Machine Vision Integrators Teach the Science of Credibility

Start with the Black Arts
Machine vision has long been described as a “black art,” ironically because of its dependence on light (so shouldn’t it be a “light art?”). Despite massive improvements in machine vision software — thanks in large part to cheaper, more powerful computers — no object-based program or set-up wizards will be able to solve every lighting challenge. “People have always said machine vision lighting and optics are a black art,” says Brian Durand, president of machine vision integrator i4 Solutions (Mendota Heights, Minnesota), an AIA Certified System Integrator (CSI). “Even though machine vision software has become more forgiving, a successful system still depends on the successful pairing of lights, optics, and cameras. You can’t just read a manual and come out knowing all you need to know.”

Durand suggests that customers start with the basics when searching for a machine vision integrator. Visit the integrator’s facility and meet the team, and not just the sales rep. “Is the engineer a PLC programmer who does vision on the side?” asks Durand. “Or do they work exclusively on machine vision? Check out their inventory to make sure they’re not a one-trick outfit. Deep machine vision inventory also speaks to the financial health of the integrator, which is another proof of value. Look at their past experience. Have they done jobs in your industry? And do they have proven solutions for problems common to that industry?”

More Than a Vision Expert
According to advanced-level Certified Vision Professional (CVP) Nicholas Tebeau, manager, Vision Solutions, Business Unit Industrial Solutions at LEONI Engineering Products and Services, Inc. (Lake Orion, Michigan), projects generally come in two types: machine builders and customers who just need a vision solution, and retrofit projects that can demand much more.

“If you’re just adding vision, the machine builder can treat you as an ancillary system,” Tebeau explains. “You can just focus on the machine vision solution. However, for retrofit projects, you may be responsible for the full capability of the entire solution. If it’s a robot-guidance retrofit, you need to know how to control and program the robot as well as PLCs, and even design the electrical panels and mechanical tooling. While we may not manufacture the mechanical fixtures, we often have to design those elements here. Even the vision system can demand specialty knowledge. It’s one thing to put a vision system on a slow-moving line, but if you’re installing a dual-line camera running in the tens of thousands of lines per second, you’re going to need to know a lot about which frame grabbers can handle all that data, how to handle the timing, how to get sufficient light on the object, etc. It’s not a plug-and-play solution.”

Even when the customer seeks vision only, it’s not uncommon for machine vision integrators to get involved in other automation systems related to the vision workcell. “We often get involved in designing the mechanical mounts for the cameras and system enclosures as well,” says i4 Solutions’ Durand. “More and more, it’s easier for us to take responsibility for tracking parts down to a reject location, for example. It can be a nightmare for a customer to depend on a PLC programmer to track parts from a vision inspection station to a reject station when it's not directly related to the new machine they’re designing.”

In response, when hiring new machine vision integrators, Durand often looks for people who not only understand machine vision technology but also have experience in key application areas such as packaging equipment.

AIA, the global machine vision trade association, is helping to provide objective third-party guidance for machine vision integrators and customers searching for qualified vision professionals through its Certified System Integrator (CSI) and Certified Vision Professional (CVP) programs.

“The continued growth of general-purpose machine vision depends upon having trained and qualified engineers and technicians to competently implement this technology,” says David Dechow, CVP instructor and machine vision staff engineer with FANUC America Corporation (Rochester Hills, Michigan). “The AIA has always been a key resource in training for machine vision, particularly in targeted conferences. I’ve taught classes in AIA conferences for the past 16 or 17 years. The relatively new Certified Vision Professional program is taking great strides in standardizing that training, and in quantifying results by awarding a professional status to attendees who pass a rigorous test. This program, as it matures, will be critical to the machine vision industry.”

According to Durand, whose company is one of a dozen or so to achieve CSI status, the new certification programs are a good way for integrators to help customers feel at ease with a new machine vision integrator partner. “The program is still relatively new,” he says. “While few of our customers know what CSI or CVP status entails, they do recognize GigE Vision and other standards put out by the AIA. So when we explain that CSI and CVP are from the same organization, that gives them a feeling of confidence. There’s still a lot of public education and evangelizing that we need to do to spread the word about machine vision certifications, but every machine vision integrator needs to establish credibility, and CSI and CVP are good ways to start.”

Unlike selling a smart camera, which is easy to quantify in terms of performance, identifying the best machine vision integrators can be a challenge, even for veterans in the industry. But by using good sense, and leveraging a growing body of standards and certifications, newbies to the machine vision market can quickly winnow their list of prospects, keeping the wheat and leaving the chaff.