Meet the quality inspector who never gets tired

The fact that factories are replacing human labour with robotics in many cases has been a well-known trend for many years. Less well known, however, is the use of artificial intelligence, especially in the area of quality control.
While humans may seem irreplaceable in quality control, the opposite is true, as evidenced by DataSentics’ Quality Inspector solution under Atos SE. The solution replaces the human eye during inspections and works based on machine learning algorithms. In addition to helping manufacturing companies save money on skilled inspectors today, it also offers long-term attention retention, low error rates, significant speed-ups, and thus increased final quality and reduced costs. Quality Inspector thus helps to keep pace with the acceleration of processes and increasing demands on output quality.

Fewer quality claims, fewer complaints, fewer worries

Minimizing quality claims is a common motivation to optimize controls. This is where Quality Inspector can save the most costs. Thanks to effective controls, the rejection rate is significantly reduced and the company deals with significantly fewer quality claims, repairs and complaints, which also has a positive impact on its reputation. The actual resolution of quality claims is then significantly easier, as all checks are saved, including the defects found and the final decisions of the algorithm. This makes it easy to trace the condition in which the product left the production line.

A single inspector to check all production lines

The Quality Inspector solution finds defects in products using cameras and supporting sensors. Smart algorithms then look for flaws, defects and differences in the images - often in the same way a human would. Like a human, the algorithm also learns to correctly find defects and faults. Its advantage over conventional inspectors is that it uses one system across the entire production line, so it is consistent and can be repeatedly applied to other lines.
While conventional algorithms in similar solutions only use pattern matching, Quality Inspector uses neural networks to adapt to new conditions such as new defect types, tighter tolerances or new production line layouts. Another advantage is a real-time overview of the error rate of the production process.

Tailor-made solutions

Quality Inspector solutions can be applied in a wide range of industries. Interested parties can also try it out in a demo version to see what defects it can detect directly from their production data. DataSentics can then deliver a customized solution tailored to the company’s quality control, which can be integrated on various end devices such as smartphones, handheld devices and scanners, and can operate either as an on-premises solution on an edge server or as a cloud-based solution.