Added AI model automatic generation function, etc. to the system for judging product appearance abnormalities: Manufacturing IT-MONOist

On November 15, 2021, Oki Electric Industry combined Google Cloud's "Visual Inspection AI" with the "appearance abnormality judgment system" that saves labor and automates the inspection process with video AI (artificial intelligence), and is a learning model using AI. Announced that it has added automatic generation and autonomous inspection functions.

The appearance abnormality judgment system analyzes the images of the parts and products to be inspected taken by the camera with the AI ​​edge computer "AE2100" and judges the appearance abnormality in real time. Until now, the system had to regularly carry out work to improve the accuracy of the learning model, such as parameter combination and model tuning by AI engineers, in order to deal with fluctuating defects such as scratches and dents. ..

This time, in addition to the conventional image analysis engine, an automatic judgment model function that utilizes machine learning by Google Cloud's Visual Inspection AI and AutoML has been added. This makes it possible to generate learning models without specialized knowledge.

Since the learning model is implemented in the AE2100 edge terminal, it autonomously inspects the appearance at the manufacturing site and immediately notifies the operator of the result. In addition, when an unknown abnormality is detected during operation, if the operator captures the corresponding image and marks the defective part, a learning model can be automatically generated.

The automatic judgment model creation function will be provided from March 2022.

⇒ Click here for other "Manufacturing IT News" articles

Related article

Related Links