Department News
Trajectory Generation and Control :
Seminar Date
2000-06-22
Author
관리자
Date
2000-06-22
Views
1023
1. 제목 : Trajectory Generation and Control :
Haptic Device Design and Mobile Robot Localization
2. 연사 : 이 수 용 박사
Computer Science Department, Texas A&M University
3. 일시 : 2000년 6월 2일 (금요일) 1:00 - 2:00
4. 장소 : 301동 15층 세미나실
5. 내용 :
Robotic systems have an enormous potential to reduce human exposure in dangerous situations (e.g., battlefields) and/or increase human presence in remote locations (e.g., space). Navigation of autonomous robots involves following trajectories generated by automatic motion planners. For non-autonomous systems, the robot's trajectories are provided by the operator, either a priori or in an on-line fashion such as teleoperation. This seminar describes solutions to two important problems arising in robotic applications: teleoperation using haptic device and mobile robot localization.
Haptic device is widely used for the direct control of non-autonomous robots from a remote location. We will describe the novel design and control of an exo-skeleton type haptic device developed at the Korea Institute of Science and Technology (KIST). For example, teleoperation using the haptic device is very useful in deactivating incendiary devices task which requires fine motor control and is extremely hazardous for humans. The motion of the user's arm is accurately measured and converted to a trajectory command for the robot. The haptic device has force-feedback so the user can feel the same force as the robot. It is controlled by custom-designed controllers which have a distributed architecture for extendibility and reliability.
For autonomous mobile robots, localization is an important issue since odometer errors render it impossible for any robot to precisely follow a desired (previously planned) trajectory, resulting in a growing difference between the actual configuration and the calculated configuration as the robot travels. A new localization method using range sensor data will be presented. Our new method requires only inexpensive range sensors (such as sonar), and is applicable in partially known environments, such as buildings where one might have a blue print but not detailed knowledge of the location of furniture or (moving) obstacles. Our method first localizes the robot to a visibility sector of the environment, and then to a particular configuration in that sector. Advantages of this method are that it is computationally very simple, and that it allows precise localization without any landmarks, and moreover, it can be applied from any configuration in the environment.
6. 연사 약력 :
1989 서울대 기계공학과 학사
1991 서울대 기계설계학과 석사
1996 Ph.D, Massachusetts Institute of Technology, Dept. of
Mechanical Engineering
1996 - present 선임연구원, Advanced Robotics Research Center,
한국과학기술연구원
1999 - present Visiting assistant professor,
Dept. of Computer Science, Texas A&M University
2000.8 - Assistant professor, Dept. of Mechanical Engineering,
Texas A&M University
7. 문의처 : 기계항공공학부 조영만 교수 (880-1694)
Haptic Device Design and Mobile Robot Localization
2. 연사 : 이 수 용 박사
Computer Science Department, Texas A&M University
3. 일시 : 2000년 6월 2일 (금요일) 1:00 - 2:00
4. 장소 : 301동 15층 세미나실
5. 내용 :
Robotic systems have an enormous potential to reduce human exposure in dangerous situations (e.g., battlefields) and/or increase human presence in remote locations (e.g., space). Navigation of autonomous robots involves following trajectories generated by automatic motion planners. For non-autonomous systems, the robot's trajectories are provided by the operator, either a priori or in an on-line fashion such as teleoperation. This seminar describes solutions to two important problems arising in robotic applications: teleoperation using haptic device and mobile robot localization.
Haptic device is widely used for the direct control of non-autonomous robots from a remote location. We will describe the novel design and control of an exo-skeleton type haptic device developed at the Korea Institute of Science and Technology (KIST). For example, teleoperation using the haptic device is very useful in deactivating incendiary devices task which requires fine motor control and is extremely hazardous for humans. The motion of the user's arm is accurately measured and converted to a trajectory command for the robot. The haptic device has force-feedback so the user can feel the same force as the robot. It is controlled by custom-designed controllers which have a distributed architecture for extendibility and reliability.
For autonomous mobile robots, localization is an important issue since odometer errors render it impossible for any robot to precisely follow a desired (previously planned) trajectory, resulting in a growing difference between the actual configuration and the calculated configuration as the robot travels. A new localization method using range sensor data will be presented. Our new method requires only inexpensive range sensors (such as sonar), and is applicable in partially known environments, such as buildings where one might have a blue print but not detailed knowledge of the location of furniture or (moving) obstacles. Our method first localizes the robot to a visibility sector of the environment, and then to a particular configuration in that sector. Advantages of this method are that it is computationally very simple, and that it allows precise localization without any landmarks, and moreover, it can be applied from any configuration in the environment.
6. 연사 약력 :
1989 서울대 기계공학과 학사
1991 서울대 기계설계학과 석사
1996 Ph.D, Massachusetts Institute of Technology, Dept. of
Mechanical Engineering
1996 - present 선임연구원, Advanced Robotics Research Center,
한국과학기술연구원
1999 - present Visiting assistant professor,
Dept. of Computer Science, Texas A&M University
2000.8 - Assistant professor, Dept. of Mechanical Engineering,
Texas A&M University
7. 문의처 : 기계항공공학부 조영만 교수 (880-1694)
