DLR (CC-BY 3.0).
Nowadays, in a lot of areas of production technology, industrial robots are an important component for an efficient automation. However, they are employed mainly for the production of high lot sizes of low variability. In contrast, smart factories need the flexibility and adaptability of robots in order to be able to produce costumer-specific products with high variability in small lot sizes. This requires new approaches and methods for developing software. In the methodology proposed in this project a new continuous methodology for developing software for multifunctional robot cells is developed. This is an important aspect of a smart factory in order to handle small lot sizes and costumer-tailored products efficiently as demanded in the context of Industry 4.0.
The main focus of the project lies on automation software for multifunctional robot teams and their methodology of developing industrial production processes. The superordinate goal is to investigate and to test a tool-based methodology for the control software of dynamically forming, multifunctional robot teams. The test cases are the laying of a half CFRP aircraft fuselage and the assembly of real furniture.
The main idea of the project is to decompose the building plan of the product under consideration into smaller subgoals and to define constraints and additional conditions on these subgoals. For example, if one want to build the lay-up of an aircraft fuselage in CFRP one has to respect the order of the layers but still has freedom of design within one single layer. Subsequently, the single subgoals are assigned to potential robot teams, and afterward, the process is optimised by incremental and iterative algorithms. These assignments and optimisations take place independently of the production domain, i.e., after a suitable abstraction, a CFRP building plan is processed by the same algorithm as the construction plan of a piece of furniture.
This project forms a collaboration with the Institute for Software and Systems Engineering (ISSE) of the University of Augsburg. Each partner brings its own specific competences: the ISSE has a large wealth of experience concerning robotic planning algorithms whereas the ZLP brings its knowledge about structural design and their decomposition in smaller tasks, especially in the area of production of CFRP, into the project. Moreover, the ZLP can already point to a bunch of preliminary work concerning cooperating robots and robot teams.