.jpg)
ROS Goss
ROS Goss is powered by the team at Dwight & Company — a specialized sales and marketing agency that lives at the intersection of automation, robotics, and storytelling. We work with everyone from massive integrators to stealth startups, helping them grow with the right strategy and messaging.
So why a podcast? Honestly, we never planned on doing this. But after years in the industry, we kept hearing incredible stories — not just about technology, but about the people behind it. And we realized those stories needed a place to be told.
ROS Goss is where we share the human side of a very automated industry. We’ll talk about innovation and thought leadership, but also mentorship, customer success, and the relationships that make this industry tick. We’re sitting down with some of the most inspiring voices in automation to have real, honest conversations — the kind we don’t always make time for, but really should.
This isn’t just about robotics and automation. It’s about the people building it and we want you to meet them.
ROS Goss
Shots Fired!
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In this episode of ROS Goss, two industry veterans, Pete Allen and Guy Courtin, go head to head on some of the most contested topics. Does a warehouse robot need legs? What is actually selling versus what is just hype? How is physical AI changing the software stack?
Pete and Guy discuss what’s real and what’s overhyped in warehouse robotics, from humanoid robots to “physical AI” and the next wave of automation. We also get practical about how end users should evaluate risk, ROI, and change management before they buy anything.
If you care about warehouse robotics, AMRs, ASRS, WMS and WES strategy, physical AI claims, and what automation will look like by 2027, this conversation is built to sharpen your thinking.
Bold predictions, hot topics, and differing viewpoints. Shots fired, indeed.
Welcome And Why Shots Fired
SPEAKER_00Hi everyone, welcome back to Ross Goss. I'm Mandy Dwight, host of Ross Goss Podcast. Also the founder of Dwight and Company, a sales and marketing strategy agency specifically for automation robotics companies. So for our 15th episode of Ross Goss, I'm joined by two industry colleagues that, you know, no pressure, but they're on podcasts together all the time. And they they definitely have some opinions that are the same, but most of them are very different. So I'm joined today by Pete Allen and Guy Corden, industry colleagues, both have been in the robotics space, they've been in supply chain, they know manufacturing. So we thought we'd get together and film something today that's called shots fire. It's pretty appropriate. Yep, exactly. We have our hats at the end of the table. Um we are live in Boston. We did want to film in Lexington and Concord. It turns out that we had some torrential rain, so a little bit of a gamble. So here we're in the conference room at our office. Um, but it's appropriate because it's 250 years of the US. We've done a lot of talking about Shots Fired this July, and now it's Shots Fired, the Robotics and Automation Edition. So thank you guys very much for being here today. The traffic was horrendous, so I appreciate you coming in. Um, but you know, you guys know me, so I'm gonna turn it over to you guys for introduction.
SPEAKER_03So go for it. All right, thank you, uh everybody. My name is Geek Kochet, I'm the Vice President of Industry and Global Alliance at Texas. We are a Canadian-based WMS vendor, uh, but I've been in supply chain for over 30 years, uh, with as Mandy mentioned, uh, for robotics companies, but also for other supply chain execution and planning companies. Uh so looking forward to our conversation.
SPEAKER_04Pete? Mandy, thanks for having us here today. Um, my name is Pete Allen. I have been in the robotic automation space for about seven years, started running a business unit at Berkshire Gray, helped with partnerships, went into the fun world of humanoid robotics at Agility. Robotics are running their partnership program. Currently, I am a managing partner at my own consulting company, and I uh support Jacoby Robotics, Bechnica Robotics, and Breckham Group on all go-to-market strategy and execution. So thanks for having me again. Appreciate it.
SPEAKER_00Yeah, and and the reason I thought this would be a fun conversation is you guys always have such bold opinions. Us? Yeah, I know. Very shy. So coming in today, what is the boldest opinion that you guys have right now before we kind of get into
Chinese Robotics And Security Blowback
SPEAKER_00everything?
SPEAKER_03All right, here's mine. Uh in the next couple years, we are gonna have some issues with Chinese robotic vendors. Okay. Oh, interesting. So if anybody has read the news recently, it just came out that uh PJs mentioned you bought a new car, but one of the rival companies, right, Polestar, uh, has been now banned from selling EVs in the United States.
SPEAKER_01Really?
SPEAKER_03Yes, that means there's 32 US auto dealers that have been geared up for Polestar. It's a subsidiary of Volvo, but it is owned by a company called Geely out of China. Uh, and it's being banned because of apparent risks around things like the navigation system, the cameras, etc. So the Chinese EV company will now be withdrawing from the United States market. It will be continuing in Europe. So, my big bull statement, and I've been saying this for a while, and we've seen this in other Chinese companies, right? TikTok is still under threat of being banned in the US. A massive electronics company, Huawei, which is dominant in the globe, is not allowed to have any US presence. So, my big bull prediction, and we're seeing a lot of Chinese robotic vendors obviously doing very well here in the North American market. My prediction is in the next couple years I'm gonna look at is that will there be repercussion or blowback on now the Chinese robotic vendors? Because when I read this about Polestar, well, you think about robotics, they're in your warehouse. Yeah, they know the orders, they know what you're picking, they know what your warehouse looks like, they've got cameras, they've got physical AI, which I know we're gonna talk about. Yeah, right? They are now being tied into a server that, yeah, we can say it's in the US, but really where's that patch coming from? So, my big bull prediction is look for this in the next few years. That I think some of the Chinese robotic vendors need to be very careful, or at least the customers of those need to be careful of does the US government start looking at these robotic vendors as the same way they look at a Huawei or a Polestar?
SPEAKER_00You know, it's it's interesting that you say that because there was one integrator that said to me years ago that they choose their components. They do a lot with food and beverage, and they choose their components based on where.
SPEAKER_03Governance. Yeah.
SPEAKER_04Yeah, if that comes true. That's interesting how it impacts Geek and High and Quicktron, those kind of AMRs, right? Um, and then you think about the humanoid world that I came out of, that there's a ton of Chinese companies that are doing that, and I keep telling people that I don't think the North American market is going to let them in. I think there's gonna be a battle there, but we'll we'll talk more about humanoids later.
SPEAKER_00Yeah, what's your bold prediction?
SPEAKER_04You know,
Case Handling Finally Gets Solved
SPEAKER_04I I don't know how bold it is, but I think if you look around warehouses today, probably the number one item that is still almost all manual is some level of case handling. Yeah. Right? It's a very manual process, whether it's picking, palletizing, depaletizing. I think we're gonna see newer technologies uh come into the market that address problems that are very difficult to solve. Um, and I think we're gonna see price points come down as solutions become smaller and they have less infrastructure. Right? There's some companies that are already working on this. Uh uh, you know, some of the bigger integrators are thinking about how do you do this. Um and so I think we'll see different technologies in that area. Uh Jacoby changing the game on case um on unsequenced uh mixed case palletizing, right? It's been the holy grail in the industry. People have been trying to solve it for years. The impact is if you don't have two sequence, you can shrink the ASRS, right? You can change your design, right, and make it smaller, less impactful on the on the infrastructure and put the chaos at the end of the line. Um, or other solutions like Vecno Robotics with a uh entry-level, low cost, no infrastructure change, locus style type solution, I think is going to change how companies think about automation, right? And it'll let those who are worried about risk and that big change get into it easier at much lower costs.
SPEAKER_00Yeah, oh that is interesting. Yeah. I mean, space is everything to these facilities, right? And ASRS solves that, but those are still big.
SPEAKER_04ASRS solves it, but you've got to, at least short term, find space to put the ASRS. It's still a problem, right? And you still gotta have your flow of your building going. But you know, if I look at that Vecna solution, I don't have to change the infrastructure, right? I can just put bots and a software layer in to do the orchestration of people and bots, and I'm low cost, no infrastructure change, and I'm now into automation. And there's a lot, I mean, if if 90% of the warehouses are still manually picking cases to a pallet, because they're terrified of the cost and the infrastructure problems of automation, it's an easy way to jump into it for them.
SPEAKER_00Yeah,
Humanoid Hype Versus Warehouse Reality
SPEAKER_00agree. So when we decided to have this podcast, I think it was um at Modex in Atlanta. Yeah. We were kind of chatting, and you know, I think you guys were going back and forth about humanoids, which you do, you did when you walked in this room today. So, you know, just kind of to talk to you a little bit about that. So where are they right now? Um, what's going on with that industry? Are they going away soon? I mean, I know you have bold opinions about this, both of you.
SPEAKER_04So I'll kick this one off since I kind of spent my a little bit of my career in this space. One, I I find it wildly impressive on one hand and scary on the other hand about the valuations these companies are getting with no revenue. Like they're not producing real revenue in any way, shape, or form. They are losing money left and right, their spend rate is crazy, but yet Figure's getting, you know, a 35 billion valuation. What is it? Humanoid AI just got a 1.3 billion valuation yesterday or something like that, and raised money. So this the staggering amount of money they've raised is insane, but to the marketing side of them, kudos. Like, great job, right? The reality is these robots are not ready for prime time today. Like there are very few that have been put into a warehouse or into a manufacturing setting. Like I would call it on the planet less than 50. Right? But if you hear the hype, you'd think there's hundreds and thousands of these.
SPEAKER_00My family does based on what they hear.
SPEAKER_04Right. It's it's just not true, right? It just, you know, it just hasn't happened. The agility has has more out there than probably anybody, but all of these companies are doing proof of concepts. I think you were at the uh robotic summit with me, and you had Pros from Agility, you had uh one of the leading engineers from Boston Dynamics, who are two companies I would say are honest about where they're at and what it's going to take to get there. They both said three to five years, and that we're gonna have to spend the next the time three to five years deploying these to get real world data. What they're missing in the physical AI world is real-world data to train their models. And there's no such thing as humanoid robotic data on the internet. Like AI engines like uh Chat GPT, they work great because they consume the internet and they learn, right? Can't do that for a humanoid. They have to learn how to move their actuators, pick things up, put them down, and they're single-threaded. They're trained to do one thing at a time, and they proof of concepts. And there's a whole other reasons why it's not ready, but it's three to five years before we get there.
SPEAKER_00Should we believe three to five years?
Stop Designing Robots Like Humans
SPEAKER_03I mean, my point of view is the humanoids we see today in three or five years will look very different themselves physically, right? So I think that's the first, it's my opinion, right? Because I I've always stated this, and Pete and I have had this discussion, and I would even think your former employee would argue if you look at the way you built the humanoid, the knees were inverted. Oh, right. Yeah. So therefore, digit, right? Digit had right because of bird legs. Right. Because the reality is that I look at the I look at it this way, like the humanoid built like a human is not optimized for anything. Because we as humans, right, the human form factor is amazing. We do a lot of things. We can run a marathon, we can scale Taipei 101 without any ropes, all this great stuff, but we're not optimized, right? I Michael Phelps can swim really well, but you put in the ocean against a shark, he ain't gonna win that race. What an image, right? You put Usain Bolt out there from a cheetah, he ain't gonna win that race. So we're not optimized. So why are we therefore trying to build a robot that's like us? That's sort of I read this article, it's like a Swiss Army knife. So it could do a lot of good things, but it doesn't really do everything really well. Whereas the robot, the whole reason for a robot is I'm gonna work on one or two use cases and I'm gonna optimize the heck out of it. I'm gonna make sure, for example, if I'm moving a two-ton pallet, that robot's gonna move the two-ton pallet. You and I can move that two tons, it's gonna be hard.
SPEAKER_02Right.
SPEAKER_03And we don't want to do it all the time. But a robot just built for that, it's gonna move it. So to me, will humanoids be around in three or five years more ubiquitous? Maybe. But I would argue the way it's built today will look very different, whether it's on a different base, right? Not too late. Whether it's why does it have to have two arms? Right. It can have three or four, right? Why does it have to have a hand? Couldn't it have different types of grippers? Why does it have to be mobile? Why does it have to be mobile? Why does it have to have a head that looks like our head? Can it just be a screen? Like, do we need that? So I would argue that a humanoid style will probably be around, more ubiquitous, but it's not gonna look like a human. It's gonna take parts of the human, but then optimize for other parts. That's just my my belief because I just think again, part of it I think is it's right now it's it's an ex science experiment. It's like, well, we want to make it look like us humans, right? I was just an example too, right? When we first started, humans tried to fly, we build models off of birds. They failed miserably. You go and jump off buildings trying to fly and die. All of a sudden, these two brothers figured out wait, but lift. So let's build a wing with thrust, and that's gonna let us fly. Planes today don't look anything like the birds in my backyard. But we fly all over the world. So I would argue the same thing about humanoids. So two to three years, sure. The same four factories see today, no.
SPEAKER_00So talking about two to three years, like when does it start making financial
The Economics Behind Humanoid Adoption
SPEAKER_00sense? They they might be deployed, but you know, like AMRs, like the the you know, it took a while for those to really catch up with you know making sense for the balance sheet. What about humanoids?
SPEAKER_04The biggest thing that needs to happen is the bomb cost has to come down significantly. Um I saw about a year ago a single part breakdown of an Optimus robot from Tesla, and it was like $125,000 just for the bomb, right? So now you've got to add margin on that and software on that, right? But I would say today I think the market's probably between that $80,000 to $125,000. And realistically, we know the industry, people are buying robotics at less than three-year ROI, right? And if you're not replacing enough to FTEs, and if that robot is still single-threaded in terms of a use case, it's never gonna pan out, right? So what are they doing now? They're losing money on every proof of concept they do, right? Because their bomb costs are too high, because they have too many people in their buildings, but they need to get that real world data. So from a CFO perspective, this comes back to this three to five years. Is the bomb cost gonna get low enough? Like, is it gonna reach the mark of what what Elon and uh and Brett from Figures say of 25 to 30,000? Not in three to five years.
SPEAKER_01Yeah.
SPEAKER_04Because they need significant scale, right? So it's it to me that's a huge problem for them for all the companies in the industry. And they're and there's a new humanoid every day. Oh, I swear to God. It's crazy.
SPEAKER_00I know, I know.
SPEAKER_03But I think part of that too, Pete, I think part of that new humanoid, back to what you said earlier, it's about capturing the marketing buzz. Yes, right? All of a sudden it's all you know, our mother and father, oh, the humanoid is coming to get us because so-and-so just rolled one out. That's a science experiment. It looks cool, it can do a backflip, it can dance, it can run a marathon. Is it scalable? Is it makes sense? Is it CFO gonna buy it? No. But for that sugar rush for that company, they get it. Yeah. Right. And I think part of it is hey, we work on this, maybe the humanoid isn't what we end up selling, but the things we learn from it we can then parse into, I don't know, a sortation robot that has three arms that's fixed at the end of a semi-line, and it looks better than what we have today because we learned that.
SPEAKER_04I don't know. Well, you you see this trend though, right, in robotics. Go back to AMRs. How long does it take AMRs to scale? You were you were in the early days of AMRs, right? They started very similar to humanoids, like hypey, right? Slow to scale or slow to deploy, got a few deployments here and there, and then all of a sudden it's been about what, 10 years? Yeah, and now they're starting to scale, right? They're scaling now. So I think humanoids would be on a similar curve, and they'll they'll come nowhere near the hype of uh Brad Alcock saying, I'm gonna sell a hundred thousand robots in four years. No chance in hell.
SPEAKER_00Yeah, not at that price point, right?
SPEAKER_03Well, not that price, but I also think again, it's it's you know, you go back to talk about Elon. I think he makes those bold statements, and all the finance bros are like, oh, and they fall over each other, and then they so then they hear about the next humanoid company, they're like, oh, we're you know, these guys are gonna be the big one. It's the new Tesla or Apple, let's invest in it. But you and I both know, like, no, it doesn't, that the the the math doesn't work, and in the environment you try to put them in, it doesn't work.
SPEAKER_04I here's where I would slightly disagree. I think when they become multifunctional, right, and when the safety regulations are in place and they can come out of cages, because they don't talk about they're in cages today, um, but they are, or they should be because there's no regulations. When that happens and they can move around a building and they can go from manipulating totes to handling cases to handling eaches, now they become super interesting. Agree. Right? And that's when it becomes like that, and it that becomes complementary to AMR, complementary to ASRS, and now you've got a solution, but that's a big leap where we are.
SPEAKER_03That's a huge like to have the human humanoid to me, like what we saw. There was an art, there was a 60 minutes piece, right, about uh the humanoid in the Hyundai plant doing sortation.
SPEAKER_04BD.
SPEAKER_03And it was like, oh my goodness, and it's like you look at it and you're like, and you listen to the BD guy talk about it. He's just like, if you knew how long it took to train it to do that, you'd realize that's just literally sorting like a windshield where the labor looks at it as I'll learn that in 10 minutes. Right. But now you're asking it, okay, you're gonna sort in the morning and then in the afternoon go pick each of shampoo for sorters, right? That's a to me, that leap is not inconsequential. That is, you know, that crossing the chasm, that chasm's massive. That's a massive chasm.
SPEAKER_04Might be.
SPEAKER_03Right. And I also think the other chasm across on that is a change management. So now you've got FT in your warehouse, and now you're saying, hey Pete, don't worry, the robot's not taking your job. Next thing you know, you see the robot next to you training to pick that bottle of shampoo. As much as your boss tells you it's not here to take your job, you're like, Well, I'm not stupid. I mean, you know, and it's it and I think there's gonna be that chasm to cross too, which is people are gonna say, Whoa, wait a minute.
SPEAKER_04Yeah, for sure.
SPEAKER_03Like if you want to make it do the dirty jobs that I don't want to do, fine. But now you're making it do picking, which is what I do. So I think you've got the technical chasm, you've got the the personal, you know, human element change management chasm, which I think are pretty big. And I just I don't see that in the next three or five years. Maybe, maybe I'm too pessimistic. I just the whole notion of the Jetsons having robots amongst us.
SPEAKER_00Well, that was supposed to be years past, right?
SPEAKER_03Well, yeah that was supposed like in the 80s, right? Yeah. And we don't have flying cars yet? They do with Saudi Arabia.
SPEAKER_00Yeah, tell me about that parking situation.
SPEAKER_03Yeah, it'll be a lot further if we had them here. Yeah, right.
SPEAKER_04Exactly.
SPEAKER_03But I just I I I'm wondering, because I mean, you I saw this selling AMRs. Like you go into some warehouses, and the air AMR is literally just like, hey, it's just it's just your cart, but it's doing it for you. And we already got the like, whoa, you're taking my job. Like, no. See that cart you push at Sevi? I'm just gonna make it mechanical now. You know, even that people were like, eh.
SPEAKER_04Although I think now in today's environment versus you know, eight, ten years ago when you started with AMRs, you have definitely labor problems where people don't want the heavy lifting, monotonous job. That turnover there is 2,000 percent, right? So there is a natural place where they're going to start. Yeah, right. They're gonna say, my turnover's too high, stick that humanoid there, because it can pick up a heavy thing and and move it. Right. Right? That's what it's designed to do. And then my insurance costs go down and all those other things. And then the CFO might smile a little bit.
SPEAKER_00Yeah. A little bit. Well, the workers' comp and a lot of those things. Exactly. Yeah.
Physical AI Marketing Versus Meaning
SPEAKER_00So I'm gonna put another heavy word on the middle of the table. Ooh.
SPEAKER_04Fun.
SPEAKER_00Physical AI. Or is it few words? I'm not sure. Okay.
SPEAKER_04So it's not just AI anymore. We have a whole word on top of it.
SPEAKER_00Okay. Yep.
SPEAKER_04All right.
SPEAKER_00So you're hearing it everywhere.
SPEAKER_04All the time now.
SPEAKER_00Is it a rebrand? Was it always there? Is it really a thing? I don't know. You tell me.
SPEAKER_03All right, I'm gonna take this one for I I think it's a it's a pure rebrand.
SPEAKER_00Okay.
SPEAKER_03In a lot of ways. Now, is there some element of actual AI being integrated into the physical machine? Absolutely. I'm not denying that. And I don't want to get this whole discussion of AI and you know, blah, blah, blah. It's been around since 1956 and et cetera. But when I think about like physical AI, and some of, you know, here we're here in Massachusetts, some of our local robotics companies, they're just pushing, oh, it's physical AI, it's physical AI. I'm like, we've always had intelligence and software built into the robot itself, right? We've always had to have that because even simple things like LiDAR and 3D cameras, so the robot knows when it's going down the aisle, if it recognizes an object, it steers around it. I could argue that is some form of AI because it's gonna alert and it comes back down that aisle, it's gonna look for that object. If it's not there, it's gonna say it's not there, it can go this path. If it is still there, I'm gonna go around. It's gonna learn from that. It's gonna tell its friends, hey, in aisle seven, someone put a paddle in the middle, so just be careful. So we've already had that since the beginning of these AMRs. I feel like now, because there's so much hype around AI, they're all trying to find, when I say all, all of us in the industry are trying to find a way to wrap it into something that I could then say, this is why you're investing in in AI, this is why you're investing in my company, because I'm doing physical AI. And I also think part of it, it's because we're all lemmings. So I feel as if some of the robotics companies who aren't saying this, someone's gonna say, Well, why aren't well, wait, what's your AI strategy? You don't have an AI strategy, you're not doing physical AI? No, we're not gonna talk to you. You know, I'll give you an example. A couple years ago, not related to physical AI, but I was meeting with a customer outside of the United States, and at the end of the meeting, they literally said, Well, what's your AI strategy? And I was like, wrong question. What problem do you try to solve? Are we gonna leverage AI as a tool to solve that problem? Maybe, maybe not. Might not be the right tool. So I Feel like a lot of it right now, it is a marketing hype. It is how do we keep up with the Joneses? How do we show that we're advanced in all this? But I would argue today, the robotic systems you and I have been working with for a long time, there's always been some level of intelligence baked into the physical robot that you could rebrand it as physical AI.
SPEAKER_04Yeah, I think the challenge is most of us don't know what the definition of that word is. Right? And because there's a lot of marketing hype around it. Although I I when I think about it, I kind of boil it down to a simple definition, right? There's always been this concept of AI and everybody used it for marketing. Now it's the new one is physical AI, particularly in robotics. I even see companies comparing it equating physical AI to humanoids. As if it were the same thing. And it's like, no, it's not. But here's where I think it breaks down, though, too, and specifically in robotics. I hear you on the point about robotics has some intelligence. I would also argue a lot of robots companies don't really have AI. They have software that runs their robot. It's not learning anything, it's been trained to do a thing, but they will call it AI, and now they're calling it physical AI because of investment strategy or before or marketing buzz. But there are some companies that truly have an intelligence AI-based approach, like I'll I'll use Jacoby as a good example, right? They have a physical AI approach in that their software actually improves, the throughput actually improves the more it learns the SKUs, the more it learns the environment. And if you change the SKUs, it'll actually have to go back and relearn, hey, this is this is new. To me, if I would put a definition of physical AI, it's can the software, does it have an AI capability to actually learn how to do its job better over time? And I would say to you that if we really dug into some robotics companies, probably most don't have it. And some of the newer ones that are using newer tools probably do have it.
SPEAKER_03But then isn't that an argument that the complexity of a use case a robot is handling is more of an air quotes physical AI? As you said, what does that really mean? Right. As opposed to just, all right, a machine learning, I know, you know, I know this aisle from 4 to 6.30 gets clogged because these orders come through, so I'll avoid it, right?
SPEAKER_04That is the robot doing that, or is there an optimization layer and the software doing that?
SPEAKER_03Or but is the optimization layer then being translated at the robot level, right? Is the robot that's the good question, right? Is it but isn't part of the beauty of physical AI is that at the at the software level, we're gonna learn from all the robots. That's the whole thing of AI, right? Right. Hey, Robot X is learning to your point like Jacobi. Robot X is building this palette this way and learning that these SKUs are actually off. So I'm gonna tell all my other little robot friends about this SKU so that we built we optimize the palette. Optimize the palette.
SPEAKER_04Exactly.
SPEAKER_03So now is that back at the software layer, back in the cloud, or whatever, or is that the robot layer?
SPEAKER_04That you know it's probably several combinations of all, right?
SPEAKER_03Combination of all. So then does that mean it is doing physical AI, or it's doing AI that's being then translated into the physical actions of the robot?
SPEAKER_04Yeah. Is there uh those are all good points? I think I think my simple point is that I think there are some robotics companies that it's there's just some don't get better. They don't learn. And then and but they're probably using the words physical AI, right? And I think if I'm if I'm a user out there and I'm looking to buy something, I'm gonna hear physical AI from everyone.
SPEAKER_03Absolutely.
SPEAKER_04Who's really shown that they can take throughput from 400 UPH to 500 UPH after it was installed. Right. Right. I think that's to me an example of physical AI being manifest into a result that a customer's gonna be happy with.
SPEAKER_03So reality is, I think we agree on this, like it's all marketing BS, but there is a layer that's actually leveraging the tool to do what it's supposed to do. Right. And it's up to the buyer to really ask those hard questions to figure out, hey, are you just throwing this crap around because it's the cool thing to say? Or do you actually have an AI engine somewhere that's, as you said, truly learning and adapting once it's put in? And I think that's the hard question because no offense to salespeople, but I guarantee you nine outside of the salespeople are just like, well, of course we have physical layoff.
SPEAKER_04Yeah, yeah.
SPEAKER_00But I would say, yeah.
SPEAKER_04I I think I think the the the end user should be asking when you put that system in, what was the uh SOW for throughput that you had to hit? And then six months later, where were you at? That'll tell you if you have some form of physical AI that the system learned how to get better.
Where End Users Should Start
SPEAKER_00That's a great question. I mean, speaking of working with end users, if you have an end user that and and it is noisy, right? They're hearing physical AI, they're humanoid, blah, blah, blah. They're hearing all these things. If somebody comes to you and says, I want to do robots, as they do. Yeah, where do you tell them to start? Like, where do you guide a somebody like that?
SPEAKER_04What's what's even funnier, right? They not only do they ask about where do I start with robots, they always start with, I have this unlimited budget. I want to automate everything. I want lights out, right?
SPEAKER_01Let me write down the name of the that's the one that I the lights out part I think they start there and then they get the price tag of that.
SPEAKER_04And then they're like, yeah, our budget's not that big.
SPEAKER_03One robot.
SPEAKER_04They throw open their mouth a little bit, they're like, oh, right.
SPEAKER_03What do you mean I can't do lights out?
SPEAKER_04For 50 million bucks, I can't get lights out? Come on. Um I I think it it depends on the end and who they are, right? Like if I were consulting them, right, I would look at what have you done in automation to date? Let's say you're a company and you've got 10 DCs, right? And you've automated nothing. Right? There's a reason you've automated nothing. You're risk averse, right? Or your abilities are so constrained, you can't do anything, right, from an infrastructure standpoint. That's just gonna that's not gonna throw off your production, right? So in those cases, I would say to them, let's find entry-level solutions, right, that are going to complement, you know, or are going to impact areas that you're struggling with, right? So I'd go back to the pain point of labor, right? We all know that's where it starts, but where are the highest turnovers happening? You know, is it in case picking, is it in uh palletizing, is it in each's, probably not in each's is my guess, uh, because it's so easy to do, right? Um, but I I'd find those places and say, let's start with a solution that's not going to impact infrastructure because that client is that risk-averse one, right? And it's gonna be low cost, it's entry level, I think locus, I think the case handling, uh aisle case picking from VECNA, those are great solutions that are no in no infrastructure impacts, low cost, have uh strong ROIs in most cases, and that's simple. Now they're into automation. Now five years go by, and now the question is how do I get more out of my building? Right now I might start to think about ASRS and things like that. But again, it goes back to who they are and what's their profile. Now, if they're another customer who's already done a green field with ASRS and they're fully automated, now I'm thinking, how can I optimize that more? What what other complementary technologies might fit that building that's not done already today, right? So those are some ideas.
SPEAKER_03No, I I I agree 100% on that one. I got shots fired, it didn't work.
unknownI know.
SPEAKER_03No, but I I I think what Pete said, I think the one part that you said I think is oftentimes overlooked is what is your appetite for change? What's your appetite for change management? That's well said. Are you risk averse? Because I think too often, to your point, oh, we want robots. Okay, let's take the money aside. Then you start talking to them, and you're like, oh, by the way, yeah, you want to put an ASRS in there? Well, you have to shut that down, pull all the SKU out, put the ASRS in, then pull the SKUs back in. Meanwhile, that section is closed. Like you can't use it. Oh, you want to put a uh goods to person. All right, I gotta sequester this area, and I gotta do all this stuff. Oh, wait a minute. Oh, by the way, your labor now is gonna be interacting with robots zipping around. Yeah, they'll be fine. Are you sure about that? You know, because now all of a sudden they look at this little cute robot and they're like, it's trying to take my job. Now you could tell them all you want, it's not, but in the back of their minds, they're already thinking it. Yeah, it's that change management. And I think that's the part that too often we glean, we gloss over.
SPEAKER_04Although I will say though, that in a lot of cases though, when you put the robots in, you find out that your turnover stops. I agree, but it's it's it's getting so they're afraid they're they're afraid of it, then they start working with it and they're like, Oh, this is awesome. Right. I don't have to pick stuff up all the time anymore, and I don't have to walk 10 miles a day and all this.
SPEAKER_03Oh, yeah, I'm gonna keep this job. I agree, but it's like how do you back to cross the chasm? You get over that. Yeah, and I think to your point, Pete, I think it's it's a lot of it is the first thing we want robots. Why we want more efficiency? All right, what's the ROI? Well, if I invest X amount of dollars in this robot, I get Y return. They always gloss over change management, risk, and all that. And I think that's the part that's really hard. And I think that's the part that a lot of I think projects will fail because of that because they don't take that into account. The ROI and all that, they'll you know, they'll run the numbers. But it's also they put a robot in and people are like, no, I'm not, I'm like, or I'm using it, or I'm I'm moving the constraints somewhere else and it's creating issues.
SPEAKER_04Or my favorite one is they put it in the AMR that's supposed to be a optimized, you know, the robot to the person for picking, and instead they've turned it into a follow-me robot, or the person follows the robot around the warehouse, right? And you've lost your ROI completely because you're using the product completely wrong. And I think that's the that never happens, does it?
SPEAKER_03Never, never. Or you build an ASRS and it just sits there and it's it's just a tombstone. You're like, oh yeah, it's it's there. Look how cool it is. Like, why is that working? Well, so I think that's the thing that I think what I would tell people is that you have to, and I think this is where you know folks like you guys who have experience and who can help these customers look at the hard questions and not just say, I just want a robot because it's cool, but understand what's the end case, what's the risk, what's the impact on your business and your risk and your people. And I think those are the things that because I think if you go to the robotic vendors, of course we could solve that. Right. What do you need? I got it all, right? Right, right, right. There you go. You I think you go to some of the integrators, all they see is oh, we're gonna put in a three million dollar system. Yeah, we can do that. I don't care that I just chopped up your warehouse into a way that looks, you know, it's bastardized, but I made my money. Yeah, I think it's really, and I don't think this is the part I think the market is missing. There's who do you go talk to, right? Integrators, no, robotic vendors, no, supply chain, supply chain consultants generally. Yeah, but even they, I would say you go to a Deloitte or McKinsey or KPMG, they're gonna, they, they're, I think they're they're not there yet. They don't really know the the sort of down and dirty in the warehouse. So I think there's a gap there. I think there's a massive gap of who these customers should go talk to. I think some of the tier two consultants are starting to fill that, but I think there's a great opportunity for people like yourselves who are trusted to say, all right, just give me the inside baseball. Like, what does this look like? Like, what do I need to think about? Because I've told some of my customers this hey, there's like 700 OEMs out there. There's a solution for your use case. Right. Multiple solutions for your use case. But that doesn't mean it's the right use case or it's the right robot, right? That's where I think, that's where I think in the market for these customers, they need to be able to have that conversation. And I think it's still lacking.
SPEAKER_04Yeah, and to your point, I think companies like Breckham Group, comp companies like John Sandigates company, right? They're great resources for people who have been operators, they know the technology, they can really help customers figure out this journey. Yeah, right. And I think understanding the journey you're on is the important thing from an end user and not thinking behind the sky.
SPEAKER_00Yeah, absolutely.
Drones And Orchestration Software Next]
SPEAKER_00Yeah. So we're thinking future state now, right? So future, two years from now. What still exists, what's gone.
SPEAKER_03Pete and I are retired. We're gone.
SPEAKER_00Don't leave me here.
SPEAKER_03We're out of here. No, I got college payments to make. I'm not going anywhere in your top suit.
SPEAKER_00Yeah, so what's overhyped today that'll still be around? What's gonna be gone?
SPEAKER_04I don't know about what's gonna be gone. I mean, humanoids will continue to be overhyped for sure, because they're not gonna be ready yet. And that the hype is growing because more and more companies are getting into it. We just saw a company in in Boston area just open up. What the heck was the name? Um it's it's conquered.
SPEAKER_02Oh yeah, Walden Robotics.
SPEAKER_04Walden Robotics, that's it. Yeah, yeah.
SPEAKER_00Interesting.
SPEAKER_04Yeah, so I I think you'll see that continue to be hype. I I think what we're gonna see in the next year that's gonna become real is the stuff like Jacoby's doing and what Vecta's doing, where it's like, hey, you've you've solved two hard problems that nobody has solved before in unsequenced mixed case palletizing and entry-level automation for aisle case picking. Like those didn't exist a year ago, right? And now people are starting to hear about it, and they it modex started to kind of take off for both of them. It's gonna be real next year and it's gonna be you know starting to scale. And I think that's gonna be interesting. And I'm biased, but I like cool new technology, hence, I was in humanoids, found out it didn't work, and left.
SPEAKER_00Whoa, okay. I know, but solving spice it up a little bit.
SPEAKER_04Yeah, I know.
SPEAKER_00What do you think, E?
SPEAKER_03Uh, I'm gonna say, yeah, I'm gonna agree that it's overhyped on humanoids. I think we're gonna see in two years. Uh, one thing we didn't touch upon, I think, is more drones.
SPEAKER_00Oh, interesting.
SPEAKER_03I think more drones as they expand their battery life and usage, right? I think you see more drones in the warehouse for like cycle counting, security, and things like that. I think we're gonna see more drones outside the warehouse. Um, last mile delivery. We joked about this before. I think drones for returns, right? Um, I think you know, I look to our friends in Seattle, like Amazon is gonna have to start thinking about that. I think it's interesting. Um, I've years ago, I said kind of flippantly or tongue-in-cheek, like, hey, pretty soon, like in New York City or in Boston, right? If you go rent an apartment or buy a condo, one of the things they're gonna tell you is, hey, we've got a drone airport on this on the roof. So if you've got orders, the drone will bring it to the drone airport, drop it off, and then we'll go get it for you, bring it, and if you want returns, you bring it up there. The other thing I think we'll see more of in the next few years, and I think we're starting to see it now, is software to do the orchestration on all this MHE. Yeah. Right. And I know companies like Geek Plus and Locust are there trying to have all the MHE, right? They're gonna do goods to person, ASRS, AMR, but at the end of the day, I don't think it's gonna be a one-stop shop for robotics. It's still gonna be, hey, I want Vecton to do case bake, I'm gonna have an ASRS over here, I'm gonna have a you know, robust AI over here. Who's gonna orchestrate all that? And and throw in the labor, right? Because we still have labor involved. So now we're gonna have to have some type of, dare I say, control tower or some concept that's gonna orchestrate all of this moving parts within the warehouse. I think it is gonna use some AI to be smart about it and to learn. So I think we're gonna see more and more of that. I think you know, when I was at Six River, we were just starting. So, oh, we just put Six River robots, you know, just put Chuck in the warehouse. Now all of a sudden it's like, well, wait, truck's in the warehouse, but there's a right-hand robotic system over here that's gonna do, you know, some picking. So how do we orchestrate? Oh, and by the way, there's an old conveyor over here. You have to pull stuff off of that. So to me, I think in the next few years we're gonna see more and more of software that's gonna be there to help orchestrate all of this stuff.
SPEAKER_04Different than a WMS or WCS kind of layer?
SPEAKER_03I think a WMS is gonna stay as a system of record. WCS is gonna be the control layer. So now let's say it's a WES, which I think is also kind of like physical AI. Yeah, it kind of has five people, they're gonna have seven different WESs. But I think some version of what we think a WES is gonna become more prevalent. Is it gonna come from WS vendors? Sure. But it's also come, I think, from third parties, and I think that's where we're gonna see. I think, because I think you and I have seen this, right? When I started at Six River to compare it today, AMRs now are kind of ubiquitous.
SPEAKER_04Yeah, yeah, yeah. They're scaling. I mean they're scaling, right? Absolutely.
SPEAKER_03The coolness with a secret sauce now is like, nah, it's an AMR. Yeah. Do you want to follow my robot or swarm or whatever?
SPEAKER_04Other than that, it's Do you want material handling? Do you want picking? Like what do you want?
SPEAKER_03It's an AMR is an AMR. Right. Now it's still a lot of runway for that. But now the question is all right, I have an AMR and I have a VLM. How do I make sure they work together? Because the orders, I have an order coming in, I have five items in the VLM at six that the AMR is going to pick. I kind of want it all to come together at the same time. The truck's leaving a three. Right. So I think that's where we're going to start seeing more of in the next couple of years in drones.
SPEAKER_04The drone thing I'd push back a little on. I'd because I look at it and go, okay, drones were almost not, they were not the hype of humanoids, right? But they were hypey in the last couple of years, and they seem to kind of disappear on us all of the same. I agree. It's like, are they really gonna be used in a warehouse? Is that a good setting? Or is it outside doing other things? I'm not sure yet.
SPEAKER_03I think it's both. I think you're absolutely right. They were so hypey, and then they got overhyped by AI and then humanoids. I think the physical problem with the drones in the warehouse in the past, which we rarely talked about, is their battery life. Right? Oh, I could get a drone to go psycho count. Oh, by the way, it's only got 15 minutes of battery life. Oh, I've got a million square foot warehouse, it can barely touch part of it before it has to come back and I didn't realize it was that bad. It's not better, but it was, you know, I remember talking to some and they're like, oh, we can set it out, and it could be on station for like five minutes. I'm like, so you're gonna count anything. I think the other part too with the drones is and I this maybe tied into this is sort of camera technology, where it's getting better. Whereas when I first started seeing drones, and you really start asking the question, like, yeah, we kind of guesstimate. So we we cycle count because we look at a palette and we see that it's full, so we just assume it's full, right? Or we see a couple of things missing, so we guesstimate, but we can't really count how many items are, you know. So I think combine the technology with the drone being able to stay on station longer with better camera technology, and I think we're gonna start seeing it pick back up. And I think it's gonna be quiet, right? Because it's not as it like you said sexy, it was overhyped, and I think people were like ran all over it and they're like, oh, it's not that good. Now it's gonna come back. And I do think like last model, like Walmart's starting to get more and more into drone delivery, obviously Amazon. But I think we're gonna start seeing more of that as you know, us in the consumer world being like, Oh, we're in Boston Commons, I need some sunscreen. You have my phone address, drone shows up, drop, right out the DoorDash to the Boston Commons.
SPEAKER_00Yeah, I mean there you go.
SPEAKER_03So now two or three years, we're gonna have it'd be like that. Remember that commercial from Audi where it was like all the little drones were following it around, like this is when it was overhyped.
SPEAKER_04Yeah, yeah.
SPEAKER_03I think we're gonna start going back to that.
SPEAKER_04Okay, we'll place a bet. You on drones, me on humanoids.
SPEAKER_00Shot fired. I need to get a drone, a drone company on this podcast too, now that I'm thinking
Rapid Fire Takes And 2027 Bets
SPEAKER_00about it. But I mean, to close, I want to do a quick rapid fire. This is called shots fired. Shards fired.
SPEAKER_03Shots fired.
SPEAKER_00All right. First one AGV or AMR? AMR.
SPEAKER_04AMR, sorry. We're both on the same page.
SPEAKER_00Wow.
SPEAKER_04That's I just the whole guided vehicle. It's like gone. No.
SPEAKER_00Like, no, really? No. Sorry. We're smarter than that. This one may be different. Humanoid on a real production floor in five years, yes or no?
SPEAKER_04Yes.
SPEAKER_03Well, uh humanoid as we see them today, or humanoid bastardized?
SPEAKER_04Here's the thing. So we'll we'll dig into that a little bit, right? Picture Star Wars.
SPEAKER_03Yeah.
SPEAKER_04Right? How many humanoids do you see in Star Wars? Tons of them, right? They're in their servers in the in the bar area. There's C3PO, who's your butler. Yeah. You see industrial ones, buildings, stuff. I was gonna say another word. Um right, so I think we're gonna see that, and we're seeing it now. You got wheel-based, you got leg-based, you've got hand grippers, you got change the end effectors. We're seeing it, right? I think the reality is though, the number one company in that space is Boston Dynamics. They've been doing it longer than anybody else, right? Their robot is going into production in a proof of concept, kind of right now. Yeah, in three years, it's gonna be humming down there. They're investing tons into things. I I think they win the race, by the way, for my two cents. It should have been my bold prediction earlier. Who wins the humanoid race? Boston Dynamics, Atlas. Atlas is the most amazing humanoid out there today. It's performative, it can lift incredible weights that none of these other ones can do. It's nimble, and I love that it changes its own battery. Going back to batteries.
SPEAKER_02That's a big deal.
SPEAKER_04The battery lives in these things are like two to four hours, depending on what they're doing. Everybody else, a person has to touch it or it has to back up to a charger and charge itself, which means you need two robots to do the job, not one. Right?
SPEAKER_01So it goes your MTE costs.
SPEAKER_04Yeah, there goes back to the CFO. Uh, but when Atlas pulled his battery out and put it down, put a new one in, I was like, dang, that's cool.
SPEAKER_03Yeah, all right. But let me count on your Star Wars example. How many ro how many droids do we see on wheels zipping around? Oh yeah. Oh yeah. That's true.
SPEAKER_01R2 D2 does not look it looks like a trash can. He looks like a trash can.
SPEAKER_04Did he really do much though? He saved people. I guess.
SPEAKER_01He saved people for being crushed in the in the compounder.
SPEAKER_04But he couldn't pick anything up because he's the one who couldn't. We're in warehouses. We need manipulators.
SPEAKER_03Well, yeah, but R2 D2 can like run all the machines and all. He can't be in the WC. See us. So I would I would argue Okay, but so so but I think the humanoids the humano everything you just said. Is it gonna be in a wheelbase? Is it gonna do that? I think that, yes. So we're gonna take the humanoid, but we're gonna decouple it to different ways.
SPEAKER_04Yeah, yeah.
SPEAKER_03So is it really a humanoid or is it now just humanoid aspects that are optimized for a specific job?
unknownOkay.
SPEAKER_00Last one. Just quick rapid fire. What is the robotic solution or automation that everybody's gonna be talking about in 2027?
SPEAKER_03I'm a hate to do this for you, but go for it. I think like like I think what Vecna is doing, I think is pretty cool. I agree with P was saying earlier. I think the biggest issue we've always had is case pick and leveraging it. You know, when I was at Six River, it was each pick using the follow the follow the me follow me solution. Yep. I think now if we could do case pick, I would even say, and this is a push to Vecna, if you can also have it do case pick at multiple levels, that's a game changer, right? Because I have so many customers I talk to who that is That's hard. I know it's hard, yeah. Putting it out there, like I think if that's at least picking at the first level is great, but now if you can take it the next literally the next level, I think that's gonna be a true game changer because to me, that to you probably repeat, least disruptive in an existing warehouse, addressing a massive problem, which is case pick, but we're not changing aisles or formats or putting stuff in an ASRS, right? Now you're just putting these robots in there, one human attached to it, or in a zone, and you're picking at a case level. I think that could be really interesting because I think it's something that we just haven't seen done properly. Right. Um, the AMR is doing each pick, as we said, I think it's ubiquitous. We've got plenty of options, we see it all over the place. Now the case pick is the next level. Yeah, yeah.
SPEAKER_00And quickly, Pete, because I'm getting the wrap it up from Katie over there. So, what's what are we gonna see in 2027?
SPEAKER_04To me, it goes back to the mixed case palletizing. For years, it's always been this sequence was the only way to do it. Super impactful from a cost perspective. Very few companies did it, right? You sell one or two to a company, that's all you ever do. But everybody in the industry has been looking to solve unsequenced mixed-case palletizing. The largest integrators on the planet are trying to solve it right now. You know, the Dematics, the Korbers of the world, you know, Fortnite's of the world, Honeywell Intelligrated. Um, and now with Jacoby software being here to do that, I think that we're gonna see scale it crazy next year, particularly because the integrators are looking at this saying it's a game changer for my design. I can lower the cost and lower the impact on infrastructure so the ROIs are better. So I think we that's where we see some impact. Now I'm biased, of course, but I do think that is a huge game changer in the industry.
Closing And Final Shots Fired
SPEAKER_00Yeah. All right, so we're gonna go for lunch at 21st Amendment, which is another Boston thing, right? So thank you, Pete. Thank you, Guy, for coming to the podcast today. Um, thanks everyone for tuning in. Mandy Dwight, Pete Allen, Guy Corton. Um, hope you enjoyed Shots Fired.
SPEAKER_04Shots Fired, Shots Fired.