Realtime Quality-Inspection with 3D-Scanning at Inline-Production-Processes
Acquisition of real-time information of technical systems is becoming more and more important to companies all over the world. One of the biggest beneficiaries of transparent processes can be found amongst companies applying industrial production methodology. According to this demand, a 3d-scanning technology in combination with deep learning algorithms helps to overcome current problems at real-time quality control. The projects aim is to support the current quality control process of a German Tier 1 automotive supplier to become more efficient. The goal is to deliver real-time quality information of the produced parts to employees, management and to further business intelligence systems.
Instead of the conventional procedure by manual inspection, the detection process will automatically review the parts at the assembly line without the need of further support by workers. As a key element of this strategy, an image-based, predictive quality control system will be established at the entire assembly lines. The collected data will support the internal BI (Business Intelligence) with necessary information for further optimization steps. An additional benefit will be the reduction of labour and savings of raw material. Additionally workers will be relieved by less repetitive tasks.
The project targets to support the current quality control process of the Mühlhoff Umformtechnik GmbH with inline 3D-Scanner in combination with a smart connected infrastructure. The goal is to deliver real-time information about current quality of the produced parts to employees, management and to further business intelligence systems. Instead of the previous procedure, the detection process will automatically review every single part of the assembly line without the need of support by workers. An algorithm under use of deep-learning technology will help to review and compare the reconstructed 3D-image data of the scanner.
One of the main targets is to use this sensing technology to support the digital transformation of the company. Goal is to observe current quality aspects in real-time and verify quality aspects of the automotive customers automatically. Additionally, a deep-learning technology provides a prediction about when recalibration/maintenance of the production equipment is necessary. The data will also be used for the business intelligence in order to track quality and ensure the traceability from raw material to the final products.
