An Intelligent Simulator for Telerobotics Training

被引:10
作者
Belghith, Khaled [1 ]
Nkambou, Roger [2 ]
Kabanza, Froduald [1 ]
Hartman, Leo [3 ]
机构
[1] Univ Sherbrooke, Dept Comp Sci, Sherbrooke, PQ J1K 2R1, Canada
[2] Univ Quebec, Dept Comp Sci, Montreal, PQ H2X 3Y7, Canada
[3] Canadian Space Agcy, St Hubert, PQ J3Y 8Y9, Canada
来源
IEEE TRANSACTIONS ON LEARNING TECHNOLOGIES | 2012年 / 5卷 / 01期
基金
加拿大自然科学与工程研究理事会;
关键词
Telerobotics training; intelligent tutoring; robot manipulation; path planning; demonstration generation; PROBABILISTIC ROADMAPS; SEARCH;
D O I
10.1109/TLT.2011.19
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
Roman Tutor is a tutoring system that uses sophisticated domain knowledge to monitor the progress of students and advise them while they are learning how to operate a space telerobotic system. It is intended to help train operators of the Space Station Remote Manipulator System (SSRMS) including astronauts, operators involved in ground-based control of SSRMS and technical support staff. Currently, there is only a single training facility for SSRMS operations and it is heavily scheduled. The training staff time is in heavy demand for teaching students, planning training tasks, developing teaching material, and new teaching tools. For example, all SSRMS simulation exercises are developed by hand and this process requires a lot of staff time. Once in an orbit ISS astronauts currently have only simple web-based material for skill development and maintenance. For long duration space flights, astronauts will require sophisticated simulation tools to maintain skills. Roman Tutor addresses these challenges by providing a portable training tool that can be installed anywhere and anytime to provide "just in time" training. It incorporates a model of the system operations curriculum, a kinematic simulation of the robotics equipment, and the ISS, a high performance path planner and an automatic task demonstration generator. For each element of the curriculum that the student is supposed to master, Roman Tutor generates example tasks for the student to accomplish within the simulation environment and then monitors its progression to provide relevant feedback when needed. Although motivated by the SSRMS application, Roman Tutor remains applicable to any telerobotics system application.
引用
收藏
页码:11 / 19
页数:9
相关论文
共 28 条
  • [1] Angros R. Jr., 2002, Proceedings of the First International Joint Conference on Autonomous Agents and Multiagent Systems, P1372
  • [2] [Anonymous], 2006, Planning algorithms
  • [3] [Anonymous], 2005, AAAI
  • [4] [Anonymous], 2001, SPRINGER TRACTS ADV
  • [5] Using temporal logics to express search control knowledge for planning
    Bacchus, F
    Kabanza, F
    [J]. ARTIFICIAL INTELLIGENCE, 2000, 116 (1-2) : 123 - 191
  • [6] Bares W. H., 1998, IUI '98. 1998 International Conference on Intelligent User Interfaces, P81, DOI 10.1145/268389.268405
  • [7] Anytime dynamic path-planning with flexible probabilistic roadmaps
    Belghith, Khaled
    Kabanza, Froduald
    Hartman, Leo
    Nkambou, Roger
    [J]. 2006 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA), VOLS 1-10, 2006, : 2372 - +
  • [8] MODEL-DRIVEN QUESTION-ANSWERING SYSTEM FOR MIXED-INITIATIVE COMPUTER-ASSISTED CONSTRUCTION
    BROWN, JS
    BURTON, RR
    ZDYBEL, F
    [J]. IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS, 1973, SMC3 (03): : 248 - 257
  • [9] Choset H., 2005, Principles of robot motion: theory, algorithms, and implementation
  • [10] Christianson DB, 1996, PROCEEDINGS OF THE THIRTEENTH NATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE, VOLS 1 AND 2, P148