AI is the secret weapon for higher productivity

“We’re just seeing the beginning of AI applications in the factory,” says Måns Bjuggren, Director, Strategy and M&A at Mycronic. “In our business, the potential for AI extends to the manufacturing of photomasks, mounting of printed circuit boards and even flat panel displays.”

Less downtime, more revenues

He cites a McKinsey report that suggests AI can improve forecasting accuracy in manufacturing by 10-20 percent, which translates into a 5 percent reduction in inventory costs and a 2-3 percent increase in revenues. Other benefits include predictive maintenance to reduce unplanned downtime.

We expect the AI field to grow explosively as it becomes capable of accomplishing more tasks.

Måns Bjuggren, Director, Strategy & M&A

Learning from the data

“This all rests on our ability to process and analyze the large amounts of data (Big Data) being generated in the factory,” says Romain Roux, Head of Inspection R&D. He explains that Machine Learning, a subset of AI, is already being applied to areas such as automated optical inspection (AOI) to significantly increase first-pass yield, eliminate potential human error and boost productivity. It’s a huge advantage to be able to eliminate defective units before they waste more time, materials and energy.”

Deep learning possibilities

But according to Romain, AI in its purest form, sometimes known as “deep learning”, is an even more advanced methodology that imitates the ways humans gain certain types of knowledge. Deep learning, a subset of machine learning, requires much larger amounts of data storage and complex algorithms. This is a field in which the US and China are racing to take the lead, he notes.

Within SMT assembly, the focus is on real-time 3D AOI and SPI inspection of components using feedback-based machine algorithms and camera images.

Romain Roux, Head of Inspection R&D

Mycronic AI Center in Silicon Valley

“Mycronic is also investing in R&D in this field and coordinating activities with our AI Center in Silicon Valley, with a focus on solving key production challenges across all our divisions,” says Bjuggren. “Within SMT assembly, the focus is on real-time 3D AOI and SPI inspection of components using feedback-based machine algorithms and camera images. AI is being tested together with clients in applications such as high-value boards for aerospace and other industries.”

“We expect the AI field to grow explosively as it becomes capable of accomplishing more tasks” concludes Måns Bjuggren.

Key trends shaping the future

NOTE: This is the fifth of seven articles on key trends shaping the future
of the electronics industry – and how to capitalize on them.