The top 5 problems vision inspection can solve and why its future is exciting

Posted by Brought to you by Matthews Australasia on 25th July 2017

Automated vision inspection has application across a host of industries, doing a host of different checks. And because it’s objective, automated and highly repeatable, it can solve many issues. On top of that, the rise of the Industrial Internet of Things (IIoT) means the future for vision is exciting.

Here are the top five problems automated vision inspection can solve.

In this article, we’ll look at how automated vision inspection can solve typical problems that manufacturers face. The top five issues we’re asked about are:

 

  • flaw detection
  • identification and validation
  • verification
  • measurement
  • positioning and robotic guidance
  1. Flaw detection

Detecting flaws is possibly the most fundamental automated vision inspection task. Flaw detection is vital to quality control (QC) because it enables manufacturers across sectors — dairy, food & grocery, beverage, wine, beverage and so on — to find blemishes, scratches, cracks, discoloration, pitting and so forth, through empty container inspection, packaging QA and other types of flaw detection.

Machine vision inspection has four advantages over the human eye: it is much faster, much more accurate, can detect defects that are invisible to humans and can operate around the clock, 365 days a year.

Flaws in products or their containers tend to be random, so vision flaw-detection algorithms look for changes in colour, texture, pattern or for set structures. iQVision uses high-sensitivity, high-resolution image sensors and cameras to capture high quality images, which allows powerful processing software down the imaging chain let the QC system take the action for which it was set up.

 

  1. Identification

Products that are perfect in every other way but have mismatched, improperly positioned or unreadable codes, or no code at all, face the strong potential of being rejected by retailers. Automated vision inspection is an excellent way to verify that a code is present, properly positioned and formed correctly — whether it’s a 1D or 2D code, such as barcode, date or batch code. Vision systems automatically identify and reject items with missing, incorrect or unreadable codes, so only properly coded items leave the factory door. Via system integration, codes can also be validated to check they are correct for that product. It can also look at barcodes outside the carton and match them to the retail units to make sure the right products are in the right carton.

 

  1. Verification

Another common use for machine vision systems is to verify parts, assemblies and packaged goods. Often, verification is combined with other tasks, such as measuring part dimensions or reading product barcodes, to give a full product inspection, inline. Examples of verification uses are for: blister packs, moulded parts, solder joints, bottle cap and safety seals, print, PCB assembly, cable wiring, and for features such as threads, holes, notches and so on. Matching packaging components is another use for vision inspection, where all packaging components (lids, containers, bases, shrink labels, outer packaging) are verified as correct for that product.

 

  1. Measurement

Automated machine vision inspection is also commonly used to accurately check dimensions and tolerances that need to be highly precise, such as fill-level measurement.

 

  1. Pick & place and robotic guidance

Machine vision systems have great application for robotic guidance in enabling pick & place processes so that robots can quickly and accurately distinguish the shape, orientation and position of parts. This is commonly used in identification, sorting and packing operations. It works by the vision system detecting a part using shape-based object identification, then sending that part’s position and angle to the robot. The robot, which picks the part, then places it in the correct orientation. This pick & place process can also include quality assessment, so any parts not meeting the quality criteria are excluded. Pick & place is very useful for highly automated lines where quality and process control are combined.

 

The future

With the right cameras and lighting, automated machine vision inspection provides precise, repeatable QC to ensure manufacturing accuracy. On top of this, the future for vision is exciting because of the Industrial Internet of Things (IIoT). For decades, machine vision technology has demonstrated its ability to inspect, measure, scan, or otherwise identify a variety of products throughout a broad spectrum of applications.

Due to the increasing affordability of machine vision components and systems, a broader array of solutions, more capable hardware, and smarter AI-based software algorithms, machine vision is positioned well for substantial growth within the IIoT movement. You can read more about IIoT here.

 

Check out Matthews’ great resource library. It has a host of detailed information that’s all free to download! There are whitepapers, presentations we’ve done to industry bodies, infographics for manufacturing, case studies, articles from our thought leaders, vids showing solutions in action and more!