Through the Line Podcast: Pallet Adhesives, AI Label Checks, and Freezer Hacks

Hello, everybody, and thank you for joining us for another End of the Line podcast, where all of us PMG editors sit down and talk about the things that have come across our desks this week. I’m Liz Cuneo, editor-in-chief of Healthcare Packaging. Joining me are Derek Teal, editor of ProFood World; Sean Riley from OEM Magazine; and Matt Reynolds from Packaging World.

This week, we are, of course, still digging out from Pack Expo as we cover the show and produce dozens of videos. Matt, you mentioned having some Pack Expo innovations you wanted to share. What have you been seeing? What looks cool?

These are innovations that may have been more cutting-edge years ago. They just never had as much of a reason to be adopted deeply into the industry until recently.

You may remember that this year and late last year, Packaging World has been covering Walmart. At its distribution centers, Walmart takes in single-SKU pallets, breaks them down, and builds mixed pallets designed specifically for one aisle of a Walmart in, say, Topeka, Kansas. If it’s a personal care aisle, it’ll have all the Crest, all the Unilever products, in the right amounts for that specific aisle. So they’re taking in full pallets, breaking them down layer by layer with robotics, and rebuilding pallets designed for one or two aisles at a time.

Historically, they’ve done that. That’s not particularly new. What’s new is that Walmart has automated the process significantly. That puts a lot of different strains on transport packaging — the pallets and the secondary packaging — especially when you’re removing layers with compression force from end-of-arm tooling instead of with human hands and the dexterity people have.

Under that new regime at Walmart — and it’s the 800-pound gorilla, so when it moves, everybody moves — there’s a technology we’re seeing a lot more of that’s coming back to the fore. When you’re building pallets, people often use slip sheets between products to keep them from moving around and protect them a little. More and more companies, like Henkel Adhesives, Robopac, and [Jawatt], all at Pack Expo, have done tiny adhesive dots between pallet layers. It was a very nice product that kept pallets from shifting and very stable, and there were great use cases for it.

But now, with Walmart’s very tight pallet specifications, which are actually getting narrower, brands shipping product have to be much more careful with dimensions, size, and pallet patterns so Walmart’s automation can handle them. I don’t even know the name for the technology, but adhesive used not for closing cartons, which is what we usually think of, but for tying together layers of a pallet, has existed for a while. Now it’s being used again to create straighter, more dimensionally precise pallets that are much more palatable — forgive the pun — for Walmart’s tighter spec.

So what’s old is new again. Brands are awakening to the fact that to serve Walmart better, they need better transport packaging. You’re seeing all those companies I mentioned, plus Signode with its strapping and pallet handling. From a wider perspective, pallet handling and pallet building, and being very precise with them because of Walmart’s new automation needs, have come back to the fore. It doesn’t sound sexy next to all the new innovations at Pack Expo, but it will have far-reaching implications and impact.

Interesting. That’s pretty cool. I thought of glue dots when you mentioned the little adhesive dots. Maybe it’s a similar product.

It is. Think of a nozzle that applies a bead, or whatever pattern, to close your Crest toothpaste carton. Instead of applying it to carton closing, they’re applying it to pallet locking. We’ll be seeing more of that, I’m sure.

Is there a sustainability angle? Sean and I were about to ask the same question. Go ahead, Sean.

I just wondered: Are there sustainability implications to not using slip sheets and just using the dots?

I’m sure there are. I don’t know the details, but a lot of these adhesives are water-based, and they’re used in such small quantities, especially at the pallet level, that I don’t think it’s affecting corrugated recyclability, if that’s the question. As for reducing slip sheets or using them in different ways, I don’t know how much that moves the needle. At the full-pallet level, load containment is really the name of the game. Once you get down to individual SKUs and units, the package-to-product ratio changes.

I’d also say this is largely a back-of-house situation. Retailers and distribution centers do a really good job of recycling slip sheets and corrugated, and they have [whole-floor] methods for shrink wrap, too. They do a better job than consumers. So I’d trust a Walmart DC to do the right thing with whatever corrugated comes from case breakdown and repalletizing, and with slip sheets and water-based adhesives.

Interesting. Very cool. I’ve been working on some of my Pack Expo reporting and came across one I think is pretty neat. It’s from GlobalVision, and it’s called Verify Comply. They’re selling it as another set of eyes for your compliance artwork. Think of a food item that calls out an allergen. It checks your nutrition facts and checks that you did call out the allergen — or flags that you forgot to mention soy — and keeps you in line with what you need to comply with.

It’s AI-powered, so I wanted to bring it up, since I know we like to touch on AI in every one of these. It’s an AI label compliance assistant: another set of eyes for your labeling. I gather you don’t even necessarily have to feed it your specific parameters; it checks everything over to make sure it complies. So it can be used in a lot of different industries. Sean or Derek, anything AI in your world these days?

Always. And I was going to mention this now that you’ve brought it up: It’s interesting to see AI find its way onto the machinery, rather than just front-end data checking, so that it’s actually being used to verify things like you were describing. I’m still amazed by how much trust is being put in it in such a short amount of time. Maybe it hasn’t been that short now; it’s been a couple of years. But just the trust of putting it on there to verify things like food allergens. I guess if it’s another set of eyes, there are other sensors along the way. I’m just still surprised by how quickly we’ve grown to trust our overlord.

If you don’t mind me interjecting, I saw a lot of that at Interpack. Some of Esko’s programs are kind of the golden goose here. For example, if you’re in California, you can’t use a certain type of logo because of SB 343, but you can elsewhere. Having that extra set of AI eyes checking every package design across an entire portfolio, which for a Unilever or Procter & Gamble has to be massive, is huge. But they always encouraged having a human being somewhere — a homunculus pulling all the levers and making sure things are operating the way they should. So even the latest and greatest at Interpack was not fully divested from human oversight.

Exactly. That’s what I saw a lot of at Pack Expo. AI is being incorporated — “baked in,” as Matt likes to say — into different products. They’re not touting it as AI the way they may have in the past, but it’s there, and a lot of what it’s used for is creating actionable data. Instead of a user being flooded with all kinds of data, it collects it, organizes it, and looks through it for humans. But in the end, it doesn’t just take action on its own. It offers suggestions for the human: Is this bearing about to go? Is this machine running slowly? Maybe a bottleneck is forming at this point in the line because this machine is slowing down. It’s something people can look into, rather than blindly being told, “Here’s the problem, go fix it.” It’s more, “Maybe this is the problem; go take a look.”

So, as Matt said, AI is there, but there’s definitely a human element. I think the bigger push hasn’t been just AI, but humans taking control of their own destiny rather than letting AI control it.

Absolutely. And GlobalVision is actually part of Esko, Matt, so you teed that one up nicely. It’ll flag things and say, “You might want to reconsider,” or, “You didn’t call this out.” But there’s always a human looking, which I find comical, because humans make mistakes too. That’s why we’re using AI, and then we’re doubting AI. So I think working in tandem is the secret sauce for getting as close to 100% accuracy as possible on some of these systems: human and machine working together.

I think that’s the best we can hope for right now, because the AI cat is out of the bag.

Can’t beat ’em, join ’em, which it looks like a lot of companies have done. I heard a lot of AI at Pack Expo, usually on the equipment side: inspection and detection, or in this case verifying artwork, with cameras and that type of thing. AI is going to be here. Derek, did you see a lot of inspection and detection on the food side?

Maybe not as much as we would have seen in the past. There were a lot of data companies this time around. I can’t remember whether I mentioned it on End of the Line or another of our podcasts, maybe one at Pack Expo, but Eagle Inspection Systems introduced a new product — I’m blanking on the name, something “6.” They were touting it as using AI as an assistant to help find foreign contaminants in food products and to extrapolate where along the line something might be going wrong. That was unique about it. There were others at Pack Expo, also seen at Interpack, using different frequencies to detect contaminants. So it was inspection and detection, but not necessarily using AI, rather different wavelengths to detect different types of contaminants.

You said something interesting there that I’ve been seeing as far back as two or three years ago. Rockwell has been positioning some of its vision and inspection systems as not only preventative, removing something that might contain a contaminant, and not only predictive, saying a defect is happening on every 100th tube of toothpaste, but prescriptive. The system knows that if a defect occurs at a certain frequency, there must be a bearing on machine four in the filler that’s near the end of its life, and you’d better check that bearing because you’re seeing, say, a smudge on one side of the tube.

So the AI recognizes a fault or error that’s creating scrap, and it not only recognizes that the scrap is happening but where upstream the fault is, so you can go fix it. Prescriptive is a pretty interesting use of AI.

It’s just so much quicker at detecting patterns than humans are, and that’s where the benefit of AI comes in. If something creates a pattern, it can go back and say, “A plus B equals C; go check this out,” much faster than humans can. I think we’re seeing a lot more of that.

It hasn’t been a great year for recalls in the food world. Can that help?

Potentially. It can definitely help with recalls, but a lot of what we’ve been seeing has been on the bacterial side, which is an indictment of a company’s cleaning and sanitation practices. I’m not saying AI can’t help with that, but a lot of it comes down to human error.

Interesting. Thank you so much, Derek, unless you have anything else to add.

I do. Speaking of human error: Back in August, we saw soldiers losing their ice cream because freezers failed to freeze. They were hacked. Then just a few weeks ago, supermarket freezers were hacked and the heat was cranked up, and fires actually started. A lot of people don’t realize that supermarket freezers can go both ways. They can cool down and heat up, because they have to maintain a set temperature whether it’s too cold or too warm. And most recently, as Liz [knows], a blood bank’s freezer system was hacked.

It’s very early days with all of this. We don’t know the motivations or who was behind it. But what’s interesting is that a lot of blame has been directed at one company’s controllers in particular — I’m not going to name the company — for being left exposed to the internet. Not the intranet; the internet. That gave bad actors access. Why they were left exposed is a good question. That’s what investigators think happened, but again, it’s very early days, and we don’t really know how or why. What we do know is that it comes down to the controllers on these systems, and that hackers were able to gain access and crank up the heat on these freezers.

So what can be done about it? I had a great call today with Kristin King, author of “Securing What Feeds Us: Cybersecurity in Food and Agriculture.” I’ve talked with her about cybersecurity before, but I wanted to talk about this specifically. A lot of this seems to be on the consumer end, but I was curious what food and beverage manufacturers can take away from it, and whether there are problems they need to address.

The answer is yes. One thing that struck me: When you get a new router or printer, one of the first things you should do is change the default password from “admin,” which can usually be found in the documentation on the company’s website or in the instruction manual. A lot of people don’t change those. This is speculation, but it’s possible that was the case with some of these freezers: Maybe someone didn’t change the password. We don’t know whether any of these incidents are related. Maybe someone hacked the U.S. Army’s freezers and someone else thought, “I’ll try that with supermarkets.” We don’t know if there’s any relationship at all, or whether anyone changed the password.

How come this wasn’t noticed earlier? First, the bad actors turned off the alarms that were supposed to notify people if temperatures went beyond the set point.

A temperature excursion.

Exactly. They turned those off, so that’s how they got away with it. But some freezers were hitting 122 degrees Fahrenheit, or 50 degrees Celsius.

That’s hot.

You can’t tell me somebody wasn’t walking by one of these, seeing melted product, and thinking, “Something’s wrong with this freezer.” So again, it’s the human element. How did humans miss this? Did they have checks and balances in place to ensure something like this didn’t happen?

What Kristin wanted to get across to manufacturers is to have a plan in place. Don’t assume that because you have the latest and greatest technology, everything is hunky-dory. You still need a plan, and you still need emergency procedures. More important, probably beyond anything else, is having a culture in your plant where people feel empowered to say something if they notice something. Who knows, maybe somebody did walk by, notice the freezer was pretty warm, and think, “Eh, not my job. I’m a stock person in another aisle; that isn’t my department.” You don’t want that happening in your plant. That goes for food and beverage plants, cold storage, warehousing — any of those environments. In general, not just with freezers, you want to make sure you have a plan in place for your cybersecurity systems.

And we’re going to see more of that, because the retailers we talk to, who own those shelves, are leaning more and more into smart, connected shelves that know the expiration date of the milk on the shelf, so they can lower the price as the date approaches and move that milk faster. I always look for the milk with the longest runway in front of it. There are a lot of advantages to connected shelves and connected freezers. But now we’re seeing the potential dark side: exposure to bad actors.

Really interesting. That access is a two-way street.

That’s what I talked about, too, and I’ll just give a teaser; we can discuss it in another episode. Remote access can be a problem when it isn’t secured. But in the case of the blood bank, remote access is actually what allowed them to see that things were going sideways. So it’s a two-way street, like you said, Derek. Good and bad. We’ll see what happens.

That’s how they caught it, with remote access?

Yes. It wasn’t actually in the blood bank itself. It was at an offsite storage location, and they could see remotely that temperatures weren’t where they should be.

Get that message to all the IT people out there: Remote access was the hero in this situation. I was immediately thinking it was at fault, but that’s a great way to put it. It stepped in and saved what could be saved.

Absolutely. Very good. Gentlemen, thank you for your time today. I hope everyone enjoys this episode of End of the Line, and we’ll talk again next week.