 From around the globe, it's theCUBE with digital coverage of IBM Think 2021 brought to you by IBM. Welcome back to theCUBE's coverage of IBM Think 2021. We're going to talk about the edge, like what is the edge? How it's going to evolve. And we're going to take a look in an autonomous vessel use case, which is quite interesting with me is Rob High, who's an IBM Fellow VP and CTO, IBM Edge Computing. Rob, welcome, it's great to see you again. Thanks, Dave, appreciate that, good seeing you too. Yeah, so let's start with the basic question here. People's like, oh, what is the edge? Like it's one big thing and it's not, it's many things. But how should we think about the edge and why should enterprises feel like it's necessary to begin to lean in? Well, so let's just start with the use cases. What edge means is the ability to put a camera on a manufacturing floor, perhaps juxtaposed with a robot monitoring the work that the robot is doing, using AI visual recognition to detect whether what that robot is doing is producing high quality parts or not. And to be able to do that in real time, to be able to use that analytics then to quickly remediate any kind of quality issues, helps lower cost. It helps increase your yield and it helps increase the overall efficiency of your production processes. Or if not that, then putting it in something that's perhaps a little bit more familiar to us, the idea of an autonomous vehicle, being able to drive and do driver assistance, do driver safety kinds of features. All of that requires compute and having that compute where people are actually performing these tasks based on the data that they're receiving at the moment that they receive it, being able to process that real time, be able to give them the feedback that allows them to make better decisions, to be able to do that, not only with lower latency but actually with better protection of their data, better protection of their personal information or private information if you're thinking about the business in which they operate, be able to do that even when the network fails, be able to do that without necessarily having to transmit tons and tons of data back to the cloud, especially if you end up not actually using that anywhere. That's what edge computing really means. Yeah, so it sounds like the edge is not, maybe we shouldn't think of it as a place but the most logical place to process the data of depending on latency and other factors. It's a good way to look at it. So it's, yeah, just where we do our work. Yeah, where you do the work, right? That makes a lot of sense. Thank you for that. So, you know, we always, we're talking about the pandemic changing the way we think about things. And I wonder if you can comment on the edge context as we come back from work from home or remote work. You know, think 2022, we hope it's gonna be face to face. Could edge play a part in that? Has the pandemic made you think differently about the opportunities at edge? Yeah, and in fact, what we've seen as the pandemic is actually beginning to accelerate digital transformation. If you think about it, you know, any store that wanted to survive this pandemic could only do so by basically introducing a digital presence, you know, the ability to buy online. And even if you're picking up at the store, picking up the curbside, you know, you can't go into a restaurant without getting that QR code that gives you, you know, your digital menu. Trying to get workers back into both the factories as well as the warehouses and offices and to do so safely, be able to ensure that they're wearing their face masks and socially distancing properly. All of these things I think have driven digital transformation. And if you think about the task of, you know, buying online and picking up the store, well, stores better have a pretty good idea of where their inventory is. They need to know exactly where that product is so they can quickly pick it and get it available to the client before they arrive at the store. And so that's edge computing. We need edge computing to be able to automate the processes of inventory tracking down to individual items and where they're located throughout the store to be able to do the recognition for whether people are or not maintaining their social distancing or wearing the PPE to be able to ensure that our processes are as automated as possible to limit the amount of human interaction that's required in order to perform these processes. All of that I think has accelerated both digital transformation as well as particularly the use of edge computing in all of our businesses. I think about, you know, the forced march to digital in 2020. And if you weren't a digital business, you were out of business. But my big takeaway from what you just said is the digital transformation is just starting. And now people really have some time to think about that digital strategy. And as we think about doing things, you know, more safely, maybe with less human intervention. We love autonomous vehicles examples just because there's technically they're challenging. But I wonder if you could tell us the story of the Mayflower autonomous ship. It's an upcoming journey. It's going to be cruel across the Atlantic, unbelievable, collecting data. You know, talk about how edge relates, you know, to that story. What can you tell us? Well, first of all, let's just simply talk about the task of navigating a ship from one port on one side of the world to another port across the ocean, across the Atlantic. You know, the ocean is a dangerous place. Yes, it's wide open. It's, you know, lots of water. But the reality is it's full of barriers. Of course, you got land barriers. You've got other ships. You've got marine life. You've got debris that gets dropped in the ocean. And so the task of navigating is actually quite difficult. And again, to the same point that we made earlier, you have to have local compute in order to really make those decisions fast enough with enough acuity, with enough clarity to be able to safely navigate around those kinds of obstacles. So we have to put compute in the ship. So the Mayflower ship is, as I sort of implied, a ship that will be autonomous. There are no human beings involved in operating the ship. It has to be able to, on its own, both recognize these obstacles, recognize another ship, recognize a boat, recognize, you know, that cargo container that happened to have fallen off some other ship and it's floating through the ocean. Recognize, you know, marine life, whales, and other fish and birds that might be in the way. And to be able to do all that entirely without any human invention. So that compute power is really a prime example of an edge computer. It is compute in the business of navigation, making decisions about the things that it sees and making decisions about how best to circumvent those issues. Now, along the way, I should also say, part of what the Mayflower ship is gonna do is not only exercise the task of navigation and prove that these algorithms can efficiently and effectively bring that ship from one side of the world to the other side safely, but along the way, it's going to conduct science. It's going to collect water samples for the chemical makeup of the oceans at various points along the way. It's going to be sampling for microplastics or examining phytoplankton for its health and life in this. It's gonna be the detecting wave motions and the wave energy that might be indicative of how the world is transforming in the presence of global climate change. These science packages that are going to be formed are also being performed autonomously without inhuman intervention. And that actually opens up a very exciting potential future, which is the idea of these autonomous ships navigating the oceans, collecting data that can then be brought back for the scientists to examine so that they, the scientists, are not having to go out and spend weeks and months at a time in these perilous conditions, these potentially lonely conditions, collecting that data, but rather they can remain safely at land, the ship will collect the data and they can analyze that data from their home labs. So this is actually a really exciting project, but one that I think will demonstrate not only the idea of edge computing, but also the advances in navigation and marine science. Yeah, because I mean, the ship has to navigate itself, not only is it bringing back data, but this is a great, great example. I mean, a lot of the work in machine intelligence today is done in the modeling side. This is inference going on in near real time, which we think is where the action is. That's why we love the autonomous, because there's a lot of IBM tech involved in here as well. Is there not? I mean, you've got to have software and you've got your edge devices. You've got automation capabilities. I mean, it's not, right? This is like serious technical challenge. Yeah, well, we were approached by the primary team on this project and it didn't take us long to realize the utility that some of our technology would have to advancing their project. And so you're right. I mean, we have things like Operations Decision Management, ODM, which typically is used in the financial services industry, but now it's being applied to the rules of navigation we call the Coloregs. We've got our AI services that do visual recognition, because obviously we've got to be able to detect and identify the things that the ship is seeing along the way and be able to distinguish what those things are. We have our IBM Edge Application Manager, which is being used to manage deployment of these kinds of workloads, and frankly, all of the workloads that are hosted in the ship, getting that managed and deployed onto the ship. And of course, all these things have to be integrated. And so that's just a small sampling of the kinds of technologies, but it's a good example where I think the Edge kind of represents the combination of what we have all been working with in this industry, which is how do we bring technologies together to solve a problem as an integrated solution? You mentioned financial services. So I wonder if we could, you know, beyond shipping, maybe what are you seeing in other industries? Are there any patterns that are developing where clients are saying, hey, we need sort of this capability? What can you tell us? So Edge computing is at its probably greatest demand right now in manufacturing, in industrial 4.0 kinds of environments where most of the industry, the industrial industries and markets have grown up largely dependent upon operations technology, OT. But one of the things that people need in these kinds of environments is the additional benefits that come from AI. We talked about using AI to do visual recognition on manufacturing processes, looking at quality inspection, for example, but there's other aspects of production optimization of worker safety. We talked a little bit about that around predictive maintenance and asset management. These kinds of additional things that are necessary to really run your factory efficiently or your drilling rig or your energy production systems, all these kinds of industrial processes can benefit from the advances that are occurring in analytics and then, of course, having localized compute to do that with to both do those kinds of decisions in real time, but also to offload the amount of transmission that we end up transmitting back to the cloud. So industry 4.0 or manufacturing is one big area. Retail, we talked a little bit about that, but you think about point of sale terminals and the idea of being able to do offers at point of sale to be able to do price checking to help you navigate the stores, digital signage, all the user experiences, spillage and spoilage and loss prevention. These are all kinds of use cases that will benefit retailers. Lot of demand and, of course, again, the need to be able to do that locally within the store. We had to touch a little bit on automotives. The whole automotive industry right now is going through a really fundamental transformation where virtually every automobile now is being imbued with more and more compute capacity and localized processing for doing driver safety and car maintenance and even short of full autonomy, which of course is another topic in its own right. Lots of experiences that can be brought there as well. So lots of opportunity and distribution, manufacturing, retail, banking, virtually every industry that we've looked at has some opportunity for leveraging to benefit the Dutch computing. Yeah, it's hard to get cars right now because the chip's short as, but I wonder real quick, if you could talk about 5G, you hear a lot about 5G, there's tons of hype there. How should we be thinking about 5G? How real is it? What's your take in terms of its impact on the edge? So a couple of thoughts here. One is 5G obviously is accelerating and it has the effect of accelerating edge computing because one of the benefits of 5G of course is lower latency and higher bandwidth and that kind of opens people's minds to the potential to leverage the network connectivity of equipment that otherwise is hard to connect. You think about the factory floor for a moment and all the kinds of equipment you have on the factory floor, if you had to hardwire all that equipment to get access to the compute power on that, that could be a very expensive proposition. You'd like to kind of wirelessly connect that equipment and that's one of the things that 5G brings to the table because some of the spectrum that 5G uses has less potential to interfere with that equipment than you would otherwise. So I think that what we're going to see is 5G will sort of disproportionately benefit, I'll call them industrial or commercial use cases as compared to 4G and LTE which were very much centered on consumer use cases. 5G is accelerating edge computing and in many ways 5G actually depends on edge computing. Doesn't mean that we can't do edge computing without 5G, we can, we can certainly do it for DLT even wire line but I think 5G is going to have a very symbiotic effect on edge computing. Yeah, just like Wi-Fi was enabled or mobile but this is much larger potential. Rob, we got to go. Thanks so much for coming on and sharing your insights. I'd love to have you back. Awesome seeing you. Thanks, Dave. Appreciate it. And thank you for watching everybody. This is Dave Vellante for theCUBE's coverage of IBM Think 2021. 2021, we'll be right back.