

Supervisory control of a DaVinci surgical robot

达芬奇手术机器人的监控

• DOI: 10.1109/iros.2017.8206389
• 发表时间: 2017
• 期刊: Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2017
• 作者: Chow, Der-Lin;Xu, Peng;Tuna, Eser;Huang, Siqi;Cavusoglu, M. Cenk;Newman, Wyatt
• 通讯作者: Newman, Wyatt