• Font Size:
  • A
  • A
  • A

Feature Articles

Machine Vision Checks Printed Marks for Next-Generation Supply Chains

by Winn Hardtin, Contributing Editor - AIA

For the past decade, the AIA has chronicled the use of machine vision for checking type and color quality on the printed page because that is where most printing was done – on paper. Today, thanks to government regulations in the pharmaceutical and food and beverage industries, including ePedigrees, and the realization that track-and-trace solutions give manufacturers another productivity tool to increase margins and reduce liability, machine vision print applications have moved beyond the printed page to plastics, metal, and virtually any surface that can bear a mark, code, or character string.

“Putting codes and marks for track-and-trace applications may have started in the pharmaceuticals, food, and automotive industries, but my personal opinion is that track and trace will go everywhere, across every industry,” says Donato Montanari, General Manager, Machine Vision Business Unit, Datalogic Industrial Automation (Bologna, Italy). “The idea is that companies need to track every single component in their supply chain. They’re marking everything, and sometimes that means marking it on the product itself, not just a label or paper. This is an important change in the printing industry, and a very significant opportunity for machine vision, because reading marks is getting more and more difficult. On paper, you’re generally reading black print or barcodes on white paper, which is the best contrast you can get. Today, it’s black on black, or yellow print on yellow material. It’s much more difficult to read.”

Machine vision camera manufacturers, such as SVS-VISTEK (Winchester, Virginia), focus their product development on functions that specifically address the direct-part marking and printing industries’ needs. “Many of the applications we get for printing involve reading various 2D codes and stamped-parts markings,” explains Gerard White, Director of Sales at SVS-VISTEK. “Our customers choose us because we overclock most of our cameras using quad-tap CCDs and dual Gigabit Ethernet outputs to generate some of the highest frame rates of any cameras on the market for a given sensor. And we use sensors specially chosen to provide high dynamic range, such as TrueSense Imaging, which provide 12-bit output. When the mark is stamped or etched in metal, for example, you need a camera with wide dynamic range and on-axis lighting to read the mark on a high-reflectivity part for optimal contrast and capturing surrounding surface detail.”

SVS-VISTEK’s SVCam-ECO camera line, for example, provides 150 fps at VGA resolution, while the higher-resolution SVCam-EVO line offers resolutions up to 8 megapixels (MP) and also up to 150 fps thanks to four separate read-out channels on the TrueSense Imaging CCDs. At the recent Automate 2013 conference, SVS-VISTEK introduced the EVO Tracer, a new micro 4/3rds-inch lens mount camera that is compatible with consumer-grade photography lenses, allowing for total control of focus, iris, and zoom in an industrial-grade camera.

“When you’re sorting parcels based on written codes or text, for example, many vision solutions will move the camera up and down to accommodate packages of different sizes rather than adjust the camera’s focal point,” White says. “That can add $5,000 or more to the cost of the system. By using consumer lenses with piezoelectric focusing motors, we can eliminate the need for motion-control systems to move the camera. The same logic applies to robot guidance applications. Also, the larger aperture size on the 4/3rds mounts allows you to use-high resolution sensors with large pixel sizes, and allows you to run the cameras at higher frame rates while still collecting enough light to generate a good image for OCR [optical character recognition] and other imaging processing.”

Printed labels are both the key and the map to automated parcel sorting applications. “Machine vision systems can not only read the label, but also capture a picture of the condition of the carton before it heads to the customer,” adds Montanari

Standard Products, Specialist Integrators
While SVS-VISTEK’s focus is on hardware, Datalogic Machine Vision combines hardware experience with its software. Having ultimate control over both hardware and software ensures the overall solution performance, according to Montanari. However, when it comes to printing and reading applications across diverse industries, there is one more important piece to a viable solution: an experienced system integrator.

System integration is not always provided by third-party companies. When it comes to large OEMs, in-house engineers with vision expertise do most of the system integration. However, most companies do not have the internal resources or machine vision expertise, and then third-party integrators are the answer. “At Datalogic, we have both internal teams that specialize in supporting large OEMs with their system designs, as well as an extensive network of specialized integrators that we match with customers’ requests,” Montanari notes.

The prevalence of track-and-trace printing applications has grown steadily over time, but according to Montanari, the last year has revealed a surprising momentum. “If you go back a few years ago, only large companies used track and trace. But in the last 12 months,” Montanari says, “we’re experiencing an explosion of vision solutions requests for track-and-trace applications from small and medium companies in almost every industry.”

 

Comments:

There are currently no comments for this article.


Leave a Comment:

All fields are required, but only your name and comment will be visible (email addresses are kept confidential). Comments are moderated and will not appear immediately. Please no link dropping, no keywords or domains as names; do not spam, and please do not advertise.

First Name: *
Last Name: *
Your Email: *
Your Comment:
Please check the box below and respond as instructed.

Search AIA:


Browse by Products:


Browse by Company Type: