Skip to main content

AI • IOT • NETWORK EDGE

Upgrade Your Factory Now!

Kevin Jackson IoT Chat

A conversation with Kevin Jackson @Kevin_Jackson

If you were wondering when to upgrade your factory infrastructure, now is the time. The global pandemic has thrown supply chains and customer demand into chaos, and only the most agile manufacturers will be able to adapt.

Join us as we talk to industry expert Kevin Jackson about the steps you can take to accelerate your digital transformation. We discuss:

  • How to get the most out of 5G, AI, and IoT in smart manufacturing
  • The right (and wrong) approaches to cloud migration
  • Why manufacturers need to expand integration with their ecosystem

Plus, in our next Twitter Chat—co-hosted by Kevin—we’ll trade ideas on the same topics. Join us on Wednesday, September 9 at 10 a.m. PDT, to explore new thinking on the agile factory. 

Apple Podcasts  Spotify  Google Podcasts  

Want to know more? Check out the articles we reference on the podcast:

Transcript

Kevin L. Jackson: You don’t want to waste your money and effort trying to transform or digitize processes that really are not delivering value.

Kenton Williston: That was Kevin Jackson, one of my favorite experts on cloud computing and cybersecurity. I’m your host, Kenton  Williston, the editor-in-chief of insight dot tech. I’ll be talking with Kevin about manufacturing technology in a time of uncertainty.

Before we get into the interview, I’d like to invite you to join Kevin and myself as we continue this conversation in our Twitter chat September 9. Just use #IoTDevChat to participate. Kevin, welcome to the show. What manufacturing issues have you been focused on lately?

Kevin L. Jackson: First, Kenton, thank you very much for having me on the show. You know I’m really into IoT, but with respect to manufacturing, really, all industries. It’s all about digital transformation.

Although the adoption of digital technology is really important for every industry vertical, manufacturing for some reason has really been slow to take up the challenge. I believe this is mostly because of the complexities are driven by global supply chains, the tight schedules and resource constraints. The unexpected, rapid, and extremely severe and widespread challenges presented by the global pandemic has forced them to reconsider everything.

A particular interest of mine is about 5G capabilities, the ready push for Internet of Things, Industry 4.0, machine learning and predictive analytics.

Kenton Williston: Yeah, that all makes sense and I totally agree with you there. Industry has been relatively slow to adopt, individually, some of these technologies, and at a higher level, embrace this idea of digital transformation. I think that’s understandable. There’s a lot of risk aversion here. Generally speaking, factories, their key metric is uptime. You’ve got, often times, really thin margins and you’ve just got to keep things running. It’s difficult, to say the least, to justify a huge change in direction and taking your production lines down to do something totally new.

As you said, I think the current context has given a lot of folks, pause to reconsider that because there have been so many disruptions, both to the supply chains, like you mentioned and even on the demand side where there’ve been some very rapid swings in what consumers were looking for, not even business customers for that matter.

Of course, things have calmed down. You know, there’s no shortages of toilet paper anymore, and that’s a good thing. I think there’s still a lot of uncertainty in the manufacturing sector. There’s still a lot of economic uncertainty. There’s geopolitical instability. That makes me wonder, what in the world, manufacturers can do to survive and maybe even get a little bit ahead in this environment where there’s just so much uncertainty?

Kevin L. Jackson: Well, the supply chains are still in bad shape. I’ve recently heard about some manufacturer, a particular one I won’t name, is stopping the availability of few of their products. Announcing, it won’t be available until 2021. Manufacturers really need to recognize, and in their heart, accept that we are in an era of continuous and rapidly accelerating change. The standard operating processes and long-held norms that they have operated on for years, need to be revamped and digitized.

Companies also need to take a much broader view of their position in the industry value chain. It’s not about an individual company anymore. It’s really about the ecosystem. You and your suppliers as well as your customers, manufacturing executives need to build and nurture value-based ecosystems if they wish to compete in this ever-changing environment.

Kenton Williston: Yeah, I definitely agree with that. As you know, on our program, our insight.tech program, we’ve been covering a lot of these topics lately. Some just surreptitiously. Not surreptitiously, serendipitously. Slightly different idea, but we’ve been covering on our program, sometimes serendipitously and sometimes, post-pandemic, to address these exact questions, the idea of having a more holistic view of your operations.

One example that comes to mind for me is, we had a really interesting article we did with a company called Noodle.ai that was all about using AI technologies to have a more flexible and agile approach to managing your supply chain. Like I said, the old ways of doing things, often you’d have just a spreadsheet-based approach to say, “Okay, we’re going to need X number of pallets of this. We’re going to have to have our maintenance crew out, Y number of times a month to keep all the equipment running,” and this sort of thing. A straightforward, regular process.

Now it’s like, well, when things are less predictable, we’ve got to think a little bit deeper into the ecosystem and have a little bit more advanced planning. Also, on the short-term, be able to really turn on a dime. That’s a lot to ask for. Having that really deep view of your ecosystem and also really quick agility. That’s tough. What are some of the key technologies, manufacturers should be looking at to enable that kind of flexibility that really hasn’t been there before?

Kevin L. Jackson: Yeah, you’re right. It is hard, but from my point of view, the convergence of artificial intelligence and the Internet of Things on the manufacturing and assembly line are really introducing, both opportunities and challenges to these companies. The blending will redefine the future of industrial automation with IoT acting as a digital nervous system and artificial intelligence becoming the brain that makes the overall control decisions.

These intelligent and connected systems are capable of self-correcting, self-healing. This innovation would drive unique processes that can enable these companies to establish competitive edge in their marketplace. This is the only way they can survive, if they expect to thrive.

Kenton Williston: For sure. For sure. A couple of examples again, that come to mind for me there, of what I’ve been seeing in this sector. One, there’s a company that we’ve been talking to, called Innovance. They had a really interesting case study that I was reading just recently, where they were using machine vision technology for QA, QC kind of purposes. Of course, that’s a pretty useful technology in any context. What made it really interesting in this example is, because they were using machine-trained models for this QA, QC kind of work, when the pandemic came along and they needed to repurpose their production lines very quickly.

In this case it was moving to production of N95 masks, which of course is a very important thing to have available to the world. They were able to take a production line that was doing completely different kinds of products, and retrain the inspection equipment that was doing totally unrelated kind of inspection, and was able to retrain it for this new purpose very, very quickly and spin up that line.

Another example that comes to mind. Something that, Kevin, you and I have worked together on, was a piece about a company called Relimetrics, again where, this was an example of electronics manufacturing where these self-training systems are able to adapt all kinds of different configurations on the fly. Again, rather than having to have a whole bunch of manual processes, you can maintain that high level of agility, very quickly adapt, and on a dime, produce a new product that you weren’t doing before. Maybe even just add a new variation of a product.

Kevin L. Jackson: Yeah. Absolutely. These companies need to understand, as I said before, change is constant. They need to always be thinking, “What’s next? What does my customer need? How can I deliver value to the ecosystem?

Kenton Williston: Yeah. Absolutely. Of course, I think there’s a kind of deeper question here which is ... I mean, this all sounds great. Let’s deploy all these really cool, cutting-edge new technologies that can enable us. This incredible level of agility and visibility in everything that we’re doing. Of course, the existing factory infrastructure doesn’t necessarily lend itself well to that.

If you look in a typical manufacturing line, you often have a lot of equipment that’s been there for years. May not be connected to anything. It may not have any particular level of intelligence. It’s just a basic piece of equipment that’s doing the same task, over and over. Maybe like a rotating piece of machinery that, all it does is spin around and that’s it. Simple kinds of equipment and assets.

It seems to me like, before we even get into the conversation of all the amazing things that IoT and AI can do for you, that there’s an underlying question of, how do you lay that basic foundation of a factory infrastructure in place, to be able t o support all of these new technologies? Where should manufacturers start there?

Kevin L. Jackson: Well, first of all, they need to understand that some of the things they’re doing right now and they’ve done for years, have no value. They’re doing it basically, out of muscle memory. They need to reexamine their core business processes and the associated key metrics with those processes. The new IoT systems will need to collect, store, process, and analyze that data so that it can optimize the control of production line devices, robots and autonomous vehicles.

You don’t want to waste your money and effort trying to transform or digitize processes that really are not delivering value. The telemetry that these sensors on your manufacturing lines produce, will be interpreted by artificial intelligence through the use of data visualizations, patterns, and correlations. These resulting AIoT systems will then proactively enhance operational efficiency.

They’ll be able to make accurate operational predictions and enable the humans to make more intelligent business decisions. This is not the way of the future. It’s how companies, manufacturers are operating today. If you’re not doing it right now, it’s time to catch up.

Kenton Williston: Yeah. I totally agree with you there. I’m thinking about, again, the real fundamental baseline of the infrastructure seems like, one of the key places to start is just getting all this equipment connected. Again, here I’ve seen lots and lots of recent examples that I think are pretty interesting and useful. We’ve done a lot of work on the insight.tech program to create some really interesting article showcasing.

Cisco for example, has done a lot of really interesting work in the connected factory space where, getting all these devices connected may not even necessarily mean, running a lot of cabling around. It could all be wireless. I know you’re really into the 5G space, so why don’t you tell me a little bit, what’s the starting point for just getting everything attached in the first place, to the IoT?

Kevin L. Jackson: Well, I’m glad you mentioned wireless, because when people think about wireless, they look at wireless LANs, the existing wireless LANs. They may say, “Okay, I already have that. I don’t need any more.” The problem there is, the current wireless LANs don’t have the bandwidth. They may not be able to transmit the large amount of data that these intelligent sensors on the shop floor will be sending. If they did transmit the data, it would have high latency. Your robots will look like they’re drunk. They won’t be able to operate and do their jobs. This is where the advantages of 5G will really enhance the floor.

When people think about the 5G, once again, they make an assumption it’s just about their smart phone and it’s just about wide area networking. That’s not true. The small cell 5G technology is what’s going to be put on these factory floors. The manufacturers will need to replace their existing wireless networks with this 5G small cell technology. This will enable the high bandwidth within these localized areas. It would also bring the edge computing capabilities right next to those assembly lines.

This is how you can really leverage things like machine vision to enhance the quality and reduce the cost of manufacturing these products. The reduction in cost, enhanced quality will turn into your competitive advantage in the marketplace.

Kenton Williston: Yeah, that makes sense and I’m glad you mentioned the element of edge computing. It seems to me like there’s two critical elements of getting all of your operations instrumented so you can start taking advantage of this data. One that we’ve been talking about is the network. Then the other which you mentioned, was the edge computing. I think there have been so many developments in that area, just in the last couple of years that have really changed what’s possible.

I’m thinking of, in particular, when it comes to computing platforms for AI and machine vision, there’s really been a revolution, both in the hardware capabilities and the software capabilities so that you can really deploy an incredible amount of capabilities at the edge in a low-cost, low-power, highly reliable kind of platform. I’m wondering what you’re seeing in that regard. What’s happening at the edge that might be surprising to manufacturers? They may not be aware of right now.

Kevin L. Jackson: Yeah. The most surprising thing is that your compute and your storage is no longer relegated to the data center. You think of a building with a lot of computers inside of it, and that’s where all the magic’s happening. That’s not true in edge computing. The actual infrastructure itself is all software. The server is all software, is actually referred to as Infrastructure as Software.

The compute capability, it’s liberated from these physical buildings. It will travel across the 5G network and sit right there next to your assembly line so that it reduces the distance and latency required to do work. It’s a distributed cloud that enables this edge computing. You may also have heard of fog computing where all of these sensors are talking to one another, and the data is flowing back and forth. Compute capabilities flying back and forth. One sensor may be using storage from another device. I mean, it is really, well, futuristic, but it’s happening now and it’s happening today.

Think about virtual reality and augmented reality. When the worker needs to go and fix a machine, they want to have the manual right there in front of them, and with images overlaid on to the physical device, and with information as to how to fix it. What the steps are. What you should see. All that is driven by this distributed computing architecture and edge computing.

Kenton Williston: Yeah, totally. It is amazing. I mean, it’s just a totally different world that we’re in right now, and it’s the gap between how things have been done. I was describing earlier, where you may have just incredibly simplistic rotating machines that, all they do is they spin. That’s what they do, to the now, which seems super futuristic of the idea that you’ve got a data center that’s so outdated. Your data center really is everywhere. The intelligence is everywhere. It’s an amazing jump.

Having said that, I think there still is a really important role for the traditional data center. An important role even for public and hybrid clouds to play. I think this is another place where manufacturers have been, again, understandably hesitant to get too far into it because there’s reliability concerns, security concerns, which I think are absolutely legitimate. I’m wondering, what do you see as the role of the cloud in these industrial environments? What’s been happening that can address these concerns around reliability, security and all the rest?

Kevin L. Jackson: Well, the foundation of all these advances is really cloud computing because that’s what delivers the low-cost connectivity, storage and compute resources. If security was a major impediment to cloud adoption, we wouldn’t be seeing it deeply penetrate, literally every industry. That’s really a red herring about the security and availability. Education is the cure to any security or availability concerns. Cloud transition failures are mostly caused by an insistence on pursuing an inappropriate use of cloud. This is trying to use cloud computing as you would use, a traditional data center.

I mean, one of the most prolific strategies for cloud is referred to as lift and shift, where you take the exact same architecture that you have in your data center, and you just lift and shift it into a cloud environment. Although it’s the most broadly adopted strategy, it’s the absolute worst strategy because it doesn’t leverage the capabilities of cloud. Those that have done this as an initial move, saying that, “Well, we’ll get to modifying our applications at a later date,” seem to never get there because they can never get the budget to modify their applications to take advantage of the software-based infrastructure.

It’s important. It is critical to understand the business advantages of cloud and to redesign your business models and your IT to take advantage of those changes. Cloud is not about IT. It’s a model. It’s actually, five different models for consuming information technology. It’s a different model of consuming information technology. Different model for securing the data that’s associated with that IT. It’s a different model for acquiring the IT. The economic law’s completely different. So, all this drives different business models.

Kenton Williston: Yeah, absolutely. I totally agree with you there, and a risk of just becoming buzzword soup. When I look at what all of the major cloud providers are doing, whether it’s Google Cloud Platform, or Microsoft Azure, or AWS, there’s so much that’s available, that is radically different from how you would do things traditionally in the data center. I’m thinking of, again, all the buzzwords like microservices and DevOps and all these good things. They really do present a radically different way of being able to deploy intelligence into the factory.

A lot of these things even have specific features that are incredibly relevant for IoT sorts of applications where you can, for example, develop these microservices in the cloud and then push them down to the edge. It’s just a totally different way of thinking about how you do computation.

Kevin L. Jackson: Right. That’s the mobility aspect. It’s not just the humans that are mobile. All of the services across your IT environment are mobile. They act as components and modules that you can switch in and switch out, based upon the service you’re delivering, the customer you’re serving, the location that you are. Even the laws that apply. A lot of organizations are dealing with the GDPR, The General Data Protection Regulation which is, by the way, enforceable globally. It’s not just about European companies. It’s about data. It’s about data.

Even your manufacturers need to think about the origin of the data. How it’s being used? Where the data is being transformed and moved? If you recently heard about the High Court European ruling that Facebook has violated the privacy of European citizens by bringing their data to The United States and manipulating that data. They now have actually said that privacy shield, which is the agreement between The United States and Europe on protecting data is invalidated.

The manufacturers that are operating across the Atlantic, this affects their everyday operations. You need to really be plugged in to how the management of data can affect your manufacturing line.

Kenton Williston: Yeah. Absolutely. To tie this back up to where we started the conversation, I think these new approaches to manufacturing, i.e., using AIoT, the combination of AI and IoT technologies. Moving towards a more wireless mobile kind of model for connecting and sharing data. Moving to a more cloud-centric Infrastructure as Software model for how you manage your factory, and across your production lines, across your suppliers. I think all of these new ways of doing business are really key to achieving that agility that’s so critical right now, when times are so uncertain.

I think, right along with that there’s a need to adopt these sorts of technologies very quickly right now in order to respond to the environment that we’re in right now. At the same time if I put myself in the shoes of a manufacturer, I’m going to be worried about, “Okay, this all sounds great, to have a way to respond to the current context, but is this really going to work for me over the long-term?” What can manufacturers do to, both quickly transform their operations, but also ensure that they have a future-proof infrastructure?

Kevin L. Jackson: Well, see, you sort of, like I say, hit the nail on the head right there. Things will change constantly. They first need to abandon an expectation of any type of long-term consistency. That’s in the past. The pace of change is accelerating across all industries, and manufacturing isn’t an exception. Continuous change is the norm. Survival will depend on an organization’s ability to build their business ecosystems, innovate quickly, and deal with unexpected industry shocks like COVID.

Kenton Williston: Yeah, that makes sense. I see we’re getting close to the end of our time together. Kevin, any closing thoughts you’d like to share with our audience?

Kevin L. Jackson: Yes. The world is digital and virtual. In the manufacturing sector, that means 3D interactive modeling, virtual reality, augmented reality, training using those advanced technologies, additive manufacturing, and remote interactions. You have to know, that’s what’s needed, and plan on using it.

Kenton Williston: Very exciting. Well, I can’t wait to see what happens next. Certainly, I hope we’ll have a chance to chat again about all these topics, Kevin. For now I’ll say thanks for joining us. Where can our listeners find you online?

Kevin L. Jackson: Well actually, I’m looking forward to the Twitter chat that’s coming up. It’s really easy to find me online. If you’re on LinkedIn, just search for Kevin L. Jackson. On Twitter, it’s Kevin_Jackson. Absolutely, don’t miss my articles on insight.tech.

Kenton Williston: All right, Kevin, thanks for the plugs. Like I said, the Twitter chat is coming up, September 9 at 10:00 Pacific. We’ll be, again, talking about manufacturing infrastructure, and would love to have you join us. Just use the #IoTDevChat to participate.

This has been the IoT Chat podcast. If you enjoyed listening, please support us by subscribing and rating us on your favorite podcast app. We’ll be back next time with more ideas from industry leaders at the forefront of IoT design.

The preceding transcript is provided to ensure accessibility and is intended to accurately capture an informal conversation. The transcript may contain improper uses of trademarked terms and as such should not be used for any other purposes. For more information, please see the Intel® trademark information.

About the Author

Kenton Williston is an Editorial Consultant to insight.tech and previously served as the Editor-in-Chief of the publication as well as the editor of its predecessor publication, the Embedded Innovator magazine. Kenton received his B.S. in Electrical Engineering in 2000 and has been writing about embedded computing and IoT ever since.

Profile Photo of Kenton Williston