Institute for
Robotics and Process Control

Deutsch   English

Hybrid Robot Control Concepts

Project Description

Purely position controlled robots are still state of the art in industrial assembly lines. Due to a permanently increasing degree of automation, robot control architectures become more and more sophisticated. Since program developers demand more than simple position feedback, several sensors like force/torque sensors, distance sensors, or computer vision systems are supposed to be integrated into a robot control architecture. To keep an easy handling of such complex systems, a programming interface based on very simple manipulator movements has been developed at our institute. Due to multiple sensors, several controllers have to be consolidated, i.e. a hybrid controller, that enables position, velocity, and force/torque control in six degrees of freedom is supposed to be researched. It is possible to determine desired values for each degree of freedom separately. By this approach, one degree can be force controlled, another can be velocity controlled, and all remaining degrees of freedom may be position controlled.

Beside the intuitive programming interface another approach is to obtain a modular low-maintenance software architecture, which is supported by the use of the middleware MiRPA. The Figure shows the current control architecture used for the realization of this open concept, which supports the embedding of different kinds of sensor systems in a very flexible manner.

To solve complex robot tasks, it is possible to create skill primitive nets. The efficiency of such skill primitive nets is presented in some videos. The well-known robot tasks object placing, bayonet bulb insertion, object alignment in an L-shaped corner and insertion of a battery into a cell phone have been realized there. The combination of robot tasks allows the solving of very complex assembly tasks in an easy way, while all control details remain behind the programming interface.

Software architecture for a Skill-Primitve-based robot controller.

Further information may be found in our Publications. In case of any further questions please contact Daniel Kubus.

It took 0.24s to generate this page.