Cognition-enhanced, Self-optimizing Assembly Systems
| Vision |
Self-optimizing socio-technical assembly systems within a production network which autonomously define, reach and maintain optimal operating points based on cybernetic models |
| Objectives phase 2 |
(1) Design methodology for self-optimizing assembly systems in a production network |
In the subproject "Cognition-enhanced, Self-Optimising Assembly Systems", self-optimizing assembly systems are being developed, in order to realize greater flexibility in modern automated assembly processes. The goal is a fully automated economic assembly, for customized individual products or complex products with small lot sizes. Therefore, two industrial applications in the field of optical assembly and aerospace assembly are used for developing generally valid methods and control concepts for the configuration of self-optimizing assembly systems.
Retrospective, research in the field of model based self-optimizing optic assembly focused on the integration of the measurement system as well as the processing of the measurement data to meet the requirements needed for the feedback to the optical model (simulated by using the ray-tracing software Zemax©). To enable a self-optimizing process, it is essential to map the measurement data in compliance to the DIN standard, otherwise a sustainable change of the target set is impossible. Another research topic was the simulation of component- and assembly tolerances. In case of a successive build-up of an optical system, the question arises in which manner the assembly sequence influences the result of the assembly process. Therefore, a simulation method has been developed to determine the most advantageous assembly sequence in a self-optimizing assembly process.
Further emphasis is the development of (semi-)autonomous extension modules for the laser to be assembled, which results in a more efficient alignment of the components and, at the same time, to a higher functionality of the assembled laser. In detail, we studied the automated alignment of the outcoupling mirror of a solid-state-laser. On the one hand, this extension module enables a function orientated alignment and, on the other hand, a compensation of aging and wearing over the whole life-cycle.
Main focus in field of aerospace assembly rests in the development of a method for the structural mechanical modeling of the component’s deformation behavior. The model of the component deformation behavior is used for analyzing the detected deformations during the process in order to calculate compensation forces and movements of the handling system. The depicted method for modeling the deformation behavior is based on the Matrix Structural Analysis (MSA). The MSA uses beam theory for modeling mechanical structures. In the use case an airplane shell element equipped with stiffening elements (stringer, frames) was modeled by the MSA method.
In addition to modeling the component behavior, the demonstrator for self-optimizing aircraft shell assembly has been extended in order to stabilize and reshape the entire mounting area of one frame (stiffening element) simultaneously. Therefore three industrial robots were vertically arranged on linear axis (see fig.?). With this arrangement the robots can apply defined forces calculated by the process control in order to compensate geometric deformations in the airplane structure component.
In the area of the extension modules, we plan to extend the algorithm to align the out coupling mirror with regards to self-optimization. In the field of function orientated assembly, the individually developed process steps are to be merged to obtain a complete assembly process. In the field of aerospace assembly the demonstrator destroyed by the fire at the WZL in January 2016 has been rebuilt. On this new demonstrator the process control based on the structural mechanical MSA model will be implemented and validated.
Find out more:
Technical Demonstrator "Self-optimizing Assembly of Large Components"
Technical Demonstrator "Self-optimizing Assembly of Laser Systems"