Abstract Todays’ applications of manipulators in industry are restricted to environments which are known in advance and specially adapted to robots, as these manipulators are not flexible enough to move reliably in unknown or complex environments. In this paper, we present a multi-agent based reactive planning and control system for redundant manipulators. This approach enables the reactive motion of manipulators in an unknown environment. In the system described in this paper, an agent is responsible for calculating and controlling the motion of one joint, therefore it is called “joint agent”. A joint agent is responsible for evaluating the sensors on the according link, for calculating motions of the joint which minimize the distance of the endeffector to a cartesian goal and which avoid sensed obstacles. These local algorithms possess only small complexity compared to the algorithms needed for global optimal motion planning and can be calculated very fast. Thus reactive motions are possible. The agents are coordinated by communication. Additionally to the coordination by communication, a superior level is responsible for optimizing the agents’ actions by evaluating complexer criteria. Besides that, the local sensor data is integrated into a global model of the environment on this level. This paper concentrates on the local reactive level. Several local criteria in order to generate the reactive movements are presented. Our testbed, a redundant manipulator covered with tactile sensors, is presented in detail.
[1]
Vladimir J. Lumelsky,et al.
Motion planning for a whole-sensitive robot arm manipulator
,
1990,
Proceedings., IEEE International Conference on Robotics and Automation.
[2]
Yoram Koren,et al.
Design and motion planning of a mechanical snake
,
1993,
IEEE Trans. Syst. Man Cybern..
[3]
Petra Bohner,et al.
A multi-agent approach to distributed control for task-level programs for cooperating manipulators
,
1994,
Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS'94).
[4]
Gregory S. Chirikjian,et al.
A binary paradigm for robotic manipulators
,
1994,
Proceedings of the 1994 IEEE International Conference on Robotics and Automation.
[5]
Vladimir J. Lumelsky,et al.
Sensor-based motion planning in three dimensions for a highly redundant snake robot
,
1994,
Adv. Robotics.