Darwin AI Evolving the Islands of Automation

Estimated reading time: 2 minutes

When Canadian artificial intelligence company Darwin AI was founded in 2017, machine learning and deep learning were still relatively new terms. In the past five years, CEO Sheldon Fernandez and his team have been working with this technology to develop some foundational IP to simplify implementation. About a year ago, Sheldon took a “part happenstance, part deliberate” opportunity to develop a vertical offering for EMS manufacturing. Here’s what happened.

Sheldon, it’s nice to meet you. Would you briefly introduce your company?

Sheldon Fernandez: We’re based out of Waterloo, Ontario, Canada, and we're organically connected to the University of Waterloo, which is kind of like Canada's MIT. Two of our co-founders are professors at the institution, including Professor Alexander Wong, Canada’s Research Chair in AI and Medical Imaging. 

We’ve been working on foundational machine learning and deep learning technology for the past five years. A couple of years ago, our large industrial and aerospace clients were telling us about their supply chain challenges during the pandemic and reshoring sensitive electronics manufacturing work back to North America, specifically printed circuit boards (PCB). We thought that created an opportunity for us. 

When we looked at PCB manufacturing, it became apparent that while the SMT placement workflow was highly automated, there was a need in automating back-end production and final assembly. This laborious part of the process was where EMS companies and OEMs were still employing manual inspection. These inspection tasks are tough to crack from a traditional machine vision perspective, and we wondered, “Can AI bring anything to bear on this problem?” We spent about a year developing a hardware and software solution which fits into the typical assembly line for PCB manufacturing. It also does post-assembly analysis, and what's really fascinating is how quickly an operator can program our product. 

We often hear that AOIs are good at what they do but are laborious to program and maintain. With our system, there’s not a lot of manual work. You give the system a good (i.e., golden) board—or a couple of good boards if there's a union of different components—and our AI system creates a map of where components should be in less than a minute; away you go. You can tweak it after that, and it’s striking how quickly you can configure the product. 

We brought our mini system to IPC APEX EXPO, and the response was fascinating. So many companies were intrigued by finally automating back-end production, and we're really excited about becoming a part of the community. 

To read this entire conversation, which appeared in the March 2023 issue of SMT007 Magazine, click here.