Integrated layout and path optimization of manufacturing cells

• Applicant: Professor Dr.-Ing. Bernd Kuhlenkötter
• Subject Area: Production Automation and Assembly Technology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 537603255

To facilitate the planning of automated industrial manufacturing cells and robot-supported production plants, simulations are increasingly being created in virtual models of real systems to plan and optimize the cell layout, program sequences, or robot paths offline even before the plant is commissioned. In the virtual model of the entire plant, cycle times can be determined, reachability checked, and robot movements simulated to operate the real plant as efficiently as possible later. To set up such a simulation, one first needs an existing cell layout planned by the users and is now manually optimized based on the simulation. However, this procedure is time-consuming and requires expert knowledge to find the best possible solutions. Given the continuously rising costs for skilled workers in automation technology and the general shortage of skilled workers, a methodology is being researched to enable automated layout and path planning of manufacturing cells for industrial robots. In this way, end users without expert knowledge can find optimal solutions for their planning problems in minimal time. The project aims to develop generic planning procedures that can be adapted for specific applications and are thus suitable for a wide range of cell and robot configurations. In contrast to existing optimization approaches, the degrees of freedom and restrictions from cell layout and process are to be considered in parallel. The methods are to be implemented in an open-source library and tested experimentally. The data used, and the resulting prototypical simulation framework will be published extensively on an open-source platform both during the project and upon completion. The basic procedure in the project ranges from the acquisition of resources to the parameterization of these and the delimitation of a solution space to the formulation of an objective function and multi-objective optimization of a solution for a cell layout and the robot trajectory. The applicability of different optimization methods, e.g., evolutionary algorithms or gradient-based methods, will be investigated during the prototypical implementation. A unique feature is the final validation of the achieved project results, based on exemplary projects and demonstrators at the LPS and in the research building ZESS. Digital images of the robots and components used are used, and the results of automatic optimization are compared with those of manual optimization by experts.

• DFG Programme: Research Grants