AI technologies for real-time product analysis on the production line.
Machine vision is an artificial intelligence (AI) technology. It uses machines, rather than the human eye, to take product measurements and to make judgments in real-time based on the information collected directly from your production line. The image captured by a machine vision camera is transferred to an image processing system. It's here that information captured from the image is analysed and reported to a human operator.
A significant advantage of machine vision is its ability to operate at line speed. Manufacturers can vary the product inspection rate depending on the required quality level. Facteon machine vision systems can operate in dangerous environments or tight spaces.
A commonly deployed alternative to machine vision is manual quality inspection which requires the human eye to inspect the product. Your human operators then decide whether the product meets pre-set quality requirements.
Facteon is able to integrate robust vision systems with smart manufacturing and AI software solutions that provide analysis in real-time.
Machine vision applications
Machine Vision For Defect Detection
Facteon has developed a smart weld monitoring software solution. It utilises the latest computer vision technologies to enable manufacturers to perform superior weld quality inspection. A manufacturer can assess the quality of every weld delivered, as opposed to a small sample. This allows manufacturers to act on quality issues in real-time improving weld quality, reducing waste and ensuring compliance with ISO standards.
Machine Vision For Bin Picking
Decrease labour costs, improve worker safety and increase production capacity by introducing machine vision for bin picking. This technology uses a camera to continuously feed information to a robotic system for bin picking. Similar to the human eye, the robot notes the exact position of each object in the bin. The major benefit drawn by the manufacturer is the capability of the system to assess the position of parts and process this information in real-time.
Machine Vision For Pattern Matching
Feature recognition utilises an image processing method to extract the characteristic contours of the image. The software determines whether the image input meets a pre-determined standard. This technique can be calibrated to handle a range of applications, such as classification or product appearance inspection.
The alternative of deep learning technologies on the market require many samples to train a neural network. Feature recognition is more economical and effective in comparison with other deep learning technologies when deployed for the matching of simple patterns.
Integrating Machine Vision With IIoT Technologies For Enhanced Production Quality
Often, a manufacturer's ability to integrate with modern Industrial IoT (IIoT) solutions is limited by the manual nature of the inspection process. When a robotic solution is deployed, it can be integrated with IIoT software and hardware with ease. This allows for the assessment of production trends in real-time, empowering your team to make decisions based on actual production information, rather than historic data or a hunch. Learn more about IIoT software and hardware integration for quality inspection here