309: How Physical AI Is Changing What's Possible in Warehouse Automation

0:00Automation comes from manufacturing. So Henry Ford didn't make an automated warehouse, he made an assembly line. Welcome to Manufacturing Happy Hour, the podcast where we get real about the latest trends and technologies impacting modern manufacturers. Manufacturing Happy Hour. Each week, we interview industry experts that are at the top of their craft and give you the tools, tactics, and strategies you need to take your career and your business to the next level. And now, your host, Chris Lukey. What's up, folks? This is episode 309. We're talking about warehouse automation versus factory automation.

0:47This is an episode that features owner Uranli, director of Intelligrated, and Josh Cloher, general manager for NoMagic in North America. We talk a lot about manufacturing on this show, obviously. We talk a little bit about logistics, and this is one of those episodes where we compare the two and see where different spaces are accelerating, where others are playing catch-up. We're going to be looking at how physical AI is changing what's possible in warehouse automation, where the manufacturing industry is ahead, and how folks in warehouses are closing that gap. This episode is done in collaboration with TRG, the robotics group of MHI.

1:32We do a big party with them every year at Modex and Promat. We'll be with them again in April 2027. So hey, let's jump in and meet up with Josh and Oner.

1:51Josh, you're welcome to Manufacturing Happy Hour. We're going inside a warehouse today, so to speak. And this is a very much an adjacent topic to Manufacturing Happy Hour, but a part of, let's say, the end of every manufacturing process. So very appropriate that we cover this on today's show. Onur, my first question is for you. I recently came across an article that you wrote. And the first question you ask in that article is, if your palate could speak, what would it complain about most? And to be clear, this is not the palate like your mouth, your taste buds. This is a palate inside of a warehouse.

2:34Yeah. Thank you, Chris, for this introduction. Absolutely. I mean, we are actually in my company, Intelligrated, one of the things that actually people are looking for is actually how they can actually move that large pallet in their warehouses and then package those in a way that actually go in and out their warehouse. And then I noticed actually there's a big element of pallet in my career journey. So I looked at this as, okay, where do I see pallet and how I actually met with this friend of mine that was with me, that has been with me throughout my career. And I just actually wanted to look at to see if what does pallet actually take us and tell us.

3:15So I actually took the readers in the journey where pallets actually start from the manufacturing floor and all the way to transportation carriers and air carriers are on the docks or the ships, comes on the trucks and goes into our customers' warehouses. And this is the same medium that handle maybe the world's largest trade. And then that pallet can probably tell us a lot of what we are actually fixing in our warehouses, which is a lot of times I've seen the companies are actually trying to fix how they can actually receive those pallets of merchandise, how they can actually put away in their warehouses, how they can transport within their warehouse, how they can store and how they can break that pallet.

4:07And then the goods inside those pallets, how they can actually store and replenish the parts within those pallets. So I just actually wanted to make sure that we understand if we can hear what our pallets are telling us in terms of how we design our warehouse storage systems, design our warehouse automation systems based on what the pallets are telling. and then maybe we can actually start there as opposed to trying to fix that palette at the end of the downstream. Maybe we can go upstream and then fix that palette to begin with so that we don't have all those bigger problems. So that's the overall, the theme that I'm trying to accomplish in here in that article.

4:48So I'm glad actually we brought it up. It's very relevant to manufacturing environments as well as the warehouse environments. Yeah, I have a follow-up question to that. And this is for both of you, Josh. Josh, maybe you'll lead us off on this one, but I'm curious. Let's paint the picture for the folks that aren't living and breathing warehouse life every day. How are like the warehouses and the automation organized today, right? Across the majority of manufacturers and logistics companies. Josh, is it high tech? Is it low tech catching up? Are people making the right moves? What would you say?

5:25And then owner after that, I'd love to hear your thoughts as well. Yeah, yeah. So I think the majority of warehouses are still manual today. So there's still a lot of moving items, walking around, you know, manipulation tasks. There's still people walking around with paper and trying to figure out, you know, where do I get the things that are in my racks? And how do I build these orders to go out today by this, you know, deadline that we have? I think it's something like 80 % of warehouses are still manual. But those that have invested have seen a lot of improvement in efficiency overall. And there's been a lot of investment in the middle of the warehouse.

6:09So a lot of what Intelligrated helps with, a lot of larger material handling companies, a lot of material handling manufacturers create these big systems that move things from place to place and get them to the right endpoints for whatever the goal is, whether it's fulfillment, whether it's put away, whatever the workflow is. So there's a lot of things changing, not necessarily in that big middle where you move things around, but on the edge. And that's where no magic comes in. That's where robotics really shine is unloading trucks. That's picking items. That's building mixed pallets, building your orders.

6:45So kind of working with the bigger systems to change how you take it from the system and get it to something that can be shipped out the door or to receive it from a truck and to put it into the system. On the edge is really what's changing and where AI will become much more applicable here in the future. Josh, I'm going to ask you a bit more about NoMagic and kind of just painting the picture of what that story inside of a warehouse looks like in a second. Owner, I want to hear your thoughts on this topic as well. Josh, actually, you touched upon a very important thing. Yes, there is a big majority of the warehouses and manufacturing facilities are actually still not automated.

7:30It's maybe paper-based. But also, there are some majority, maybe minor majority, I should say maybe 20%, is still like sophisticated, high, special or motive. But there is this forgotten middle, right? I mean, Josh was just talking about like that forgotten middle is the companies who are right in the middle. They're trying to organize. They're trying to automate. And don't know what to do and have to do with it. But they're actually understanding the necessity of that automation. So I think with that, maybe companies in the forgotten middle, they don't necessarily need like that multimillion dollar process.

8:11And I think what we could come in place, like the thought leaders, Josh and I, maybe, can you just educate that forgotten metal into like, we can just find what is the biggest draw block or bottleneck that you have in your warehouse. Let's start there and let's maybe pilot there. If it's a pilot receiving, if it's order picking, if it's something that is actually keeping the operators in that warehouse up at night, maybe we can start there and help that forgotten middle expand so that the automation start from that aspect so i think that's that was really important so and also i think what i'm seeing is the companies since it's the big automation uh a task that's becoming on their uh uh their to-do list and they don't know how to go about that they actually liken themselves maybe how can we just automate like the likes of amazon I think that's a big task, maybe Herculean task, so to speak.

9:11Why don't they actually just minimize the impact that they can do, going back to the pain point that I was talking about, and that middle forgotten can be actually really a game-changer for their operations, is what I'm actually thinking. So I have a follow-up question, Josh. I had mentioned I was going to ask you this. You were talking about NoMagic. you mentioned a lot of the work taking place right now, a lot of the strides are being made at the edge. And I think this ties into what you do. So no magic does. And I quote physical AI for end to end warehouse automation. You gave us a little explanation of what that looks like in general.

9:51But can you describe this in the context of a story that really paints the picture of this? Yeah, 100%. So physical AI is a big, broad topic, I would say, and we bring it to market here in the form of item picking. So we're focused on building orders at each level in the e-commerce world, primarily for our partners and our customers. And physical AI really just means that you're having something physically in the world that has a brain that can react in real time to what's going on in front of it. But I think the most, you know, kind of impressive thing from an AI perspective that we're doing right now is we're able to recover from, you know, downtime scenarios with this vision language action model where the robot is literally just looking at the items in front of it and trying to figure out what went wrong and how can I in this moment, you know, react and move the system to solve this problem.

10:53So it's a really kind of cutting edge technology that's out there. And we've deployed it in a production ready way. There's also a lot of systems out there in general that are using this type of technology to create more reactive robotic systems. So you'll see this trend in the marketplace where they're doing things that robots just could not do before. When you think about robots in manufacturing, you know, traditionally where they have been deployed most, they're doing something very rigid. They're going from here, they're moving like this, and they're going to there to do a material handling application or to do a welding process or something like this, where it's very defined what they're doing.

11:27But in the warehouse, we have a lot of challenges with this, where we need to have some flexibility to the way the robots move. And we're seeing more and more of that driven into the warehouse through new use cases with AI. We're going to talk about flexibility here, and I wouldn't be surprised if we touch on it with the next question. This is for both of you, but owner, I'm going to start with you. How is automation in the warehouse different from automation in the factory or are they closer together and there's not a ton of difference? And the reason I ask this question is we have a manufacturing centric audience here, but obviously they dabble in logistics.

12:09There's no way you can get around that in a manufacturing business. So, owner, I'd be curious to hear your take on how is automation, if it is in the warehouse, different from automation in the factory? And then, Josh, I'll pass it to you after that. Yeah, I think Josh touched upon it a little bit, and I think I'm going to expand on that. So in the manufacturing, especially starting with the automotive manufacturing, maybe that's what we can start with. There is a whole host of automation robotics applied into assembly lines and then end of line. It's just like purpose built. The robots are doing the one single task that they do that anybody and human being can do so that they do it perfectly.

12:54because it's one task and that automation is done perfectly. And then there is no other thing that actually there is less exception, there is less flexibility, and more rigid. When it comes to warehouses, especially in today's warehouses where outbound tool family handling as well as the inbound universe logistics, There is a whole host of inflexible requirements. There's a lot of variables. There is a lot of exceptions. And when the robots or automation fail is when those exceptions happen. I mean, I think the manufacturing audience, when they look into a big project for automating their warehouses, that's probably a biggest problem that they have to be looking at.

13:49How much of a variability that they have to deal with? What is the inflexibility or flexibility that have to be put into considerations? What is the exception that happens? And what happens when things go not according to what they planned? And what is the corrective action that the robots or automation can handle? So I think those are the questions that actually experts like myself and Josh can help. fixed. But I see what you're saying in terms of manufacturing can automate, whereas warehouse automation can sometimes not come as automated as well. But we're getting there, depending on the SKU level and then depending on the customer's requirements, if it's need to see your B2B.

14:34And I think there's a way to automate. It's just only a matter of the pain points and the exceptions and the variability. Josh, what would you add to that? Yeah, I really like the idea that Onar mentioned earlier about the journey of a pallet. And I think you could take a similar approach to the journey of an item in the warehouse. And if you think about manufacturing, everything is usually very reliable in how you're going to experience the work item that you're working on or you're processing or you're packing. And if you go into consumer packaged goods and you're building packaging to go around the primary packaged good, to go around the secondary packaged item, and then palletizing it, right?

15:24So you can almost take this. In the warehouse, we take that item that you built in manufacturing and it gets touched many different times until it gets to that customer. And that package wears a little bit every single time. So even within a SKU set, the SKU 1, 2, 3 might look different five times out of 10 just because of how it's experienced its journey from the manufacturing floor all the way to the fulfillment center. So you have to have a much more understanding of variability and how you're going to manage that variability within the warehouse if you're going to deploy robotics and automation to manipulate those items.

16:06yeah and and that's one of the things that comes to mind when i think of automation in logistics just the you know i think a lot of people think of like the u.s postal service or any short sort of shipping where it's like look at all the different type of boxes shapes uh rigidity as well you have soft packaging you have hard packaging right i think that's one of the easiest ways to think about all of the variability that comes into a warehouse that you need to automate around. Next round of our interviews coming up right after a word from our sponsor.

16:50All right, start by saying that I am not subscribed to many newsletters. Like you, I just don't have a ton of time. However, PTC's industrial newsletter is one of the few newsletters that I make an exception for. And you can subscribe by going to manufacturing happy hour dot com slash PTC. Instead of more noise, this newsletter is a quick hit of clarity about how data from design, manufacturing and service can work together. PTC's industrial newsletter is built for manufacturing leaders that want to take meaningful steps to modernize their operations. It's a biweekly curated read on AI automation and operational strategy with a clear focus on the digital thread and closed loop intelligence that helps manufacturers connect decisions across the entire product lifecycle.

17:37Less guesswork, fewer silos, and making smarter choices earlier. PTC's industrial newsletter. Subscribe at manufacturinghappyhour.com slash PTC. And now, back to today's episode.

17:58so maybe one of my follow-up questions then is what is some what are we not talking about enough when it comes to automation and implementing technology inside of the warehouse right now what is maybe it's not getting talked about enough but maybe one of the things that you see that maybe you feel the rest of the industry could be doing a bit of a bigger job to address. Josh, I'll send that to you first. And then owner, if you want to pick that up from there, we'll go in that order this time. Sure. Yeah, I think one of the biggest challenges with building a business case around automation in the warehouses is that a lot of the facilities still run just one shift a day.

18:44And it's very challenging to build a business case for a short-term payback for this small shift operation, short shift operation. And what we found is that they usually are flexing to their workload and that there's some big peak coming as Black Friday comes around. They're getting ready for that big peak and they usually have to throw temp labor at the facility. So they have to build for this peak to come, but then on a day-to-day basis, they have an operation that is much smaller when they're in those normal order times. And what we see is it's much harder to build a business case around automation because they're trying to get this short-term payback.

19:27So building in an automation strategy from the beginning, like during the construction of your building or during the phase where you're trying to pick a 3PL or during the phase where you want to go and figure out what's your distribution strategy across the country here, that's really the time that you should be thinking about automation, not once you've already started to operate the building and deployed a big ASRS system. And now you're thinking, oh, well, maybe we can add a robots or maybe we can add AGV forklifts or maybe we can do this. Doing that point solution business case is a really hard problem after the fact.

20:03And it should be something that you're thinking from the beginning. Josh, just a real quick acronym check here. ASRS system for the folks that aren't familiar. What does that stand for? Automated storage and retrieval system. Excellent. Owner, what would you add to this? Yeah, actually, that was a really good segue that Josh laid out for me. So, you know, any in his football or soccer fans out there, so he did a very banana actually crawl. So I'm actually going out and then had my hand on the wall. Let me actually just actually talk to you about like the paradigm shift, right? So maybe we need to just look at things differently and asking different questions.

20:46So one of the things that I'm actually looking into this, going back to the journey that I took everyone into the manufacturing world for pallets. So the question probably should be like, how much automation complexity are we purchasing to compensate for reliability that entered the system on the pallet? I mean, so this is not a really palette problem. It's a system problem. So we're trying to really like beat the dead horse when the palette shows up in our system and how can we actually put a system in there so that we can put that into our ASRS system that just mentioned, Josh mentioned. I think the answer is not really the operation is a premium palette.

21:27It's just like how we can actually redesign our system based on that palette comes into. and then the real question is like are we tolerating the inefficiencies that's happening and then coming into our way in a way that actually we can maybe like start thinking differently with that paradigm shift. So a couple of things maybe like is it like less mechanical problems like in glass and the tire and the more efficient system designs and then so how can the automation handle our palettes It's just like the problem that way you can start and dig in more. So maybe we should just actually turn that question around.

22:08Are the pallets entering our facility ready for the automation that we're designing for? So, I mean, the question is that we're designing that big ASRS system and that we're ending up receiving inefficiently made pallets. Are they ready to get into our system? Maybe we can actually have an ecosystem where our manufacturing facilities are designing the better pallets for the ASRS system we're actually designing to build in. So is it really like connecting one another? So I think that's the maybe holistic approach that is requiring a little bit of a paradigm shift. And I think that's where real maybe gems stand and that state that we have to find a little digger.

22:46I mean, dig a little bit deeper. I'll ask this very directly. who is further ahead right now the manufacturers or the warehouses when it comes to their automation because when it's all let's say it's a warehouse inside of a manufacturing facility right I feel like it's easier to stay in sync but that's not always the case the warehouse is sometimes removed from the manufacturing facility it might be a completely different company as well so I'm curious who is leading the charge right now and who is maybe playing a bit of catch up and josh you've worked with companies that were you know both you know manufacturing and automation focus or i should say manufacturing and warehouse focus now you're a bit more dedicated on the warehouse side i'd be curious to hear your perspective on this first yeah uh automation comes from manufacturing.

23:46So Henry Ford didn't make an automated warehouse, he made an assembly line. And I think that the manufacturing side, the processing side of manufacturing is quite far ahead in terms of how much automation they are using in their day-to-day life. I think warehouse is playing catch up there for sure. But then what these two different kind of workflows, they really need different things where manufacturing needs precision and control and a variety of ways to do very specific tasks. Warehouse needs more flexibility, needs even a more inexpensive model for the hardware. So to sacrifice on precision, we get this flexibility and cost reduction, right?

24:37So we've always kind of joked that we almost need like a big plastic robot to make it work in warehouse for some applications. So there's a lot of differences in the need. But I think in terms of use of automation, warehousing is definitely trying to keep up with what's going on in the manufacturing world. Yeah, I like the historical context. I think you put it very concisely at the end. Warehousing is trying to keep up with what's going on in the manufacturing world. Owner, what would you say? I think I agree with Josh. Also, maybe we need to add another layer into it. So the problem that manufacturing is solving is a little different than the problems that warehousing is solving.

25:29Right. So manufacturing is solving what comes in and out of their assembly lines or or or bring the pieces together and then how they can actually just cut down this time for that machine or product or whatever that's being built. And so that goes to the next level. It's all process driven. And then there is a lot of maybe work to be done and the work has been done to eliminate the redundancies. And that's why they're ahead. And then that's why they're actually, other minute they were second is actually in form in manufacturing because the volume is there. And then there is a lot of efficiencies to be gained.

26:11But warehousing, because of the downstream element and because of the variety and because of the complexities and then types of the things that's coming in, because not every one single warehouse is handling the same one type of skews. And that's why I think automating warehousing is more complex and then maybe lagging than the automation happening in the manufacturing. But we're catching up. We're catching up because there is this conundrum into consumerism. There's this conundrum of e-commerce and the B2B e-commerce. And then all those things and requirements, the SLAs and expectations coming from the demand side is helping the operators in warehouse to look at this differently, maybe with the mindset of manufacturers.

27:04So I think we're catching up, but manufacturers are always good processes that we can learn from. And I think it's going to be a good conversation to be had in who learns from who. It just made me think of something here. When you think about manufacturing, you're trying to create this process where one thing happens right after the next one. And work in progress or WIP is kind of something evil to manufacturing. You don't want to have this buildup of WIP in between your process. But in the warehouse, everything is whip. If you think about a case picking operation, every pallet on the floor is work in process.

27:50It's something that you have to figure out how to go and work on and then go and grab something over here and then go and grab something. It's so much more distributed labor across a wide square footage as compared to you kind of stay in your spot in manufacturing and move it to the next person or the next process. So it's much harder to figure out how do you build automation around this problem, right? Yeah, that's a great point. Like the fact that, hey, in a warehouse, everything is whip. I like how you brought that up. I think that's a great way to put it in perspective for our manufacturing audience.

28:34The other thing that came to my mind is warehousing is one step closer to the customer than manufacturing is. And any time that happens, that's where complexity comes in. Right. And we've talked about why there's more complexity there and why that's causing some of the lag to automate warehouses a little more than maybe we've automated manufacturing in the past. So great points across the board. Last point of discussion as we get to the end of our conversation. How is the industry ecosystem helping everyone evolve? You're both very involved with TRG, the robotics group, which is the automation focused segment of MHI, the world's largest material handling and supply chain association.

29:21I mean, maybe use that as some context as to how is the ecosystem helping everyone elevate and, you know, continue to automate and play a bit of catch up to make sure the warehouse is just as automated as the factory. So, owner, I'll start with you and then Josh, we'll go to you to close out. Thanks for shining out to TRG and MHI. Actually, I think us as practice or thought leaders, we have a big task to actually help the industry to combine our expertise with what's happening in the manufacturing and then bring our expertise in warehousing and then tie them all together. I think you cannot separate in the ecosystem of supply chain where literally the supply chain means bringing the supplies even upstream from maybe materials that are actually assembled in the warehouse of a factory floor, then assembly line and then the finished goods.

30:26and those may be moved into another warehouse in the manufacturing for transshipment into international destinations or domestic destinations. And from those warehouses, the ecosystem of transportation and the receiving ends of it. And then maybe those parts are going into another manufacturing in the United States for another assembly. And then those become a part that I, as an operator in the warehouse, receive for my end customer that could be a business or that could be a consumer so if you look at all that ecosystem and then parts of it there is some redundancies that can really help alleviate the problems that we're trying to solve and then i think one could be better of uh if we can actually look holistically if you're talking about one manufacturer that also has manufacturing facilities but then direct the consumer business.

31:23The best way to do this is actually looking holistically. What is going to happen for my ant warehouse so that they can ship those ant products to my customers better if I can actually look at my packaging, my pallets, my way of sending them the products better. So if I can do a little bit of those works in my manufacturing, would that help downstream? so that my internal customers or external customers can fulfill the needs of the final customers. And I think if you look at that, Lance, we can do a lot of achievements done. But overall, I'm having a great time seeing reshoring and bringing everything to Make America because it's helping us in the warehouse automation ecosystem the better making of things for manufacturers.

32:18it's a hand-to-hand conversation so there's a lot to be gained but I think we're moving to the right direction Josh, take us home Yeah, it's hard to build a company that can be all things to all people in a big market like warehouse automation and the more that providers work together that provide different things and that are complementary to each other the better for the market you see a lot of consolidation happening right now So there are big groups that are coming together and finding things that they can bring under one umbrella that they can bring to market and be successful. But I still think that manufacturers of automation equipment, integrators, consultants, companies like mine that focus on intelligence and robotics, those are companies that really need to find a better way to work together in general to help serve the market and not be so isolationists focused on their customers and kind of scared to bring in another company to their situation with their customer.

33:23And I think TRG is a place that helps this kind of relationship building side that helps us learn more about how are we complementary with each other. So it's a great place to work together. And Oner and I, for instance, are on different sides of this warehouse automation market where he's focused on the big CapEx projects with big sortation systems and lots of different technology. where we're very focused on the item picking side. And we have the entire company is on this one mission for this one type of solution inside the warehouse and only a couple of different workflows. So there's a lot of variety and we need to figure out how to bring that together and be complementary to each other as we face the market.

34:14Well, Josh, owner, I appreciate you bringing your perspectives, not only on the industry, but the community around that. I think we see what the community can do firsthand every year when we get together at Modex or Promat. Josh, I was actually thinking back to one of our conversations a long time ago. What, it's 2026 right now? I think the first time you jumped on the show was back in 2022. And we started the conversation by me asking, it's like, hey, man, if we were having this conversation over a drink, where would that be? and you said Monday Night Brewing in Atlanta, which Josh, we finally got to grab a beer there a year ago, correct?

34:55That's right. And we are actually in the process of throwing another big party with you guys at Promat in Chicago in 2027. So keep an eye out for that. But I have a feeling, owner, Josh, Josh, you and I will be grabbing a drink in Chicago in the near future as well come April 2027. Can't wait.

35:28Hey, everyone. Thanks for tuning in. If you want to connect with Josh, if you want to connect with Owner, make sure to head to the show notes page at manufacturinghappyhour.com. As mentioned at the beginning and at the end, we collaborate with the Robotics Group. We throw big extra innings after party at Modex and Promat every year. Our next party is on April 20th, 2027. So if you're really proactive, mark that calendar. Whether you're listening to this before or after that, always keep an eye out for what MHI and the Robotics Group TRG have going on. Thank you for making this episode possible.

36:10Stay innovative. Stay thirsty. We'll catch you again next time. Thanks for listening to Manufacturing Happy Hour. Powered by the Industrial Network.