How AI just made label placement better

It was only a matter of time before artificial intelligence (AI) worked its way into the industrial packaging world. But who would have thought it was here already? Not me. But, it is.

Earlier this year, Cognex sent along some information about a system for ensuring proper label placement on cartons. At first, AI didn’t exactly leap out. But digging into Cognex’ label placement guidance and verification system shows a decided AI presence.

In short, it’s all about “optimal label placement.” Who’s not in favor of that? And as we all know and Cognex points out, any carton can have a range of labels on it from case marking and hazmat to branding including logos. Given that landscape, there’s got to be a best place to apply that shipping label. And, there is.

Now, AI is not all on its own here. The Cognex system combines 2D and 3D vision systems with AI. Quite simply, the idea is to capture a high-resolution, high-quality image of the carton and its surfaces.

“These advanced technologies work together to generate the best location candidate for placement of a new label. This might be a ‘clear’ spot on the box, or a custom-defined location,” explains Cognex.

Then came this: “The deep learning (AI) tools come pre-trained, making out-of-the-box (system) installation possible.”

Doesn’t sound like much needs to be done to make AI operational here. That’s got to be a positive. But what exactly is deep learning?

Cognex can help you out. In short, deep learning, the company says, is AI for applications “that are too difficult or time consuming for rule-based machine visions systems.” It goes on to say, traditional deep learning enables users to analyze extremely large image sets (collected in this case by 2D and 3D cameras) quickly and efficiently.

Translation: Deep learning with this label system enables rapid decision making on the line, determining placement almost without hesitation despite the uniqueness of each package’s label placement options.

By the way, Cognex also gets into edge learning. But we’ve probably gone deep enough on AI for one day.

Which leaves us with two interesting questions: How important is AI in packaging (not to mention elsewhere in the supply chain) going forward? And how frenzied should we get about AI here?

Honestly, it’s way too early to offer an answer to the first. As to the second, cool your jets.

While Wall Street has nearly lost its mind looking for AI companies to invest in (they barely exist), you can take a more measured pace.

The first step here is to acknowledge that for industrial packaging and the rest of the supply chain, AI is a feature, a product capability. It is not a product of itself. And as a feature, it fits in with other capabilities of a product or system. The Cognex system is a perfect example.

In other words, expect to see increasing references to AI in industrial packaging going forward. And use the technology to your advantage. But, please keep it all in the perspective it deserves.