Machine-vision smart cameras for a Schneider Electric smart factory
Of all the benefits delivered by machine vision for manufacturing, quality control is high among them. With smart cameras to oversee hardware manufacturing processes that get more nuanced by the day, large-scale manufacturing sites – like Schneider Electric’s own Plovdiv, Bulgaria-based Smart Factory – can better capitalize on innovations in their sector.
The ongoing trend of the miniaturization of electronic components enables manufacturers like Schneider Electric to build smaller, lighter, faster, and more efficient devices and systems. Throughout Schneider Electric's Smart Factory in Plovdiv, Bulgaria, the smallest components often play a large role in determining the effectiveness and quality of more than 4 million products per month. Situated across 12 acres, the factory is a vital manufacturing hub in Europe, supplying seven distribution centers in over 28 countries with electrical distribution and protection products for people and businesses.
But with miniaturized electrical components come various manufacturing challenges. For example, soldering is a very complicated process that requires precision. So many factors ― including temperature and imprecise placement of components ― can negatively impact the final quality of a solder joint in an electronics device.
In the end, a solder defect (such as a crack, void, or weak bond) can impact the performance of a product, resulting in customer dissatisfaction which can of course prove costly.
Manufacturers can observe, measure, and evaluate solders to identify defects through soldering inspection. It’s typical for soldering inspection systems to use high-resolution cameras and machine vision to scan through electronics components, enabling manufacturers to see the areas surrounding solder joints and ensure quality.
However, according to Martin Yordanov, Method and Maintenance Manager at Schneider Electric Plovdiv Plant, conventional machine vision has had limited success considering the level of inspection they aspire to at the Plovdiv plant. So, when the time came to find a partner that could provide a solution that could improve the soldering inspection process, the Schneider Electric Plovdiv Plant turned to its own ecosystem of partners on Schneider Electric Exchange.
Communities such as Exchange can help facilitate the types of partnerships that can tackle industry issues through co-innovation. Together, collaborative partners come up with joint solutions that solve tough business challenges. Additionally, users of Exchange can capitalize on EcoStruxure-compatible IoT innovation for automation and energy management.
"From Exchange, we communicate with our partners in a very reliable and effective way. We share ideas and best practices in our area," says Yordanov, of the Plovdiv plant. "We decided to work with Cognex because they are a well-known partner," he adds.
For Cognex, it all began in 1981 in Boston, and today they specialize in machine vision. For the last 20 years, Cognex has supplied machine vision and traceability systems to Schneider Electric plants worldwide, according to Christophe Denis, global account manager at Cognex.
"It's the only type of automation product that we design and produce. Compared to some other automation solutions providers, we only do machine vision," Denis says.
Over the past five years, Cognex has taken an extra step in the relationship with Schneider Electric, joining its Technology Partner Program and promoting their products on the Schneider Electric Exchange Marketplace.
A central solution to improve production
A primary challenge at the Plovdiv factory revolved around automation, according to Yordanov.
"It was important that, together with Schneider Electric applications and Cognex, we created a central solution that helps the effective production in the Plovdiv factory," he explains. “Cognex skillfully developed solutions applicable to EcoStruxure applications.”
Since Cognex smart cameras can be embedded in a large variety of industrial protocols, they can communicate with any type of PLC or external system, according to Denis. "It's a very versatile tool that we can integrate into any automation environment.”
"With Cognex, we managed to develop a few solutions in our process, that helped us produce better quality products," says Yordanov. He uses as an example Cognex In-Vision D900, which helped the plant improve the management of the soldering process. In-Vision D900 is part of the Cognex smart camera product range.
"All the intelligence is embedded inside the camera, meaning that it doesn't need any external computer or external controller to work on the system," says Denis. "With deep learning technology embedded, it also provides a comprehensive range of conventional machine vision algorithms."
Denis explains that deep learning solutions in their smart camera offerings learn by example.
"A neural network is a program that trains by providing some images or examples of what is a good part and what is the bad part. And based on that experience, then the system can automatically separate good parts from bad parts," he says.
"With the development of deep learning inspection solutions, the applications at the plant are much easier to solve with very accurate results," says Yordanov regarding the Plovdiv plant. "It helps us deliver quality products to our customers on time.”
“The barcode system we use, developed by Cognex, and the solution they provided us through In-Sight D900 allowed us to be maximally effective in our process,” he says.
Yordanov explains that with EcoStruxure, the plant can effectively track the process, and the Cognex solution makes it possible to improve the production of quality products for clients.
"The ultimate winner, in the end, is the client because they will receive quality products on time," says Yordanov.
According to Denis, Cognex solutions can much more accurately detect defects, helping to reduce scrap on the production lines and the amount of repetitive work during operation.
In the end, organizations can save energy, reduce consumption of raw materials, and improve the workforce's productivity, explains Denis.
By improving visibility into production lines and automating the inspection process, the factory managed to raise clients' satisfaction with ready-to-use products and minimal losses throughout the manufacturing process, according to Yordanov.
"We expect to save 40,000 euro and we lowered our waste to insignificant levels," he says.
Through the solid partnerships between Cognex and Schneider Electric, the projected return on investments at the Plovdiv factory will be less than 2 years, according to Yordanov.
A story of effective collaboration built on trust
For the Plovdiv plant project, as in many use cases, Cognex worked with a local partner system integrator.
"That's one of the strengths of Cognex is that we have more than 400 certified partners worldwide, to integrate our solutions and which provide local services to our customers," says Denis.
"We work directly with our customers to promote our solutions and define the right solutions that will be integrated by our partners―that's the way we work at Cognex," he adds.
The use of the Cognex solutions at the Plovdiv factory and other Schneider Electric factories demonstrates the high level of trust between Schneider Electric and its technology partners, according to Denis.
As for where this partnership is headed, he sees two axes of co-innovation and development.
"The first one is with the factories, to develop the global partnerships we started years ago," he says. He highlights the growing adoption of deep learning and AI to drive inspections and traceability solutions.
His second point is about developing joint solutions with Schneider Electric...
"Since we have expertise in machine vision at Cognex, and Schneider Electric has expertise in data analytics from industrial processes, like with EcoStruxure Machine Advisor solution, there is great potential for collaboration," Denis says. He explains that merging different solutions and expertise into a common approach will provide added value to customers.
"This new product embeds deep learning technology, but differently. It is easy to use and deploy and doesn't require any specific skill to set up, providing the power of deep learning to any user," he says.
In addition to solutions provided by Schneider Electric Technology Partner Cognex, a community for data center operators on Schneider Electric Exchange can help people find answers that can address significant challenges and collaborate on building innovative, reliable manufacturing solutions.
So, are you ready to collaborate in solving some of the most pressing challenges in manufacturing? Register on Exchange today.