| Research project: | "Autoendoskop" |  |
| Research area: | Automation, inspection, machine learning | |
| Funded by: | Federal Ministry for Economic Affairs and Energy (BMWi) | |
| In collaboration with: | IT Concepts GmbH, Carat Robotic Innovation GmbH, Lehrstuhl für intelligente Materialsysteme Univ. des Saarlandes | |
| Start of the project: | February 2020 | |
| End of the project: | March 2022 |
Description:
The aim of this project is to develop a highly automated, cost-effective and efficient inspection process for the series production of a wide range of products that are characterized by cavities. The system must be designed in such a way that it can be efficiently adapted to the component required by the customer and the system can be set up quickly. The tool to be used is a new type of video endoscope hardware that is to be developed and is to be guided and positioned by an industrial robot. Due to the high variety of components, a modular system needs to be developed from which an endoscope adapted to the respective product can be quickly assembled from various components. The endoscope tip should be equipped with wire actuators made of a shape memory alloy so that active bending of the tip can be achieved. The positioning of the sensors must be derived in advance from CAD data using a simulation that contains both a robot simulation and a simulation of the movement of the endoscope tip as well as a high-resolution sensor simulation. After image acquisition, errors can be found with a high degree of automation. The use of deep learning is intended for this. The IFPT deals with sensor simulation and automated error detection.
Contact person at the institute: M.Sc. Ole Schmedemann