For links and show notes, mouse over the player and click the 
.
Transcript
Brandon Coats serves as an executive-level engineer and product manager at Material Handling Systems (MHS) in Louisville, Kentucky. Coats has broad engineering and solutions development experience, from implementing designs as an individual contributor to leading large engineering and cross-functional teams on major business strategies. In his current role, he is responsible for developing material handling systems that leverage robotics, sensors, machine vision, AI, and other exponential technologies.
David Maloney, Editorial Director, DC Velocity 00:01
How can artificial intelligence drive warehouse operations? Can spending on new technologies boost supply chain resilience? And a warehouse automation project completed in less than 90 days. Pull up a chair and join us, as the editors of DC Velocity discuss these stories, as well as news and supply chain trends, on this week's Logistics Matters podcast. Hi, I'm Dave Maloney. I'm the editorial director of DC Velocity. Welcome. Logistics Matters is sponsored by Honeywell Intelligrated. From system design and emulation, to integrated warehouse automation software and technologies, to AS/RS shuttles and robotics, Honeywell Intelligrated's end-to-end solutions address the most pressing e-commerce and labor challenges facing our industry. To learn more, visit sps.honeywell.com. As usual, our DC Velocity senior editors Ben Ames, Victoria Kickham will be along to provide their insight into what's going on this week. But to begin today: We've all heard the term "artificial intelligence," but what exactly is it? Our guest today feels that there's a lot of misconceptions around AI. To find out what it is and how it can help warehouse operations, here's Ben with our guest. Ben.
Ben Ames, Senior News Editor, DC Velocity 01:19
Thanks, Dave. That's right. Our guest today is Brandon Coats. He's the global product manager for robotics and vision at the systems integrator MHS. Brandon, thank you for being with us.
Brandon Coats, Global Product Manager, Robotics and Vision, MHS 01:31
Yeah, thank you, Ben. Thank you for the opportunity this morning. It's a pleasure to be here.
Ben Ames, Senior News Editor, DC Velocity 01:37
Brandon, you were on a panel this week that made a presentation at ProMatDX—that's the virtual online version, of course, of the major ProMat trade show—and the session was titled "Unleash the true potential of robotics with AI," and in that session, you said that many people have learned about AI from Hollywood movies like The Matrix or I, Robot, where it was used to enslave humanity and other dramatic uses. But it's actually a practical tool for warehouse automation. So, to start off, maybe you could help define what AI really is, and what it means for warehouse automation in general.
Brandon Coats, Global Product Manager, Robotics and Vision, MHS 02:16
Yeah, sure. So, I'd say it's not just a practical tool, but it's something that's pervasive at this point. It's something that's quickly growing. You know, looking at the last 10 years of where AI really didn't mean anything a decade ago to what it means now, it's, it's throughout automated systems that we see today. So it's pretty much everywhere, and then it's going to be more the case that it's going to be pervasive, ubiquitous, and everyone's smartphones, and all of the material handling and automated systems of the future are going to necessarily have AI as a cornerstone of the product. So, if we look at what is AI, very simply put, I would describe it as any level or any amount of programmed intelligence. So, what does that mean? It's all of the traditional software that was there before, any of the logical statements within the software in order to make decisions according to data that's around the system in the peripheral devices. But then we have this other form of AI, that is, has a lot of hype around it, and it has a, you know, very quickly growing base. And that's more of the artificial intelligence surrounding convolutional neural networks and the like. What these are able to present in terms of artificial neural networks, are supervised learning, unsupervised learning, and even reinforcement learning, which itself presents a feedback loop between the ability for this, let's say, unstructured algorithm to go in and then take inputs and then optimize according to what the output condition of that scenario is, that you're looking at within your automation field. So, if we jump in even further, convolutional neural networks, we've got generative adversarial networks—these are quickly coming up—in which you basically gamify this convolutional neural network process, artificial neural network, to where it can self-optimize over time, adding small permutations to the parameters within it. And then you can go very quickly, training this on a GPU or a GPU on the cloud, and you can optimize at a rate that before just wasn't possible. And so it means that you can have a level of optimization for your automation technology, when it comes to this algorithm that helps you make these nice decisions on the fly, and it can happen very, very quickly and at low cost. Of course, all of this is predicated on the idea of Moore's law and low cost computing and having lower and lower cost computing coming on time and time again, year after year with a nice cost the demonetization. If we look at it again, like very high level, it's the ability to detect patterns and complex datasets. So, in essence, AI is the ability, when you leverage it properly, it's going to really start to mimic the ability of the human mind to some respect. Not so much to the extent of say, like, I, Robot. I'd say that that time is coming, and that level of sophistication, but the way we use these tools, overwhelmingly, they're for the positive impact of our economies and of our lives. So, that's how we'll end up using these tools, and that's how this will play out. But the ability to detect complex patterns within a dataset is very, very important, because then it allows you to make real-time decisions on data within your automated system. And within these warehouses, more and more of the decisions going forward will be real-time decisions versus programmed software that is just fixed, and it stays static from Day One of installation.
Ben Ames, Senior News Editor, DC Velocity 06:05
Yeah, that makes a lot of sense, the ability to detect patterns and to self-optimize over time. I think those are some concepts that might seem familiar to a lot of us human workers in our own jobs, maybe. You made the point that AI is also capable of automating processes that have long been targets for automation, but they've been too complex in the past. Could you give us some examples?
Brandon Coats, Global Product Manager, Robotics and Vision, MHS 06:30
Yeah, sure. So, if we look within a typical warehouse, we can take just a few scenarios, one of them are going to be, like, inventory control. So, where you might use inventory control might be within the buffering system of the warehouse, or like an AS/RS—automatic storage and retrieval system. For a long time, we've been taking the velocity of product going through the AS/RS, and we've been moving the the fast movers towards the front, that are more readily accessible, and then we've been taking the slow movers and placing them towards the back. With AI comes the addition to that, that's not only it's more optimized, it's better at that capability, but you're going to get other capability that comes as well. So, one idea is changeover. You have a lot of different SKUs that are going through your warehouse, you're an e-commerce business, you're going to be turning over these SKUs, maybe, like, once a month could be pretty typical. You could also go through and do it down to the week or to the day. So, you're just gonna have a lot of changeover of the inventory that you're actually putting into that system, and you need to be able to keep up with that. So, AI is going to do an outstanding job when it comes to that regard. And then if we dig into another aspect, here would be like order batching, or order execution, the batching up of the orders is often predicated on the equipment that's within the facility. Each piece of equipment has its own function that it's going to add to the overall puzzle, and if you're not communicating between these disparate pieces of equipment and technology in the best way, then you're going to have natural performance losses that come into play. So, one idea here is the idea of buffer from one system to the next. It might be easy to go through and program that buffer in terms of very, very specific usage of SKUs and which SKUs you're going to use within the buffer, how, and when. But then very quickly, that can get too complex for a typical programming strategy. And so here again, we can use AI to go in to optimize the buffer based on the SKU that's available within the system, and then as you look at the total workflow from one piece of equipment to the next within the warehouse, it can become a lot more optimized and a lot more clean in nature. And then of course, we've got order fulfillment. So, order fulfillment today, you know, it's being done a lot by labor. We see this transitioning, though, towards robotics very heavily over the next two to three year—let's say, two to three years. And then it's going to hit a point where it's expected that all of order fulfillment is going to be done through robotics. So, with that, you know, people are, we're very dynamic. We're able to, with our hands, go out and grab something, move it around, build the order up. We don't really have to think twice. But we're also prone to error. On the other hand, robotic application, once it's programmed, and once it learns the environment, then it's not going to have that same level of error within it, and the traceability is going to improve as well. So, what is enabling robotics today within order fulfillment, the this type of scenario, it's really AI. It's the ability to look out using a vision sensor, look at the scene in front of the robot with all of the SKUs with all of the material handling equipment, understanding what that is, reconstructing the scene in the digital environment, so you get this idea of digitization that's coming into play, and then once you get to this point with the scene being reconstructed, essentially all of the models within the workspace, they can be adjusted in real time by the vision system as it's doing its thing. And then at the end of the day, the order-fulfillment process just gets very, very capable. And you get really high levels of uptime when you're able to move to a robotics scenario. So, you have these within the warehouse, and then we can also look at supply chain, just how do we feed that particular warehouse? And so, the warehouse itself, in terms of what parts you're ordering, what's going to be the best case scenario for what you're placing on your e-commerce or your online shopping sites, supply chain itself can be optimized with AI.
Ben Ames, Senior News Editor, DC Velocity 10:48
Yeah, I mean, inventory control, order fulfillment, supply chain replenishment, all that, I mean, that's really the core of what a lot of our listeners do every day. So, Brandon, for a company that's just getting into AI, which is probably the majority of the market, how do they justify an investment in this really new kind of technology, and, you know, wrapping up here, do you have any advice for how businesses can can best prepare for that stage?
Brandon Coats, Global Product Manager, Robotics and Vision, MHS 11:18
Sure. If you're just getting into AI, and you want to justify the investment, the first thing you have to realize is that AI is an enabling technology. AI is also based in software. And every time we've seen software in the past become disruptive in nature, inherently, it's going to go back, and it's going to lower the cost of product—services and goods. And so as it does that, you'll just have to realize that what you may have been purchasing up to this point, in terms of your material handling equipment, you may have been getting used to something that's relatively fixed, that's not changing very quickly. With AI-driven platforms, they do change very quickly. So, as long as you are aware of this, then you can keep an eye out for the new technology that's up and coming. And justifying the investment, it's really coming down to what's going to work for your particular business or for your particular business operation. So there's—we see there's new, viable business cases and market opportunities are coming up, they're left and right, on a consequence of AI. So you just have to be aware of this, is first and foremost. And then stepping into the disruptive technology, what you understood before in terms of how business operates, and how it functions, this is going to fundamentally change. So, the way we look at this is understanding the customer value chain. And with disruptive technology, it's often short circuited. And so with that comes a whole new level of value that wasn't there before, that just, there was no answer for the particular problem. And then all of a sudden, there is an answer. And not only there is an answer, but there's a lot of competition that comes with that answer. And all of this, in nature, is forcing prices down. So you shouldn't really have to go back and have a difficult time justifying the investment. You just have to make sure that you have staff on hand, that you've got a platform within your organization that's keeping an eye out for this, and that you're willing to test new technology on a regular basis.
Ben Ames, Senior News Editor, DC Velocity 13:24
Great. Well, like you said, the ramp up for this is incredibly fast, and looks like AI is going to be that, as you said, pervasive and ubiquitous all around us here, just in the couple coming-up years here, so.… . Brandon, we really appreciate your coming on the show and sharing some of this great information with us.
Brandon Coats, Global Product Manager, Robotics and Vision, MHS 13:44
Yeah, thanks again. Happy to have the opportunity here this morning.
Ben Ames, Senior News Editor, DC Velocity 13:48
Great. And once more, we've had Brandon Coats with us today, from the systems integrator MHS. Back to you, Dave.
David Maloney, Editorial Director, DC Velocity 13:56
Thank you, Brandon and Ben. This past week, as we mentioned earlier, the supply chain industry met virtually for ProMatDX 2021. Normally, ProMat is the largest material handling show in North America, but since we could not gather in Chicago this year because of the pandemic, MHI offered the next best thing: a virtual version of this huge trade show. And it did not disappoint. I was there in the virtual space, and so we're Ben and Victoria, and they're here to share some of what they saw and learned this past week. Ben, you looked at how supply chain leaders are trying to create resilience in their operations. What did you find?
Ben Ames, Senior News Editor, DC Velocity 14:33
That's right, David. It's been a busy week for the whole team at the magazine here, checking out the latest trends and sessions, and one of the interesting ones that I saw this week was that two of the biggest trends in the sector are synching up as digitization, which we were just talking about with Brandon, has increased quickly as a reaction to the pandemic. Specifically, nearly half of supply chain leaders in a recent survey are dramatically accelerating their spending on digital technologies in an effort to deal with those COVID disruptions. So, that information is from the eighth Annual Industry Report that's produced by MHI and by Deloitte, a consulting firm. So, in terms of specific investments, the biggest jump in dollars is going to cloud computing; to robotics; and to inventory and network optimization. But digitalization—or digitization; the term varies—includes a much wider range of technologies. So, the report said that over the next five years, we'll also see rapid increases in things like sensors and auto ID; predictive analytics; robotics and automation; internet of Things; wearables; artificial intelligence, which we were just talking about with Brandon; autonomous vehicles; 3D printing; and blockchain—that's a lot of terms that our readers are familiar with. And it's an impressive list for sure, but in a panel discussion about those survey results, we heard from speaker Randy Bradley, who's a professor at the University of Tennessee, and he said, it's basically the least that supply chain companies could do to start dipping their toe a lot deeper into those waters. Bradley asked, you know, what good it was for a business to recover from the pandemic if they don't come out of it better prepared to handle future disruptions. So, in a quote, he said "the traditional mindset is usually 'Let's rally the troops and weather the storm.' But the challenge is, you're not going to be better prepared for the next storm than you were for this storm.
David Maloney, Editorial Director, DC Velocity 16:35
Yeah, that's a very interesting point from the MHI report, that supply chain businesses can't really just rest on their laurels. Victoria, you covered the MHI Innovation Awards. What new technologies did they recognize?
Victoria Kickham, Senior Editor, DC Velocity 16:49
Sure, Dave, yeah, I did. So, MHI presented its awards on the opening day of the show this past Monday, and a little bit about them: They're awards that recognize companies for developing innovative products and solutions to benefit the material handling, logistics, and supply chain industry. So, they select winners in three categories: Best New Innovation, Best IT Innovation, and Best Innovation of an Existing Product. And this year's three winners were chosen from a field of 93 entries by an independent panel of judges. Best New Innovation went to ThruWave for its 3D millimeter wave imaging technology, and that provides automated inspection capabilities for use in manufacturing and supply chain. ThruWave's millimeter wave sensors can essentially see through opaque packaging—you know, corrugated boxes, plastic totes—to automatically dimension items, count items, detect anomalies, all without opening sealed packaging. What the company leaders had said, they explained in answer to a question, is that it works like an X-ray, but it's human-safe without any shielding, it has unlimited sensor life, and it requires minimal maintenance. Best IT Innovation went to 4Front Engineered Solutions for its 4Sight Connect Digital Gate dock scheduling and gate management software, and that's a product that essentially provides end-to-end shipment visibility and eliminates bottlenecks that you often see at the loading dock—you know, when you have early or late arrivals, for example, due to poor scheduling or lack of communication. So, this really allows companies to sort of proactively schedule their docks for loading and unloading; allows them to send drivers to the right location, you know, based on availability and their arrival; and it really, among the many things that does, the leaders sad, is, it minimizes a lot of the wasted time drivers often spend trying to figure out where they need to go once they get somewhere to deliver something, so... . And the last category, Best innovation of an Existing Product went to right-sized packaging company Packsize for its X7 Automated In-Line Packaging and Fulfillment Solution. And that's something that allows companies to create and pack right-sized boxes quickly, reliably, and on demand. It's a product that's been around for a while, but the new features include customized document insertion, in line scales for capturing weight, and it's also in a smaller footprint. It also features an easy-open tear strip, as they call it, and that makes it not only easier to open the box, but for customers, it allows them to recycle it or use it for returns. So those were the three innovation winners this year.
David Maloney, Editorial Director, DC Velocity 19:23
Yeah, an impressive lineup of really great innovations that we're seeing in technologies. And, Ben a part of that, too, is automation. We're seeing a lot of buzz about automation. What else did you see at the show?
Ben Ames, Senior News Editor, DC Velocity 19:37
Yeah, for sure. You know, automation can do great things—we were talking with Brandon about it before. Victoria was giving some super examples of really fast-increasing capabilities. But in another session that I went to, the speaker, Bastian Himmeroeder, who's a partner with Miebach Consulting, pointed out that, really, each specific company needs to ask whether increasing automation is the right thing for its operations, because it might be possible to have too much technology for their business model. For example, he said that some companies might get better productivity boosts, and the all-important return on investment, actually, from lower automation or semi-automated systems instead of highly automated DCs. So the case for more automation, he said, could include companies that are very busy—they have a two- or three-shift operation; that they have sort of low jumps in peak demand; where there are high labor costs; high inventory complexity; limited land footprint—that all those things can make automation really pay off quickly. But, on the other hand, if you look at the flip side of that coin, a warehouse might look to manual or semi-automated systems if they have huge peaks in demand; if they have easy access to cheap labor; if they have lots of available land; or if they have really strict return-on-investment requirements, where they don't have a lot of time to make that investment pay off. So, it's really an interesting way to look at making automation work, sort of in the real world.
David Maloney, Editorial Director, DC Velocity 21:18
Yeah, and as we've always said, one size does not fit all in this market, that it really comes down to the individual case as to what is the right application for it. Thanks, Ben.
Ben Ames, Senior News Editor, DC Velocity 21:27
Exactly. Yep, exactly so.
David Maloney, Editorial Director, DC Velocity 21:29
And Victoria, you covered a session that talked about how quickly some companies are incorporating automation to meet their critical distribution needs. What did you find on this one?
Victoria Kickham, Senior Editor, DC Velocity 21:38
Yes, that's right, and this is very much in line with what Ben was just talking about, you know, sort of a good example of a company finding the right automation solution for their requirements. So, this was a case study presented in a session. It was about footwear company Crocs, and how it launched, as you say, an automation project really quickly. It was at one of its warehouses, it was up and running in less than 90 days. As our listeners probably know, Crocs makes those foam clogs that are really popular, and the company faced exploding e-commerce demand in 2020, as many firms did at the height of the pandemic, when online shopping really ramped up. And, as company leaders described it, they needed to automate their largely manual picking process in order to handle the growing volume, and also to prepare for added spikes during peak holiday shipping season. They were realizing this in the spring, early summer, and at the time, you know, building a complex, infrastructure-heavy system would take years, so they really wanted something they could implement quickly, but that would also address long-term needs. Ultimately, they ended up using a system from 6 River Systems, which offers autonomous mobile robots, or AMRs, and they applied it to their picking process. These are collaborative robots that move throughout the facility there, either with or without workers—that's how their system works—and they don't require any new infrastructure or the need to change your warehouse layout. The bottom line with this project from inception to go-live, as I said, it took a little under 90 days, but the on-site implementation, they say, only took them six weeks at that time. And that included piloting, testing, and launching the system. So, they were up and running and ready to handle last year's record peak holiday shipping season by October. And just to give you an idea of the results, the project, it took place at the company's Dayton, Ohio, fulfillment center, and they said they saw pick rates double after implementing the AMR system, and they also saw huge improvement in new-hire onboarding, which of course is an issue at you know, peak season when you're hiring additional or temporary help. They said training time went from an average of one week down to one day. So, it was up and running quickly, and they saw results really quickly, too.
David Maloney, Editorial Director, DC Velocity 23:50
That is impressive, fast response for changing needs.
Victoria Kickham, Senior Editor, DC Velocity 23:53
Yes.
David Maloney, Editorial Director, DC Velocity 23:54
We encourage listeners to go to DCVelocity.com for more on these and other supply chain stories, and check out the podcast notes section for some direct links on the topics that we discussed today. Thanks, Ben and Victoria, for sharing highlights from ProMatDX.
Ben Ames, Senior News Editor, DC Velocity 24:08
Thanks, Dave. It's been an interesting week.
Victoria Kickham, Senior Editor, DC Velocity 24:11
It has, for sure. Thank you.
David Maloney, Editorial Director, DC Velocity 24:13
And again, our thanks to Brandon Coats of MHS for being with us today. We encourage your comments on this topic and our other stories. You can email us at podcast@dcvelocity.com. We also encourage you to subscribe to Logistics Matters at your favorite podcast platform, and to give us a rating. We appreciate your feedback, and it really does help people to find us. The new episodes of Logistics Matters are uploaded each Friday. And, speaking of podcasts, Logistics Matters is sponsored by Honeywell Intelligrated. Be sure to check out the Honeywell Intelligrated On the Move podcast on Apple, Spotify, or Google. All episodes of their podcast series are also posted at sps.honeywell.com/onthemovepodcasts. You can also find Honeywell Intelligrated on LinkedIn and Twitter using the hashtag @Intelligrated. And we'll be back again next week with another edition of Logistics Matters, when we will look at trends in making last-mile deliveries, so be sure to join us. Until then, please stay safe, and have a great week.
