Automated Visual Quality Control in the Metal Industry

Maddox AI automates and optimises a wide range of visual quality control tasks in the metals industry through precise and reliable AI technology.

Metals go through complex production processes that require sophisticated visual quality control. With varying production batches that can lead to slightly different visual properties, classic, rule-based quality control systems quickly reach their limits and increasingly cause pseudo-rejects. Maddox AI can be used to avoid production variability and thus avoid unnecessary pseudo scrap. In addition, the solution enables companies to produce more efficiently and raise quality standards.

Areas of Application for Maddox AI in the Metal Industry

The metals industry is characterised by a wide range of complex requirements and visual quality control tasks that demand a high level of precision and accuracy. Maddox AI provides customised solutions for the metals industry to ensure these quality standards.

Our AI solution is able to address individual use cases and achieve the highest quality standards even in difficult production environments and at very low cycle times. Unlike conventional, rule-based systems, Maddox AI knows that due to production variations, OK components can sometimes be lighter/darker or more/less glossy without it necessarily being classified as a relevant defect criterion. Thanks to this capability, pseudo-rejects can be drastically reduced without the need for time-consuming manual recalibrations.

Furthermore, every automation project at Maddox AI is always also a digitisation project. The data collected during visual quality control (images, metadata, etc.) are processed in the Maddox AI Cloud software and can be used for subsequent analyses such as defect Pareto analyses, defect heatmaps, dashboards, etc. You receive proactive notifications from Maddox AI (e.g. via email or SMS) when too much scrap is generated. This ensures that excessive scrap does not occur over long periods of time.

Maddox AI Can Be Used to Inspect Various Use Cases, Such As:

Detection of surface defects: Accurately analyse metal parts for scratches, dents, rust, imperfections and other defects to ensure the external quality of metal components, even when production batches have slightly different visual characteristics.
Control of coatings: Monitoring metal coatings for uniformity, adhesion and integrity to ensure a corrosion-resistant surface.
Identification of surface defects: Detection of inclusions, pores, voids and other irregularities with high accuracy, resulting in increased quality and durability of metal parts.
Surface texture control: Detection of various surface textures, from smooth surfaces to textured surfaces.
Analysis of surface markings: Detection of unwanted markings, imprints or dirt on the metal surface to maintain aesthetic quality and keep the final products flawless.

Detection of deformations: Identification of deformations or distortions that may occur during the manufacturing process.
Checking engravings and embossings: Verification of engravings or embossing on metallic surfaces.
Precise component verification: Checking individual components for their exact specifications and ensuring error-free assembly.
Identification of part numbers and embossings: Precise recording of part numbers as well as engravings and embossings on components by means of OCR to enable precise allocation and prevent incorrect orders.
Reading of bar and QR codes: Recognition of codes for accurate identification and verification of components and products.
Traceability of products: Complete traceability of products from manufacturing to the end user to ensure safety and quality throughout the supply chain.

Case Study in the Metal Industry