Quality or ‘quality’?
The only truly effective way to solve the problem of complaints about defective products is to eliminate them before they enter the market.
This seemingly simple goal becomes extremely difficult to achieve if humans are in charge, as they are the weakest link in the quality control process of industrial production.
Constantly repeated activities quickly wear us out, generating costly errors, and even the most conscientious employees will eventually succumb to their human nature. The effectiveness of a statistical in-line inspector, on average, does not exceed 64% and, of course, decreases with time pressure.
A separate issue is the speed of such control, which does not even approach the level of machine detection, reaching above 30 items per second.
Therefore, the most effective way to achieve 100% factory defect detection, is to eliminate the human factor from the entire process.
Machine vision is modern technology's answer to the question of how to do this.
An eye that never sleeps.
Like us, machine vision looks and analyses, but it does so in a very different way. The system usually consists of a camera or scanner, specialised optics, and a suitably selected light source.
The image of each component to be inspected is analysed by a dedicated algorithm, which parameterises its features by comparing them with values specified by the user.
The NOK is eliminated from the production line and the whole process is visualised in the user interface, making it possible to analyse the system's operation in real time, adjust parameters and tolerance, and generate inspection reports.
This is particularly important as it allows unjustified complaints to be rejected.
Indeed, the system, once properly implemented, works autonomously
24/7 and its efficiency increases with the number of inspections carried out, reaching full effectiveness in a short period of time.