Visionaries 2024
“Tey want AI machines to gather that know-how before it’s lost, because they believe the next generation won’t have it”
of prey is approaching. We had a camera system looking 360 degrees, able to see up to 500m, and we could even predict if the bird was approaching or just passing by. This saved 100,000 euros. Unfortunately, our partner didn’t want to continue, and since it stopped, the project stopped as well. I think one that’s much cooler is in the house manufacturing sector. In Germany, we build a lot of prefabricated houses based on a wood frame. The walls are built in a highly automated production process, where you can build a house in three days. But the challenge is that each wood frame is unique, designed specifically for the house, and you have to place them with extreme accuracy. If you turn them even by a tenth of a degree, it can disrupt the entire process. We developed a 3D system to ensure that these walls, which are stacked on top of each other, are placed with an accuracy of 0.1mm. This has completely revolutionised the process, but it’s slow to roll out because you need to build a whole new production plant for it. Our customer is a big one, but still, you can only do one or two per year. We have many such projects. One
involves producing startup coins; we’re the only ones able to do quality control within the stamping process. Another project we’re starting is with a startup. They used them a lot for medical products, for coronavirus, when you want to have minus 80 degrees Celsius boxes to transport your vaccine. These boxes are not very ecological because they’re often single-use. Our customer wants to stop that and reuse them. We have a system that controls these boxes. This market is just emerging; it doesn’t even exist yet. We try to be in these emerging, innovative markets to help our customers become market leaders. The problem is we can’t do it twice,
and maybe that’s our fault. Once these customers are paying us, we build a system, and we have an agreement not to go to their competitors. Because we’re already working with these guys and have agreed not to go to competitors, we have to find different markets to do that again.
When dealing with a new end user, how do you deal with expectation management in terms of what a vision system can do? That’s the most difficult thing in our market, and honestly, I don’t have a good answer for that. Customers come with big expectations because they know what their mobile phones can do, such as detecting the difference between cats and dogs. But then they come to the industrial market and don’t realise that these consumer-level technologies aren’t available here. That’s the most difficult thing in our
market, and honestly, I don’t have a good answer for that. I can give you the answer we’re using, but it’s not a perfect one. Customers come with big expectations. They know Google; they know mobile
phones. They say, “My mobile phone can read everything; it can distinguish between cats and dogs. Everything is feasible; it doesn’t cost anything.” Then they come to the industrial market…. First, you have to explain to them that all
these gimmicks don’t exist in our market. When we do something, we ask “do you want to share your data with everybody else?” They say, “No”. So we have to work with data that belongs to a customer, and we can’t take it from everyone else, like Amazon, Google, and Facebook do. They’re giving it for free because they took all the image data from everyone else. If you want to do that on a smaller scale,
on this specified product, we start from scratch. We have the algorithms, but we
> AUGUST/SEPTEMBER 2024 IMAGING AND MACHINE VISION EUROPE 25
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39
