Committed to progress
TechStories made by stoba: Seamless quality management with pioneering AI-based quality inspection
Manufacturers and buyers expect flawless products. Procedures for seamless quality management are therefore essential in every industry. Even the smallest defects, such as scratches, pores, dents, blowholes or impact marks in the surface can therefore lead to expensive complaints, safety defects, loss of function and even functional failures. If these are only detected later in the production process, they cause high costs. Complex geometric or safety- or function-critical components must also be manufactured reliably with consistently high quality in order to ensure that the company’s reputation is not damaged.
Higher quality thanks to automatic optical inspection
The inspection of very small components with different surface characteristics places high demands on the inspection process used: They should be highly productive and flexible with maximum quality, as well as inspect cost-efficiently and without interference. However, complex object geometries and a wide variety of material properties are increasingly placing limits on conventional optical inspection methods: If, for example, the surface structure of a test piece is irregular or its reflective characteristics fluctuate greatly, it is often no longer possible to reliably detect and evaluate defects. This increases the risk of defects on the one hand and the proportion of pseudo rejects in quality inspection on the other. For the manufacturer, additional downstream manual inspections result in additional costs that could be avoided with optimized systems. Advanced vision systems work quickly, precisely and objectively: based on optimized technologies, they provide exact information regardless of the surface quality of the test object – even with high inspection throughput and short throughput times.
Automated optical quality inspection with artificial intelligence from stoba
stoba Customized Machinery’s InspectorONE is a self-learning machine for the visual inspection of components and products of all kinds. Equipped with cameras and deep learning software, the device inspects components and products quickly and automatically. With deep learning algorithms, even the smallest defects on heterogeneous or reflective surfaces that human inspectors often overlook can be found with consistently reliable quality. In addition, powerful insights are provided that contribute to increasing quality and operational efficiency. After the training phase, in which real data and synthetic data are collected simultaneously, InspectorONE can be used for visual inspection and make decisions independently. “The benefit for our customers is that InspectorOne ensures product reliability and quality using state-of-the-art methods. This plays a decisive role, especially in applications such as in the automotive sector. In addition to quality assurance, economic aspects are also of central importance,” says Muhammer Kör, Head of Sales at stoba Customized Machinery.
Qualified defect detection reduces the number of rejects
The system’s inherent tolerance limit for the ejection of deviating components can be adapted to the specific requirements and, if possible, not defined too narrowly. Regardless of whether it concerns surface defects such as scratches, contamination such as chips on sealing surfaces, particles in liquids, air bubbles in the material or the completeness of assemblies, e.g. circuit boards. It also supports the simultaneous inspection of different components on one machine, as well as applications in clean rooms. “InspectorONE goes one step further and detects deviations of all kinds, can understand where the defects come from and predict possible future defects using AI,” says Michael Berkner, Sales Expert.