Generalized Methodology for Aerial Manipulation Task Design based on Mechanically Mo- tivated Elementary Flight States

• Applicants: Professor Dr.-Ing. Michael Beitelschmidt; Professor Dr.-Ing. Markus Ryll
• Subject Area: Mechanics
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 546943282

Robots are becoming more prevalent in undertaking tasks that are either dangerous or must be carried out in hard-to-reach locations. A rapidly advancing area of robotics is the use of Unmanned Aerial Systems (UAS) as a platform for manipulation tasks. In this case, the UAS is called Aerial Manipulator (AM) and comes into tactile contact with the target object to perform a manipulation or processing task. Research in the field of aerial manipulation tends to focus on control architectures for specific tasks, rather than exploring the potential of a mechanical problem description. In this project, we aim to develop a framework for aerial manipulation tasks that breaks down complex tasks into mechanically motivated Elementary Flight States (EFS) to facilitate generalization. This decomposition enables the design of complex aerial manipulation tasks and missions. Each EFS has a dedicated flight control for increased precision and system robustness. Tactile navigation is used to achieve a millimeter-level positioning precision of the AM’s End Effector (EE) by utilizing geometric features of the manipulation target. Force/Torque (F/T) sensors are omitted using improved wrench estimation approaches.

• DFG Programme: Research Grants