Keith Basil, General Manager of Edge at SUSE, joins Ryan Chacon on the IoT For All Podcast to discuss deploying IoT at the edge. They talk about the importance of edge in IoT, what outcomes companies can expect from edge solutions, mission-critical edge solutions, the challenges in deploying IoT such as scale, why solutions are not open source, how security is handled in open source, and the future of edge computing.

About Keith Basil

Keith Basil brings over 21 years of hands-on experience in cloud and related industries. As the lead growth catalyst, Basil leads product management, marketing, engineering, and communications alignment for SUSE’s Edge business unit. Working with the SUSE global customer base, he is also driving development of cloud-native edge solutions that encompass cluster management, heterogeneous architectures, and zero-trust security approaches at scale. He was previously the Vice President of Cloud Native Infrastructure where he drove strategy and management of SUSE Rancher cloud-native products.

Interested in connecting with Keith? Reach out on LinkedIn!

About SUSE

SUSE is a global leader in innovative, reliable, and enterprise-grade open source solutions. Built on a tech community of over 30,000 developers around the globe, SUSE powers the mission-critical workloads of over 60% of the Fortune 500 and counts everyone from SAP, Ubisoft, and Microsoft to HSBC, Walmart, and Airbus as clients. SUSE specializes in business-critical Linux, enterprise container management, and edge solutions, collaborating with partners and communities to empower its customers to innovate everywhere – from the data center, to the cloud, to the edge and beyond.

Key Questions and Topics from this Episode:

(00:37) Introduction to Keith Basil and SUSE

(00:54) What is the importance of edge in IoT?

(02:48) Outcomes companies can expect from edge

(03:49) Mission-critical edge

(05:32) IoT in mission-critical solutions

(07:22) Challenges of IoT solutions

(08:56) Scale as a main challenge

(09:31) Why are solutions not open source?

(10:33) Deploying IoT and adaptability

(12:21) How is security handled in open source?

(14:35) Future of edge computing and adoption

(15:42) When does SUSE get involved?

(17:26) Learn more and follow up

Transcript:

- [Ryan] Are you looking to optimize your IoT data strategy and harness the full value of your data? Partner with DAIN Studios, the experts in data driven solutions. They specialize in building robust IoT data strategies that fully leverage the potential of your data. Their team can assist you in systematically identifying, prioritizing, and implementing data opportunities to drive your business forward.

By establishing key enablers such as data governance and architecture, they ensure the successful implementation of your IoT ambitions. Learn more and visit them at dainstudios.com.

Welcome Basil to the IoT For All Podcast. Thanks for being here this week. 

- [Keith] Hey Ryan. Thanks for having me. Happy to be here. 

- [Ryan] Absolutely. It's great to have you. Let's start this off by having you give a quick introduction about yourself and the company to our audience, if you wouldn't mind. 

- [Keith] Sure. Keith Basil, I am the General Manager of the Edge business unit inside SUSE. SUSE is an open source company that provides Linux and Kubernetes management software as well as edge solutions today. 

- [Ryan] So first thing I wanted to ask you is regarding just edge computing in general. So let's talk about, we talked about it on the podcast before, but from your perspective, how important is edge in IoT, what role is it playing and why is it, why does it matter so much? 

- [Keith] Yeah, it's a great question. I think it's very important to IoT but let me give some context on why I think that really is, right? So it's, we're in an interesting time right now because I think we're at a focal point or a nexus of three things that are supporting the, let's call it the transformation of industrial IoT.

One is that at the edge, we're seeing massive scale. The deployment footprints are just very large, right? So you could have 2,000 locations, 10,000 locations, a thousand trucks on the move. The management at scale challenge is one thing that we are contending with, right? Second, it's been largely known that a cloud native approach to modern application development is the absolute way to go. So you've got cloud native approaches bleeding in to this nexus point, if you will. And so there's a lot of reuse from the approaches and the learnings that we have from a cloud native world. And then lastly to bring the three things together is specific to your question around the industrial IoT space.

So we're seeing purpose built protocols, protocol diversity, use case diversity around particular sub industries in the industrial space. And those things, right, that the brownfields of the world, if you will, need to be brought into a modern kind of cloud native world as we see it but at scale. So the main challenge is that for us, we're building the open source industrial IoT edge infrastructure building blocks to make all of that work from a foundational level so that people and partners and customers can bring their own value on top of that and have a good outcome. 

- [Ryan] Gotcha. And what are the outcomes that these industrial organizations will see by basically doubling down and focusing on building that edge infrastructure correctly, using these tools, and just being able to allow them to, it seems like do more, scale better, and things like that but what are those outcomes that companies are seeing when they're able to bring edge computing in the industrial space?

- [Keith] There's a few. The building blocks I just described earlier, that needs to melt away and not be in the way of the business value that's being pushed to the edge, number one. This is something we do. We just make everything, let's say, below Kubernetes just go away because it's really overhead for a particular team or customer is trying to push things to the edge. Second, once that's taken care of, your customers are seeing tremendous increases in operational efficiency. And for some of our retail customers, there's a better customer experience, right? And there's all kinds of use cases, sub use cases that come from those large buckets. 

- [Ryan] One of the things that's been brought up in the past is talking about edge computing depending on the application but also there's this like mission critical applications and mission critical edge, as I've heard it referred to before, what does that exactly mean? How is that different when we're just talking about general edge computing? Is it just more because it's tied to mission critical elements of the application or what, when that's being used, that terminology's being used, what are we referring to? 

- [Keith] Yeah, it varies depending on the customer and the industry that we're approaching, right? So we largely pivoted to provide what we call mission critical edge for the healthcare industry. In that space, obviously healthcare is super critical. We have to work with our customers and partners to build something that's extra resilient, even beyond standard cloud native practices. And so we lean on past performance and learnings, particularly from the DoD space when it comes to applying security to those platforms in terms of compliance. So we've already got a library full of things that we can do to tweak the software to make it more secure from the operating system to Kubernetes to the application space as well.

So the second thing that we do is we do a deep engagement with customers to suss out the mission critical requirements, right? And from that, one of the approaches that we take, depending on the customer requirement is that we may build a what we call a validated design where we look at the entire stack from the hardware to the application, and we build CI/CD systems that are specific to that customer's validated design where we just run that through and do extreme testing to ensure that the resiliency meets the mission critical requirements.

- [Ryan] And what other elements of an IoT deployment in, let's say, the health care space really play a role in enabling those mission critical solutions to be successful and to work in the way that that scoping that you work with them to do is intended for because obviously the edge is only one piece of it, right? But I'm, all the other play, pieces that play a role, what are you also looking at or what should people listening to this who are looking to deploy mission critical applications be thinking about or considering when they're trying to bring these technologies into their business to benefit those applications?

- [Keith] Yeah, there's quite a few technical things that you have to be aware of, but I don't want to turn this into a technical call. So let's level it up a bit, and I'll just say one thing. The notion of a validated design is critical, right? So you know exactly all the components, software, bill of materials, the supply chain, verification, for example. The validated design that you settle on is table stakes. But the thing that people need to be aware of is that design will have its own life cycle. And so the fundamental question to ask is how do we upgrade that? How do we upgrade a particular component within the validated design? What does our testing framework look like to continue that? And then once that is upgraded, you have essentially a new stack of software. How do we get that into the field? What's the lifecycle management strategy to upgrade let's say a x-ray machine at a hospital in Australia, right? That's a critical thing that needs to be resolved. First baseline, validated design. Be comfortable with that. Have software automation, again pulling from the cloud native philosophies, to work that validated design and keep the life cycle in the software in line. And then what's your deployment model from there? So those are the two big things that we work with our customers in terms of mission criticality. 

- [Ryan] Fantastic. And just generally speaking from your perspective, what are the, are there any kind of challenges that you've seen really take shape recently in the space as far as constant challenges you have to overcome working with customers that might be worth shedding light on here so people can be thinking about those types of things and understanding that there are challenges out there when it comes to deploying these types of solutions. And here's how we should be thinking about it. 

- [Keith] Most of the challenges we've taken on ourselves to alleviate them from the customers. And so if I could put myself in the chair of the customer, I would suggest that find a strong partner or systems integrator or global systems integrator that has the solution that fits your business needs. And I say it that way because it's not so much a technology problem as much as we want to remove the pain or go back to the original value pieces we talked about earlier. We want to increase operational efficiency. We want to have awareness of the supply chain outages, environmental monitoring. So whatever that business case is, focus on that and then find an integrator or partner that has that solution pretty much off the shelf that can solve it for you. Now, I will say that in your selection criteria, I'm going to give myself a little plug here, give us a little plug, make sure that whatever the stack that they're building on is as open source as possible in the sense that you don't want to be trapped into a proprietary system, and you want a system that plays nice with the other industrial systems that you may have on your, in your, on premises. 

- [Ryan] Yeah, I imagine that plays a big role in the ability for a solution, not only to be brought in, but also to scale. 

- [Keith] Correct. And scale, as I mentioned at the top of the call, is one of the main challenges because the great thing about a cloud native approach is that you can pretty much take a declarative approach to defining your deployment, and that declarative approach becomes the source of truth for all of your deployments in that footprint. It allows us to use automation and fleet management to make sure that everything is identical. And that's largely what we see in the open source world is the tools that give us that capability. 

- [Ryan] If I'm talking to integrators, looking at solutions that are not as open source as others, what's usually the reason that something is not more open source than something else? 

- [Keith] It comes from, this is my view, but I think it comes from industry specialization, right? So you have folks who are, when you pick on building automation, there's nothing wrong with building automation, obviously, but let's say there's somebody who's been in the building automation industry for 20 years, and they set off to provide a solution for other folks in that particular space. That perspective may not originate from a cloud native world or from a broader kind of scale world. And so you find that they are building a point solution to solve for that particular use case. Whereas the view that we're coming from is we're starting at cloud infrastructure, edge, and K3S of the cloud and things like that where we look at the building automation protocol and just put our arms around it and bring it in and then everybody shares it, right? So it's a difference of perspective and thinking in terms of solving the same problem. 

- [Ryan] Yeah, I've also noticed that with a lot of IoT deployments when you get in the door of a company to deploy, let's say the first IoT deployment, whether it's in the industrial space, the smart building space, you name it, the nice thing about a well designed IoT solution, picking technologies that play well with others, that play well with your legacy systems, is the ability to add on to existing or not existing, add on to that solution and application to do other things, to solve other problems, to get other data that you might not even had or really realized you had access to. So once that thing is deployed, that solution is deployed, there are many things you can do on top of that. And we've talked to companies, I know this is a big thing in the LoRa space is that once you deploy a LoRa solution, you can use that network to solve other problems in your business once you've installed it, and I imagine that's another piece of value for using things that are more open source. 

- [Keith] One hundred percent correct. And so let me just make this phrase that we say in the open source community. Nobody is smarter than all of us, right? And so we've taken on that phrase. And we as an open source company, we have joined an upstream community called Akri, A K R I. And Akri is an open source framework that allows protocol discovery and capabilities to be plugged into Akri as a framework. So today, I think Akri supports OPC UA, MQTT, and ONVIF for camera discovery. And anybody in the world can add what's called a discovery handler to Akri to give it extended capabilities for that specific device.

And we're working with Akri so that it's the basic framework that we're offering. And then we're going to bring in the protocols needed to address the specific like sub industries within the industrial IoT space.

- [Ryan] And when it comes to open source, how are security elements handled? Because in the IoT space, I think just generally when we talk about large scale deployments with thousands to tens of thousands to hundreds of thousands of devices, there's all these different endpoints that have the ability to be interacted with, hacked, just a lot of more, there's a lot more to think about on the security level than just maybe with other solutions that a company has deployed. How does an open source solution or how does open source technology support that rise in threats when a IoT solution is deployed for a company. 

- [Keith] Yeah. There's a joke in our space that says the S in IoT stands for security, right? It's largely not there. So one of the main themes for us is to augment that security. A year and a half ago, some time ago, we purchased a company called NeuVector and what was great about NeuVector and still is today is that it is the leading cloud native security suite, if you will, that is focused on a kind of a zero trust based approach to security where we know what's supposed to be running, we know the protocol space, and we can lock down the processes and everything that are running at that local cluster at that remote location. And so, but it doesn't fully address the edge, let's call it the leaf device or the industrial IoT device security because some of the legacy protocols, to be quite frank, are very insecure. And so we're gonna have to evolve our ability to reach out, discover those things, and make sure that the security posture management of those devices is brought into this world as well. Now, I do know of solutions that are focused on that, and they are absolutely needed in this space. But we want to, as I said earlier, augment what's out there today, but it is something that everybody should be concerned about.

- [Ryan] One of the last questions I want to ask you before we wrap up here is we've talked about, like I mentioned before, we've talked about edge before. Edge definitely seems to be a driving force in the ability for solutions to scale, to handle a lot of the computation and the analysis further away from having to send all the way up to the cloud, and we're seeing a lot of companies have success in deployments that incorporate edge computing. Where do you see this going? Where is the edge computing landscape going over the next 6 to 12 months and beyond? Just what should we be thinking is going to change or be improved upon in order for people that may be on the fence in understanding exactly how this is going to benefit them will start to really get on board?

- [Keith] That's a hard question to ask only because of the diversity and creativity of folks in this space. But, I will say this, our mission is to turn let's say physical sensor data, like measurements from physical devices, into digital data that can be acted on at a different layer in the stack.

So we see a lot of use cases around AI modeling, operational efficiency, even some use cases where things are self tuning or auto tuning for better efficiency based on that data stream. So there are a lot of use cases to think about, and we want to position ourselves to be those fundamental building blocks and basically say yes to all of the opportunities that are above us in the stack and work with people and partners to make that a reality.

- [Ryan] And when companies work with an organization like yours, at what point are you usually coming into the process? Are you coming in pretty early stage? Are you coming in after certain elements are already put together and built? Or what's the typical kind of business relationship like with customers?

- [Keith] Our customers today are largely the integrators and partners, and we are, the earlier, the better, right? Because we're actually, going back to that validated design idea, we're developing reference stacks for them to bring their value add on and then go to the customer. I do want to give a plug for a really cool organization that we're involved with in Germany called Industry Fusion Foundation. So this is all about Industry 4.0, the fourth industrial revolution. The Industry Fusion Foundation folks actually have a real factory with cutting machine sensors, and our software is in that stack, and we're defining what that stack should be from an open source perspective. And that's, we've been very early with the team over there and our engineering team. So those type of relationships are the ones that we're looking for because this is a long game where if we can get coverage at the base layers, then we can accelerate the value add being delivered to our customers.

- [Ryan] Well, this has been a great conversation. I appreciate you taking the time. We've, like I mentioned before, we talked about edge but talking about it in this perspective and talking about the mission critical side of things, the benefits, the real value that we're starting to see companies be able to deploy is important. So anybody out there listening, as you are working on developing your plans for an IoT solution to understanding how the edge is going to fit in to your solution seems like there's just a mass number of benefits to drive the outcomes in the ROI that they're, that companies are looking for. For our audience wants to learn more about what you all have going on, follow up with questions, anything along those lines, what's the best way they can do that? 

- [Keith] Number one, suse.com. We have great content there for the edge. If anybody's on LinkedIn, feel free to reach out to me. I will personally respond and hook you up with the various folks in the business unit. And let's do some great things together. 

- [Ryan] Basil, thank you again for taking the time. Really appreciate it. And I'm excited to get this out to our audience. 

- [Keith] Okay. Thank you, sir. Happy to do it.