Quality Magazine
Vision & Sensors - Best Practices

Evolving At Its Best

The need for best practices within the vision technology industry has been addressed over time through system integration certification and interoperability standards.

September 11, 2013

best practices man woman working factoryThe need for best practices within the vision systems industry arose almost as soon as end-of-use companies began implementing vision systems technology in their operations. While these vision systems would ultimately help augment companies’ quality and output, they often struggled with the task of successfully integrating it into their businesses.

“Companies that do all of their integration with in-house personnel discovered that their teams needed more training to successfully install machine vision or even simply to keep up on the rapid technology advances in the field,” says Dana Whalls, vice president at the Automated Imaging Association (AIA).  So they turned to system integrator companies for their expertise. Unfortunately, that presented a new set of obstacles.

Finding the right system integrator companies with vision system experience and competency was the foremost challenge. “In some cases there were stories of companies who started a project only to discover that the integration company did not have the skill set necessary to complete the project,” Whalls explains. “They would have to stop mid-program and find another expert to come in and complete the work.”

Other problems cropped up. Some companies would hire integrators to install systems, only to realize that the systems did not work (or work as expected) at the end of projects. There also were rumors of systems integrator companies—some of whom were brilliantly capable in engineering--folding for financial reasons. “If you happened to be in the middle of a project, that could set you back to square one,” Whalls says.

While competent and successful system integration companies did exist at this time, many couldn’t escape the growing sentiment of the time that vision systems were difficult, costly and not successful.

“The bottom line was that vision technology and system integrators’ reputations were suffering, which subsequently hurt investment in vision,” Whalls notes.

Certification Saves the Day

Both end user companies and system integrators needed reliable practices to level the playing field and help navigate the new world of vision systems technology.

“The end users knew that vision systems would help them increase their quality and productivity and they wanted to increase their usage of the technology,” Whalls says. “But they were having difficulty with understanding which system integrator companies had vision system experience and competency. At the same time, system integrator companies wanted to disassociate themselves from the negative rumors and continue their growth and success.”

In 2009, AIA addressed these needs with the System Integrator Certification. The program was designed to help raise the vision system engineers’ skill levels and to create best practice models for system integrator companies. The result, Whalls says, was more successful installations of vision systems, and more faith and eventually investment in vision technology.

The following year, the AIA realized that educating system integration engineers would benefit companies with in-house integrators and system integration companies, so it launched the individual training portion of the certification program: the Certified Vision Professional program. AIA began teaching at two levels, CVP-Basic and CVP-Advanced.

“The course criteria and content were developed with a core group of leading industry professionals and the classes are taught by experienced vision industry professionals,” Whalls notes. “The fact that our instructors know how to answer questions from a relevant, real-world perspective has been very useful to the attendees.”

The Need for Interoperability

The demand for best practices in the vision industry did not stop at system certification. A certain amount of product uniformity was also called for.  Eventually, standards development broadened the market and furthered the use of the technology.

“When vision and imaging technologies were first developed, individual manufacturers produced a multitude of proprietary machine vision products,” explains Bob McCurrach, AIA director of Vision Standards. “Vision buyers were bound to specific systems – X brand camera worked only with X brand frame grabber and X brand software. When a user’s needs changed slightly, chances were that the complete technology solution then needed to change, resulting in higher costs.

“The flexibility to upgrade components in your system with another brand was also missing,” McCurrach continues.”For the manufacturers, their investment in R&D to develop new proprietary components was high and their new product time-to-market was slow.”

Groups of vision and imaging companies collaborated to jointly create camera-to-computer system interface standards that facilitated interoperability between separate manufacturers’ hardware and software. “This gave the technology a universal starting point that allowed lower cost solutions with broader availability of state-of-the-art interoperable products,” McCurrach notes.

AIA published the first true machine vision standard, Camera Link, in 2000. “Camera Link defined its own connector and cable and required a frame grabber in the PC to process image data,” McCurrach says. “Since then, there have been a number of different standards developed that take advantage of the interfaces that are readily available on most PCs. Each has its own strengths and depending on the application, one standard may be more suitable than another.”

According to McCurrach, the following factors that help determine which standard is best for a given application:

  • Bandwidth - How much data is needed and how fast?
  • Cable Length – How far can the PC be from the camera?
  • Number of Cameras – Does the application require multiple cameras to be networked together?
  • Standard features such as speed of triggering and camera control
  • System Cost

 AIA has since developed a bevy of vision and imaging standards: Camera Link, GigE Vision, Camera Link HS and USB3 Vision. In addition, the Japan Industrial Imaging Association’s (JIIA) established CoaXPress, which will continue to be refined to address evolving industry needs.