The global site of the UK's leading magazine for automation, motion engineering and power transmission
19 March, 2024

LinkedIn
Twitter
Twitter link

AI-based defect detection works like human inspectors

20 July, 2020

Omron has announced a “defect detection” machine vision system which, it claims, is the first that can identify defects without learning samples. The system uses AI (artificial intelligence) technology to replicate the techniques of skilled inspectors to detect defects that were previously difficult to spot, thus automating visual inspections.

Omron points out that skilled inspectors are hard to find, and their costs have risen sharply. Manufacturers are facing intense pressure to automate processes that have previously relied on the senses of experienced human workers. An added factor is the need to avoid people having to work close together in manufacturing sites to protect them from Covid-19, which is increasing the demand for labour-saving automated visual inspection techniques.

Such techniques need to be able to identify subtle defects, even on flexible lines producing a wide range of items.

To solve these challenges, Omron has developed what it describes as the industry's first defect detection AI that reproduces the techniques of skilled inspectors. It says that AI is now reaching the stage where it can recognise object features as reliably as humans and learn criteria automatically.

The company has incorporated more than 30 years of experience of image processing and visual inspection in the new FH Series vision system that uses AI to enhance inspection capabilities without needing to learn large amounts of data. Dedicated AI engineers are not needed to set up and adjust the system.

A new AI-based image filter replicates the techniques that skilled inspectors use to identify defects on any product background. Omron claims that scratches and blemishes that were once difficult to spot – on sandblasted metal or resin products, for example – can now be identified without using samples or making any adjustments.

Omron's AI-based defect detection technology runs on its existing FH series machine vision platform

An AI-based “fine matching” tool learns from images of non-defective products to acquire the “expertise” that inspectors can take many years to develop. It can determine acceptable variation tolerances, cutting costs and boosting productivity.

AI has traditionally required a high-end environment, but Omron has integrated the new inspection functions into its existing FH Series hardware. No special AI hardware or experts are needed.

The company says that the new AI vision system gets close to human sensitivity and is easy to use. It adds that the technology will free people from simple, monotonous tasks, allowing them to become more engaged in creative activities that advance innovation in manufacturing.




Magazine
  • To view a digital copy of the latest issue of Drives & Controls, click here.

    To visit the digital library of past issues, click here

    To subscribe to the magazine, click here

     

Poll

"Do you think that robots create or destroy jobs?"

Newsletter
Newsletter

Events

Most Read Articles