Minimize maintenance to maximize machine vision system success

Behind every well-maintained machine vision system is a well-trained staff responsible for its operation and upkeep.

By Winn Hardin, AIA August 2, 2017

Even the best climber in the Tour de France needs to visit the team doctor for a daily checkup. Luckily, machine vision systems don’t need nearly as much attention.

"Machinery in a manufacturing facility needs occasional service and checkups for its health, and vision systems are the same," said Brian Durand, president of machine vision integration company i4 Solutions. "A vision system usually doesn’t require much maintenance, but you can’t just forget about it and trust that it is working at peak efficiency."

To minimize any missteps, machine vision integrators typically take a three-pronged approach: design the most appropriate vision system for the application; thoroughly train the operators and technicians working with the system; and provide ongoing customer support that predicts, detects, and solves any problems that may arise.

Anticipate maintenance needs

Even though maintenance is the final step of a vision system’s design path, it needs to be considered at the very start. "We like to minimize maintenance upfront by design, because every time somebody is touching the system, there are opportunities for something to go wrong," said Tom Brennan, president of Artemis Vision, a machine vision integrator.

That requires due diligence on the part of the integrator. Before deploying a vision system to the customer site, Artemis Vision sets up the system in-house two or three times. "We make sure it is a smooth and documented process so when it is time to install the system, the process runs efficiently," Brennan said. "Having a written record makes maintenance easier and gets the system back up and running quickly if something does need to be replaced."

Developing the correct vision system for an application saves not only on maintenance but customer resources as well. Shelley Fellows, VP of operations for machine vision integrator Radix Inc. cautions her customers against over-designing a system or making specifications too stringent.

"You end up with a large number of false rejects, and in the Lean manufacturing world, excessive waste is a very bad thing," she said. "It’s important to follow the data and make sure that concerning trends aren’t emerging once the system is up and running."

Nearly every integrator has received a maintenance request that was the result of curiosity gone astray. For example, Radix has fielded calls about vision systems not working, only to find that personnel were changing the camera’s algorithms. As a result, the company developed a software application called Quick Compare for their cameras.

"Instead of having someone sit there for a couple of hours trying to figure out the problem, Quick Compare will tell us in 10 seconds where the actual differences are and we can take the software back to the original calculations," Fellows said.

Train like a champion

Behind every well-maintained vision system is a well-trained staff responsible for its operation and upkeep. And behind every well-trained staff is a well-equipped integrator that can instruct in a clear, concise manner. While the personnel responsible for installing a vision system or learning how to train it on new product types likely won’t be the same people operating or maintaining the system, at least in larger manufacturing facilities, good fundamentals at every stage go a long way to prevent future headaches.

"Because we’re delivering custom solutions, the more specific training for the maintenance technicians, plant engineers, and operators, the more successful their vision system will be," Fellows said.

Quality training relies not only on good instructors, but also on good material that is customized to each vision installation. Proper documentation is at the heart of good maintenance policy, and an integrator worth its salt will provide a comprehensive, easy-to-follow operational manual detailing everything from interfaces to troubleshooting techniques.

The mantra behind teaching someone how to maintain and troubleshoot a vision system could easily be: "Assume nothing."

"If they have concerns about the results of their vision system, the first thing we encourage them to look at are live images," Durand said. "If we are able to see what the camera sees, the issue often becomes obvious."

Many integrators agree training isn’t a one-and-done activity. i4 Solutions, for example, offers a service contract that sends a representative to the customer site every month. A key component of the contract, which is geared toward customers with a number of vision systems installed, is ongoing training to address employee turnover.

"We may have trained a customer site on how to use and maintain a vision system two years ago, but the likelihood of those people still being around isn’t very high," Durand said. "It makes sense for our customers to have us on-site on a regular basis, training new people as employees change departments or leave their organization."

In fact, i4 Solutions started the service contract out of necessity. "We saw that our customers weren’t making optimal use of their vision systems, and we knew that training was the answer," Durand says. "Customers acknowledge they have turnover issues and limitations on in-house resources."

The service contract also allows an expert to review the performance of all vision systems. "We can use that data to anticipate a problem before there is any dramatic impact on operations," Durand says.

Maintaining machines and relationships

Persuading a machine vision customer that they need a maintenance support contract can be a tough sell in today’s cost-conscious marketplace. At the end of the day, however, customers want their integrators to be that one-stop shop, their double-redundant backup.

"We have found that customers’ willingness to allow us access through an online meeting platform such as WebEx means that we can see what they’re seeing [without traveling to their site]," Brennan said. "It tends to be something fairly simple that would have required hours on the phone trying to figure out the problem."

As customers lean more heavily on integrators’ expertise in developing machine vision systems, integrators will continue to find creative ways to protect these big-ticket investments.

Winn Hardin is contributing editor, AIA. This article originally appeared on the AIA website. The AIA is a part of the Association for Advancing Automation (A3). A3 is a CFE Media content partner. Edited by Hannah Cox, content specialist, CFE Media, hcox@cfemedia.com.

Original content can be found at www.visiononline.org.