Lone Robots Fall Short, but Fleets Scale Warehouse Automation: Robot Orchestration Makes It Work

Alright. Let's get going, everybody. Hi. Welcome to our monthly webinar of and this year's series, Robot versus Wild. I am Josh Kovanko, chief marketing officer at Vekna Robotics, and by joined by yet another trusty pal, this time a very special guest, our chief technology officer, Zach Diedek. Zach, how are you doing, man? I'm doing good. How are you? I'm doing great. Doing fantastic. I'd like to announce that if we get forty people on this webinar, everybody wins Chick fil A. It's amazing. Did you watch the NBA playoffs, Zach? Are you an NBA fan? I am not. Did you see that? I thought it was a remarkable contest that if the opposing team misses two foul shots in a row, the whole crowd gets Chick fil A. Wow. Imagine after one missed foul shot, how the how the crowd was in the beginning. It was great. So I figured I'd do my own my own, take on that if we get forty people. Everybody gets Chick fil A. Isn't that great? Although I'm in Canada right now, I'm in the great white north, and there's no Chick fil A here, unfortunately. So we'll have to contend with that. But let's get going here. Yeah. You'll you'll have to wait. You'll have to wait. So, let's get to it. The title of this episode, I love this one. And for those of you that are Game of Thrones fans, you'll, you'll you'll like this title. The lone wolf dies, but the pack survives. And, while I love the Game of Thrones and all that kind of stuff, I think Zach's a fan too, we actually have a serious topic to cover today. And this is why we've brought up big guns, and Zach to talk to walk us through it. So let's get let's get going. And before we do, let's just reintroduce ourselves or remind ourselves what we're talking about this year. What is this whole robot versus wild concept about? Well, you know, let's just start from the beginning. There's a lot of technology out there to deploy in warehousing and manufacturing, and and a lot of it is fantastic. Just excellent technology that can be deployed today. But a lot of it really is not ready for prime time and, certainly not ready for prime time to be deployed at scale. And and that's our claim, and that's that's why we're out here every month, attempting to, educate folks, as to the benefits of, certainly horizontal transport transport, but also ground to ground where we focus our time. And while we hear great stories in other domains and other technologies, like, high reach and trailer loading, unloading and whatnot, humanoid, certainly, that technology needs a little bit, of time to cure and, and mature before it's really ready for prime time because this series is really all about, maturity and scaling. But why? Why are we talking about, robot versus wild and and in in the spirit of scaling, automation? Well, this is a report from Gartner, and, around eighty, eighty five percent or I guess, eighty two percent, I should say, of the market is in various stages of of initial deployment, okay, or fully deployed. Right? So the market is really headed in that direction, of maturity. This is a, intra logistics automation, so all the various types of automation that are being deployed in the warehouse. So we're we're all heading there, and and different technologies, as I've stated before, are at different stages of maturity, and ground to ground is certainly there right now. So what what's the narrative here? So you've you've you've started your your your deployment. You've either fully deployed or first deployment or you're beginning or you're piloting. Right? So you've deployed now what? And that's what this series is all about, the now what? Great. Always great to start things, but, you know, you really have to have an end in mind, and we have actually have a road map for that that we'll talk about in a sec. But we really wanted to focus on, okay, I've deployed three or four or two or three or five or six robots, but then I really need to get into the the meat of it and really start seeing value. What should I expect? Right? And then, before we get to today's topic, I you know, we we we've talked about this for the last three episodes. What really are the pitfalls, the common pitfalls to scaling automation? And, they are, you know, poor requirements def definition, and making sure your IT infrastructure is ready. We, you know, we we see a fair fair amount of that. Underestimating the need for flexibility, a big one. Right? Change management and flexibility. Not only flexibility of the technology, right, which in many ways Zach is responsible for, but also flexibility of your operation to handle how the technology works. Support is a big one. Changing project scope. If you were here a few weeks a few months ago, you heard David Rabinovic talking about, the warehouse as a as a living, breathing organism and how it changes. So underestimating that cultural factors or environmental factors can change in how your automation choosing an automation vendor that can handle that. Goal alignment in particular between head office and sites is another one that we see that that prevents scale. And then, I you know, I've mentioned I've mentioned change. So you know? But not just change within your warehouse, but just changing within warehouses in general and the needs of the market. So that's why we do what we're we see in terms of the common pitfalls, and we're we have this series to maybe dive into each one and how you can avoid, avoid these pitfalls. So the lone wolf dies, but the pack survives. Zach, you know, my first question to you, I guess, we'll get into maybe some examples, but, like, what was on your mind when we when we came up with this idea, of the lone wolf dies and the pack survives? Yeah. I mean, it's a good question. The the thing that's been on my mind for a while now is, you know, once you once you've got the robot working you know? And those blooper reels of the humanoid robots are always hard to watch as a roboticist because, you know, robots are hard. Getting the robot working is is a a feat in and of itself. And a lot of smart people have been working on it for a long time. You know, that last robot that fell off the car, that was MIT's robot, you know, trying trying to get out of that vehicle. Sorry, buddy. I thought it was I thought it I thought it was Carnegie Mellon's. I didn't know it was MIT robot. Sorry about that. CMU had just as much trouble, you know, back then. We've come a long way, you know, and and, you know, especially, with some of the technology that that we're working with now for horizontal movement. That kind of single robot part of it, is, you know, in good shape. I won't say solve. Yeah. But it's you know, that's mostly working. Right? The robots can get from point a to point b. So you really once you have that piece, you really quickly start thinking about, okay. What what comes next? Now I have the ability to go from one point to another. You know, that by itself doesn't provide a lot of value. But as you start thinking about scale, you start getting, you know, more of these robots into a facility. Now it gets really interesting. Right? Now this is where this is where a lot of value comes from that, you know, having more than one of these robots working at a time. Right? So that's that's where that's where my mind went right away is is, you know, looking at the things that we've been doing over the last, you know, three to five years, around fleet orchestration and optimization of the the collective of robots and not just the single robot problem. So it's a really interesting topic. So if you had to sign off now out there in in webinar land and you had a meeting to get to or you had a lunch to get to and you just wanted the takeaway, my assertion is this. The the overwhelming benefit, not only from a business value perspective, but from a technology perspective, they're intertwined, of course. But from both perspectives, the overwhelming value of this technology is is achieved via scaling. Right? It's not onesie twosie. So let's peel that back now. So let's start with a comparison. What do we mean by lone wolf and pack? Let's go through some silly examples just so we're all oriented and maybe we can we can have have some nostalgic moments while we're at it. But let's let's just make sure we're understanding about what we mean by lone wolf and pack. Okay, Zach? So first example, lone wolf and pack. Here we go. Fairly straightforward one. I can tackle this one. Right? This is, like, just a laptop or a desktop. Let's just assume that maybe it's not connected or even if it's connected. It's sort of a lone wolf device. It's on its own. It it serves the needs of one person, and, regardless of whether it's connected. It's confined by its own hardware, right, inside. Former Dell guy. Right? I used to sell this for a living. But, really, it's a it's a lone wolf device in many ways. And then you have the cloud. Right? And that's that's a really, I think, obvious way to compare, you know, compute. You know, compute for an individual, a personal computer, and compute in the cloud, which is, you know, I guess, in theory, relatively infinite, in scale and can do, so many more things and so many more dynamic ways with with, I guess, physically speaking, Zach, not unlimited headroom. But I guess as far as humans can understand, it might as well it's asymptotic to infinity. Right? Yeah. I mean, I think the the best way to think about this example for me is Yep. Imagine what it would be like to to buy a laptop today that you know, didn't come with wireless connectivity or or Internet connectivity. It's just that, you know, you have to plug in the USB stick or a CD and install your programs and then use them. I I I can't imagine that, you know, in in in modern modern times. And that's because all the value is happening, you know, in the cloud, and and we're connecting to it. And the scale of the cloud, the scale of compute and the flexibility. Right? And, you know, remember we talked about how how warehouse automation programs can fail and scalability and flexibility is really sort of part of that. So okay. I think we everyone's got it. Let's go for another one here, another interesting one here. Oh, movies. Right? Another one? Think you should see the theme here. So the old style I love old style movies, Alamo Drafthouse from Austin, Texas. Right? Just got bought by Sony, as a matter of fact. But, and movies are even getting better. I love movie theaters these days, so lots of fun. But clearly a difference between that experience and what you get in Netflix, right, which is a massively scalable platform that can deliver on demand and, offers up suggestions too. Right? So pretty obvious example. Zach, anything to add here, or did I nail it? You're, you're really giving us a a window into your viewing habits with the with your home screen from this list there. That's brave. Yeah. Yeah. And I don't and I don't share I don't share my Netflix with anybody, of course. Don't know what some of these are. In fact, I don't even think I watch I've watched Netflix in many months. But, you know, maybe I should watch Bridgerton. Maybe Bridgerton. But, anyway, yeah, I am giving you a glimpse pretty bold. Let's go here another one. Oh, this is an interesting one. This one can relate to, I guess. You wouldn't really think of an iPhone as a lone wolf, but it kind of is. Same reason as the laptop, right, versus the apps that is the real value, right, behind this. Maybe this is hits a little closer to home, Zach. Yeah. I mean, it it's a little bit similar, and we've seen this, you know, happen in a number of different technologies. But, you know, you build the operating system and you build the the device. Right? That's the phone part. You know, but where you get that, the value is really in the ecosystem that's built on top of it. Right? The App Store and all the developers, you know, across the world that are building, these features on top of that operating system. Right? You need both pieces, you know, of course, but we didn't really see, mobile phones become what they were today until you had you had that operating system with the App Store. Right? That was the that was the catalyst. So it's been really And the app's getting better and the app's getting better not not because coders got better or not because people had brilliant ideas, but also because they're capturing the combined intelligence of how people are using all this software and then reapplying that back in to make to update the software more frequently. Right? Exactly. Right? It's incredible. So it's yeah. Like a silly example here. Right? I think we all get it, but I think I think since this market is still, I guess, very much in early innings, we we may forget. These aren't just dumb pork lips. Right? There's the combined intelligence of the fleet that really dev deliver that massive value at scale. Okay? At scale. Alright. Let's keep going. Oh, love it. How can we have how can we have a presentation without tipping a cap to RadioShack? I love RadioShack as a kid. Didn't have many here in Montreal. Go ahead. Poor RadioShack. It's tough it's tough to be to be on that side of the yeah. But that's right. That's the lone wolf. The lone wolf. Right? In many ways. And then, obviously obviously, the big bad bad Amazon. Another glimpse here, Zach, into my, my account. I don't know if you could see you could see some reordering, some sunscreen, or whatever. Yeah. Yeah. Somebody's somebody's trying to sell me some Stanley's. They know I have a fourteen year old girl, daughter. And she has one. That's enough. Right? We don't need we don't need any more Stanley's. And I think I need to order some, I think I need to order some air conditioner filters as you can see right here. I think it's time to order some air conditioner filters. I mean, it's funny, but but that's that's one of the key differences, right, that's highlighted right there is is that Amazon knows what what you want to see, and they're putting it right in front of your face as opposed to, you know, RadioShack, the storefront. You know, it's got everything there. It's all, you know, hidden inside the store somewhere. You know? Maybe you find it, maybe you don't, but, it's not being surfaced to you because there's not that layer of kind of intelligence and the network that's bringing all this stuff together to you. So it's a good example. Yep. We'll skip this one, I guess. You know? I kinda got it. Uber and and taxi. Can't we'll we'll we'll get off this now and but but I'd be remiss if I didn't have a Star Wars reference if we're talking about the future and robots. Right? And we're talking about two different types of robots here. So this is really close to home, Zach. Maybe too derivative, but I think we all get it. Right? The the the I have a I have a a Roomba, works okay, battles with a a a corner plant a plant in my in the corner of my house, every day, and the plant loses every time. And, and then you have an army of droids here that are are, you know, significantly order of magnitude, more sophisticated. And that that's really what we're talking about. This is really what we're talking about today. Right? Yeah. I don't know if I wanna be put in the same category as the army of droids, but but I I see the see the point, though. It's it's good. Twelve seconds. Okay. Okay. Well, we don't need to get into battle references. Okay. So let's get into some serious stuff here. Right? Let's pivot from the silly, which is always a little fun, to to we're talking about warehouse robotics and scaling it and the benefits of of scale, the literal benefits of scale. Right? And I think it really starts with this, which is that the robots are connected. They're connected devices. This is table stakes. Right, Zach? This is this is where it starts. The robots have to be connected. So let's walk through why that's so important. Yeah. I mean Great. There are a couple key pieces here. Right? Yeah. Not only are the robots connected to each other, but they're connected to the facility as well. Right? So they're they're plugged into all the other technology that's there, which is, you know, increasing, you know, besides automation. You have more technology in the warehouse nowadays. Right? WMS, you have, people with wearables. You have people with, tablets. You have all kinds of technology there, you know, doors and, conveyor belts and all kinds of things that are now part of this network. Right, so being connected is is not just, you know, connected to each other, but connected to this broader ecosystem and also connected external to the outside, which I think we'll talk more about later. Right. Really, really interesting. I never thought of it that way, but it's like, it's not just connected to some random cloud that makes them go. There's a whole network of connectivity that's happening there. It's not just about mission. It's about mission. It's about interconnectivity with robots, which maybe we'll talk about that in a future show, with the humans on the go. Go ahead. This is the big observation with the Internet of things. Right? It's that Yep. You can have not just not just devices that you know, or computers and and whatnot, but you can actually have physical devices that are out there in the world moving around that are part of this Internet of things. Industrial IoT is is something that, you know, we fit right into nicely. Yeah. And, and we mentioned a previous episode that this that these warehouses are living organisms. How amazing is that these devices are connected into that living organism, right, in many ways? You know, you see charging. You see, you know, you know, not only that, but connected into our network so that we can monitor them. So much greatness. So much goodness. And then as the fleet expands, the network and the web, right, all connects itself. And there's some great benefit to that, which we'll we'll get to in the future. Here, well, I think we're gonna get to it right now as a matter of fact. So they're all connected. Right? And and and, and, of course, we're really just talking about our robot, in that environment, in that, in that ecosystem. But, let's not forget the gold here. Right? Some of a lot of the gold is this. So what is the gold here, Zach? I mean, the the gold is is data. Right? Information about operation of not just the robots, but the warehouse as a whole, which, you know, we can provide with our system because of that connectivity. Right? There's a richness to this data that you know, I've had conversations with with people in the space that have just been blown away by, you know, what they can see happening inside their facility that they didn't have access to. These are the two facilities. Right? Either two facilities? These are two facilities laid out differently, you know, both warehouses. But, we see things lighting up as hot spots or places where, in this case, it's where robots are stopping. There are a lot of different ways to slice this data, but, this is a good one that we look at a lot with our customers, and we, you know, look for operational, issues that that show up here. You know, maybe, you know, I've had had situations where one of these hot spots comes up and we point it out to the customer, and they say, oh, that's just where the janitor leaves their cart. And it's like, no. Great. Let's let's have him put it to the other side, and then the robots come through here nice and smooth. Right? But you have access to this kind of bird's eye view of the facility and everything that's going on inside it, because that data is is coming up, from the from the robots. This is mind blowing to me, by the way. Right? When you compare just like a standard AGB or or and and maybe we don't need to wait into those waters. But if you compare just a manual operation, which most of the customers that we deal with are moving from to this. So it's not just about moving from a to b with because I can't hire a worker to do that job. Right? That is part of it, and it's a it's an important part. It's a here and now part. But, like, the overwhelming benefit, I I would argue the portion of the, of the iceberg beneath the waterline here is are the insights and improvements you get from this. Right? From sitting with, a technologist and your customer success manager and reviewing the heat maps and saying, okay. If I'm gonna deploy ten more robots, how do I make this whole thing work better? Right? Yeah. Dare I say a hundred more robots or a thousand more robots. This is vital. Right? This is vital. Okay. Love this. Love this. Alright. Let's keep going here. I think we just you basically just covered it. When you have the data, right, what do you get out of this? And, you know, why don't you comment on on this? This is sort of something that we talk about every day. Maybe we take for granted too much. Yeah. That that's right. I mean, mostly, you you know, what I talked about earlier was kind of operational efficiency, but there's also a really interesting feedback loop, on the software side of things. Right? The data that we collect from the fleet, you know, comes in and makes the whole fleet smarter. Right? And that and that comes as an update over the air, back to the fleet. And, right, there's a there's a really nice, a cycle there, feedback getting in and making the robot smarter. Right? So that that's the that's the benefit, really, for this connectivity as far as, you know, software performance is con is concerned. And in the way that I talk about it, which is obviously very dramatic, is, when you have a forklift driver, who's experienced, you get the combined experience of twenty years of that forklift driver, which is worth money, right, in terms of efficiency. When you have a fleet of robots, you have the combined intelligence of every robot in that fleet that is that is making that individual robot and the system better. Okay? And so we're talking about step functions of of improvement. Right? And you're talking about the improvement that an individual lone wolf can get through their experience. Right? And humans are underestimated as Elon said, and and they are. They're highly complex individual, contributors. But when you get to the combined benefit, the accrued benefit of a network of robots that are all operating in unison and improving each other massive massively powerful. Massively powerful. Okay. Let's keep going here. Just got a few minutes left. Oh, my favorite. Always love always love a little chart, little graph. Shouldn't be too complicated. We've got we've got throughput throughput on the y axis, number of robots on the left. And what this is saying is the more robots you add, both AGV and AMR, deliver deliver increased benefits. Right? They have slope. They have their slope there. Here's the the CTO is like, what does the marketing guy know about slope? Right? How does he know what slope is? That's your calculus. But yes. But the slopes aren't the same. So, Zach, tell us what's going on here. Why aren't the slopes the same? And and what's and and what are we trying to convey? Yeah. I mean, it's similar to to what you just talked about. You're looking at people versus robots. So this is specifically AGVs versus AMRs. Right? The AGV is the is kind of the lone wolf in this scenario. Right? That's it's installed, and it does its work. You know, it's very precise. And if you put two of them in there, you can do two times the work. Right? But you don't get that network effect, which is the the benefit of having that fleet and having that feedback loop coming back in. So as the number of AMRs in the world, gets higher and higher, the the rate, when we go into calculus even more, right, the acceleration is is going up. Right? So the second derivative is going up. Up. Right? That's the network effect. That's what's causing that, that green line to to shoot up into the right. So that that's that's really a big difference there, is the connectivity is that's the it's the pack. That's the pack part of it. Right. And, I would say the spread just by definition, we are asserting that spread gets more severe at scale. That's the whole idea here. Right? So you may not see it one for one, everybody. Add two robots. You may not see it. Look at the spread when you have just a few robots. Right? And you may be like, ah, you know, it's no. These AMRs sort of better than when than when we had an AGV here. Right? But we're asserting that at scale, that's where the massive spread happens. Right? And I think we're all waiting and trying to motivate that that spread to, to, to deliver. Okay. Let's keep going here. We hear a fair amount. Wall Street Journal and others that are publishing these ideas about robot wranglers, really interesting concept. Certainly, at early days, you need people to help make sure the system is getting to, higher utilization. You need training, monitoring, and stuff. But, you know, at scale, is this a sustainable setup, Zach, to have robot wranglers monitoring and man supervising robots in the field? Yeah. It it it doesn't make sense at a certain point. You know, you really you can't have it be, you know, close to a one to one robot wrangler to to to robot. Right? That defeats the purpose. Defeats the purpose. And the the guy should get off the horse and get in the truck at that point. Yes. You know, you you really need to get to a point where, you know, it's someone who is already in the warehouse that's doing some other jobs and maybe a small fraction of their time is is spent taking care of the robots. You know, not not as a a full time job or multiple full time jobs, you know, managing that fleet, all day every day. These these are a small portion. Right. Right. And that's this is this is true, but it's also self serving. We remotely manage all of the robots in our fleet twenty four seven, through our pivotal command center, and we assert that that's the more scalable way to to run this. And, and, and there's also data benefits and and so on and so forth. So, you know, in case you you you've been influenced out there by this this concept of robot wranglers and, you know, the as the future of automation, we're asserting that it's not scalable. Right? It's not scalable. And and you won't see that spread as we talked talked about before if you're gonna go down that path. Okay. I think this is our last treatment, if memory serves. So, orchestration. So, I I mean, I think this is sort of the higher order benefit of it's not just about the robots working. It's not just about them maybe being connected together or into their environment and and and and getting software benefit, although that that is certainly, useful. It's not just about remotely managing them. There's even a higher order, I would argue, value of a of a system that's well run. And I think this is what we're talking about here is orchestration. Right? So just, maybe conclude with a little bit on what we mean by orchestration here and then quickly on these three examples, Zach. Yeah. I mean, on the on the left, you have the, you know, the football team with the set of plays that they're gonna run through. Right? And, you know, they've they've practiced that scenario and then, you know, they can do that, well, well, depending on the football team that you support. You know, maybe better or worse, but, the idea is that they're they're coordinated within that framework. They can do that thing, and maybe some, like, you know, slight reaction to it. In the middle, you have, like, a beehive, which is obviously distributed. Right? It has an a sort of intelligence. Right? Mhmm. The bees themselves are are pretty basic, but they have a sort of behavior that that, you know, lets them accomplish a a bigger goal, you know, and be a little bit robust to, you know, weather conditions and, you know, things like that. But, ultimately, it's kind of limited what they can accomplish with that. Right? They can they can make honey, which is great. Maybe pollinate some stuff. That's that's good too. But on the right is is kind of the, you know, what is being built now in terms of the technology that we've been talking about. You know, very sophisticated technology in the cloud, connected devices, opening up possibility for, you know, basically, unlimited configurations of physical and cyber components that come together in an optimized way to to build these really complex emergent behaviors, which is really interesting. It's a it's an interesting time to be alive to see some of this stuff come online. It's that's pretty cool. And and I think Zach's being modest. This is what our system does. At the end of the day, the massive value at scale, once you talk about all the things that we just laid out, the ability to this is this is the mind blowing part for me. The ability not just to orchestrate the work and improve the individual work of each individual robots, but then be able to, coordinate the work of all of them in a warehouse or a manufacturing facility, in in an optimized way. Right? Independent of in of individual software updates or whatever. Right? Like, assuming the software is there and it's as good as it's gonna get, then being able to optimize the work along with the WMS. Right? That's the orchestration that we're talking about here. And, and my assertion and Zach's assertion here is that, the ability of compute to conduct this work is nearly unlimited. Right? It is nearly unlimited in being able to, in real time, optimize the work for maximum output, and then it can continually improving over time. Peyton Manning's awesome. Right? Peyton Manning's awesome. But, you know, there's there's there's a limit. There's a limit to what the queen bee and what Peyton Manning can do. And the orchestration level, certainly, when it comes to AI, of course, as well, is difficult for for for ninety nine point nine percent of humans to understand. Right? But that's how our robots, deliver massive value. So good. Love it. No. I always love pig manning. So let me let me conclude here. Thank you, Zach, for, for a little bit of a lesson here. Let's just land the ship. Most of you are in that early stage thinking about deploying proof of concept, thinking about scale, and we have a road map, that is flexible for whatever speed that you wanna operate in to adopt robotics at scale. We've got clear stages, discover, design, deploy, learn, and scale to get you to scale. We got very simple steps. Our CSMs and our sales teams walk you through it. We try not to make it overly complicated, and, and so you can trust that, you know, if you're gonna go with the vendor, just dropping a robot in and pressing a button and making it go is really not how it works. You want somebody who makes it easy but has a road map, and and we're very confident that we have the road map that wins. If you're interested in this, you like the RevPAR T, you wanna watch some back episodes of the Mythbusters or Robot versus Wild, We have others as well. Go to our resource hub on beckon robotics dot com. Also, look, Clint does a great job of curating our social media feeds, and, and we've got all of this content and recordings and workflows and interviews and press releases and stuff always, going on there. We got two two announcements every day on LinkedIn. Please follow us there. We spend a lot of time there with user support, and, chime in. Let us know if there's anything else you would want us to cover. We're happy to, to do that. So with that, I just wanted to thank everybody for your time. Zach, hope you have a great weekend. Thanks. Any parting thoughts for the crowd or, you said what you gotta say? I think I said it all. I mean, the the lone wolf dies, the fact survives. That's that's what it's all about. That's what it's all about. So, so on behalf of everybody at Vekner Robotics, wishing everybody a great, weekend and Canada Day coming up, Zach. I know. I don't know where your little Canada flag is. Yeah. I'm sorry. And then, July That's your Yeah. Yeah. Fine. I thought I left one on your desk. And then, and then July fourth as well. And on behalf of everybody out there who's watching this and every series we've ever had and every person at Vethna and every person who's ever worked in Vethna and every person who's ever worked in robotics today and ever, happy birthday to Zach. Have a great one, everybody. Happy birthday, Zach. Yeah. Thanks. Bye. Bye. Later.