Last week, COO Todd Barnum spoke to students at Rochester Institute of Technology on Photovoltaics Engineering and Economics. Discussing crystal growth and the CZ system, he noted “It has to be cost-effective. The hardest truth to learn in engineering is that you can’t just design the coolest thing imaginable. It has to help a company, a customer, or someone makes more money than it costs, both in capital and operating expenditures.”

Barnum also spoke about AI and how the industry is changing. He likened Linton’s process engineers to everyone’s grandmother cooking in the kitchen, drawing on his own experiences with his 98-year-old Polish grandmother: “She never really measures anything or needs a recipe; it is all in her head. She uses her sight, her taste, smell, and basic intuition to come out with this amazing food—and it has to be in her kitchen with her favorite pan because she knows how each burner on her stove will heat and how the pan will transfer that heat. Our process engineers are similar, and we are challenging them with our vision system to help train the machine to first recognize when things are not what they should be (or the detection phase). Then, we go a little deeper in the learning curve and try to predict what main variables lead up to that detection event, and the last stage will be the prevention phase, where we use the inputs on the variables (add heat, remove heat, pull slower/faster, change rotations/gas flows/magnet strengths, etc..) to try and prevent an error from occurring and so we can create the perfect ingot time after time after time, just like Grandma.”