Research objectives and content
Mobile manipulators are robotic systems where manipulator arms are mounted on mobile robotic platforms, thus increasing the workspace of the arm, while, on the same time, endowing the mobile robot with the ability to interact with its environment by tools or sensors mounted on the arm. Systems of this type, which are subject to nonholonomic motion constraints like the ones occuring in many wheeled mobile robots, are of particular interest for practical applications. The presence of these motion constraints makes some promising sensor-based control schemes, like visual servoing, difficult to apply in systems of this type.
However, a nonholonomic visual servoing framework has been recently proposed, the initial development which, is limited to the case of a one degree-of-freedom (d.o.f.) arm and to positioning and stabilization of the system with respect to a static target. We propose to study extensions to multi-d.o.f. arms and to moving targets, as well as related robustness issues, in order to make this method appea- ling to applications in e.g. industrial flexible manufacturing systems.
It is expected that the first eight months of the project will be devoted to the theoretical part of this work and to relevant simulation studies, while the last four months will be devoted to its experimental evaluation. Training content (objectie, benefit and expected impact)
Training objectives include further familiarization with research on the nonlinear stabilization problem, on homogeneous approximations of nonlinear systems and on robust nonlinear control techniques, with computer vision research related to non-metric and qualitative vision, as well as with tools and techniques relevant to real-time implementations of the above. These skills are complementary to our past work on nonlinear geometric control, geometric mechanics and active vision.