Artificial intelligence in
production and manufacturing
can effectively reduce errors in production, save costs and resources and increase efficiency.
Unwanted errors in production can quickly lead to interruptions in operations and increased costs due to later error correction. They can easily be missed within a quality management process because they are so small and inconspicuous that the human eye has no chance of detecting them. If, for example, more than 200 products per minute pass over the treadmill, it is not surprising if deviations from the standard are not detected. This places an unnecessary additional burden on the employees (e.g. controllers), the quality of the products decreases and, in the worst case, it leads to terminated contracts.
With the right method and technology, however, those errors and deviations can be effectively reduced and the burden on staff reduced.
The latest Artificial Intelligence technology enables the complete automation of quality management processes. It detects errors, deviations, and foreign substances and independently classifies and assigns product types. This means that the technology is not only responsible for preventive error measures, but also for the processes within the production by making them simpler and more efficient. It supports the responsible personnel by reporting errors in real-time and displaying them on a management dashboard or sending them as a signal. Among other things, the signal can also be reported directly to a controller or to the control system so that appropriate measures can be taken. The employee is enormously relieved.
The name of the technology is the Blue Quality Control platform, whose core is an Edge-AI device that processes all data locally. The device, together with a set of components (video cameras, sensors, special cameras, lighting), checks each product in real-time for deviations and defects. Depending on the production process, these components can be adapted, because different products have different requirements. For example, a meat product may require a different component for detecting defects than a detailed technical device. Thanks to the modular architecture of Blue Quality Control, however, these adjustments are easy to implement.
The error prevention measure
In the error prevention measure, the product travels on the conveyor belt and is scanned for errors in seconds by the corresponding component (e.g. by a video camera). The smallest deviations, such as cracks or foreign matter, are detected and reported immediately.
Classification of product types
When classifying and assigning product types, e.g. different metal screws, which differ only in the thread (full and partial threads) and heads, the product is recognized and assigned by Edge-AI. A signal sent by the solution ensures that the corresponding product is directed to a selected and classification-compliant treadmill for further processing.
Edge AI means a solution, in which the AI algorithms are installed directly on the device collecting the data. This could be a camera, augmented reality glasses or a dedicated device with any sensors attached. This eliminates the need to transfer data to a server or a cloud, as is the case with traditional methods. One of the advantages of such solutions is fast and secure data processing.
By using the latest edge devices, such as those from the Nvidia Jetson family, it is not only possible to execute advanced AI algorithms, but also to use them in an energy-efficient way.
• Reduced error rate
• Reduced costs
• Increased level of quality
• Relief of employees
• Automation of quality management processes
More Information about Blue Quality Control: Link