Episode Transcript
[00:00:03] Speaker A: Welcome to INA Insights where prominent leaders and influencers shaping the industrial and industrial technology sector discuss topics that are critical for executives, boards and investors. INA Insights is brought to you by AINA AI, a firm focused on working with industrial companies to make them unrivaled segment of ONE leaders. To learn more about INA AI, please visit our website at www.aina.AI.
[00:00:40] Speaker B: Good morning and welcome to the Titanium Economy podcast hosted by INA. Today we are joined by Ross Sabalzek, Senior VP and General Manager of the Industrial and Commercial IoT at Silicon Labs. Ross leads the innovation in IoT connectivity and energy efficient solutions driving advancements in smart metering, building automation, asset tracking while emphasizing sustainability and security. Silicon Labs, which is headquartered in Austin, is a global leader in secure, energy efficient IoT solutions specializing in wireless technologies and IT innovations are powering smart homes, smart cities, industrial automation, retail, healthcare applications worldwide. We are excited to hear Ross's insights on the industry and Silicon Labs future. Welcome, Ross, to the Titanium Economy podcast.
[00:01:26] Speaker C: Thank you. I'm excited to be here.
[00:01:29] Speaker B: Perfect. So, Ross, I'll first start with. If you could start by giving us an overview of the Silicon Labs, including its core mission, products and applications and the primary market you serve.
[00:01:40] Speaker C: Sure. So at Silicon Labs, we're focused on bringing wireless connectivity for the iot.
And if you look at our history, when we started down this journey, well over a decade ago, the IOT wasn't even the term for what we were trying to do. We were trying to make smart wireless connected devices that would allow you to control and monitor various types of devices and different applications. Our journey started really, our whole pedigree has been in wireless. And our belief all along has been there are significant advantages to using wireless for connecting devices versus being in a wired technology.
So over the last decade or so, we've built up a complete wireless portfolio that targets a wide range of different markets. So we target home automation, we target true industrial. Think of a smart metering and sort of electric infrastructure. We have what we call a home and life business. So that's appliances and medical. So we're really targeting all of these large market spaces from a technology point of view. You know, one of the interesting things when you look at wireless is there's not a one size fits all. So you have this wide range of technologies where you can think of very like personal area networks which are more maybe Bluetooth, where you're using your phone to connect all the way up to networks that cover kilometers of range with, you know, thousands or tens of thousands or even millions of devices. So there's not one wireless that fits all. So we have a whole portfolio of wireless technology from Bluetooth to wi Fi to there are some networking standards based on a standard called 15.4. So these Zigbee and you might have heard of Matter and Thread, WI Fi and then even technologies like ysun that are used in Smart Grid. So we've built that whole portfolio because customers are trying to solve unique problems that need different technical kind of solutions to get there. So that's our mission.
[00:03:52] Speaker B: Perfect. No, and that's actually very helpful. And we'll through this we'll actually jump into some of the things and especially the areas where you're driving innovation. But could you, you know, now Silicon Labs, which was started in 19, sort of 96, it's been there for more than 25, 30 years now. Could you talk about some of the key milestones or breakthroughs that have shaped Silicon Lab success over the years?
[00:04:15] Speaker C: Sure, yeah, like. Well, when we started, you know, our mission was to be a disruptor in analog mixed signal. And at the time we had started, the idea of doing very sophisticated and high performance analog in standard, what you would call kind of foundry grade CMOS processes was really unique. And we thought that that was going to be transformational in the industry because the big foundries, like if you think of a Taiwan semiconductor, the amount of money they can invest in their processes is so much higher than you could do if you had your own kind of captive process. We thought there would be a huge disruption on riding that wave, which had predominantly been for digital, for compute and things like that, to ride that for analog. So in our history we've done several heavily analog breakthrough products. Our first products were, if you remember, all the way back to B90 modems. So with dialog modems, the original boards were kind of a pcb, you know, maybe this large, with a sea of discrete components. We got that down to an integrated circuit. And you know, we've sold billions of those devices when we were in that market. But it kind of highlights that our core had always been analog disruption. So we moved from that. We played in the cell phone space. We also did FM tuners. So cell phones for a while had FM tuners in them. We were in the majority of cell phones that had those tuners. We did terrestrial TV tuners and a variety of other type of products. About 12 or 13 years ago we wanted to get off of trying to continually do analog disruption in new markets to find a market we could penetrate and then continually innovate within the same market and the same customers. So we took all of that expertise, we built up over a decade doing very difficult high performance RF and then pivoted it to, you know, target the IoT space. So, you know, our dream was to take low cost microcontrollers, bolt up very high performance wireless with them, and then give customers this really cost effective turnkey solution to allow them to connect their products. And so that's kind of the journey we've been on. And now we're in a mode of the systems are getting larger and more complex, the requirements from our customers are getting more, are expanding. So you know, we continue to grow that MCU footprint to do more and more of the customer system, but always bold it up with high performance wireless to allow them to connect their data to a cell phone or to the cloud or to a gateway. You know, that's really been the journey of how we got to where we are now.
[00:07:20] Speaker B: Yeah, and I think so. And just picking on that one. Right. And especially on the topic of innovation, especially. Right. There is a lot of things that are changing in the ecosystem. The number of devices is expanding exponentially, of course. Right. The concerns around the security of the data is becoming more and more prevalent. The number of standards and the number of manufacturers and the platforms is also expanding, which sort of increases the complexity of these things. And on top of that, you really have to differentiate yourself. You have to drive applications that are creating real value for customers. Can you. And I know Silicon Labs has been like leading in many of the solutions that you are developing right. In that particular space. Can you talk a little bit about that innovation process and how you're driving, you know, the value added solutions for the customers?
[00:08:11] Speaker C: Sure, yeah. I think, you know, over time the requirements that or the expectations is maybe a better word that customers have of us as a vendor providing the solution has grown significantly. So if you go way back in time to our first devices, we were doing wireless technology sub gigahertz, so lower frequency. But if you think of a garage door opener or security sensor in your home or the wireless keyless entry for your car, you know, those were some of those applications. And at that time there was no expectation that we would provide software or stacks or security pen testing on our software. Now the idea was give me a high performance radio and the customer would say, I've got it from here. And you know, as you look over time, customers have gone from their differentiation being I can build a high performance wireless network to Silicon Labs, you give me the high performance wireless Network, I'm going to focus on my value add on top of that.
And so, you know, some of the things that have, that is driven for us in innovation has been really a lot of the wireless networking stacks. So first and foremost, you know, customers expect us to deliver them, not just an ic, but an IC with a network stack that is robust, that is field proven, that is hardened against security or against attacks using security. So those are some of the innovations we've had to bring to the table. And we've often had to make those innovations prior to customers asking for them. So we saw early on with the IoT that as more devices get connected to the cloud, or our CTO likes to say, if you break the air barrier, so if there was this, your MCU was not connected to the cloud, it was cordoned off somewhere, that's a different level of security you have to worry about versus I'm now talking to the cloud and I'm using maybe IP protocol that someone can actively attack me. Security becomes much more important.
So, you know, those stacks, the reliability and the innovations we've done around the networking, then in security, you know, hardening devices against all a whole bunch of different types of attacks, physical attacks, if someone can get their hands on the device, remote attacks and things like that. So we were early on embracing security and we have some of the highest level of certifications for our security. But again, customers were saying that is something that we require and where we innovate. The other thing that we rely on is that deep analog heritage. So we try to think of what is the vertical market that we're going after, what's the customer application and what's the pain point they have not just on the wireless connectivity, but also on integrating other things into their systems. Sensors, low power, you know, battery monitoring, things like that. So that's kind of been the basis, you know, the core of where a lot of our innovation is come from. And finally, we've been very active in the industry alliances and sort of working groups and trade associations. So there's a new home standard called matter, for example. And we were an early adopter in matter. We actually our engineers wrote a lot of the core technology that's in one of the wireless standards underneath called red. And by virtue of being part of those alliances and leading those alliances, you know, that's also allowed us to be one or two clicks ahead of our competition. And the final point I would make is as you think about device networks growing to not tens but hundreds or thousands or millions if any of Those devices have batteries. You know, a huge pain point for customers are changing batteries.
And you have to change batteries and maybe a couple of door window sensors at home. That's painful. Imagine if you're running a hotel and you've got a thousand.
So. Or you're in a retail environment with electronic shelf labels that are battery powered and you have 10,000 of those in your store. Having extremely low power and long battery life is another, I would say, core area where we've tended to innovate.
[00:12:51] Speaker B: Yeah. And I think that actually has been one of the core things even, you know, I've been reading about. Right. And it becomes a little bit of a, you know, a balancing act because either you want to go towards a very low powered device that sort of lasts the life of the device, or if the consumption is very high, then you need to figure out a way to power some of the devices. Right. And I think you're probably trying to balance that out depending on the usage and the life of the devices. On, on, you know what the power consumption metrics should be?
[00:13:23] Speaker C: 100%. Yeah. If you have, you know, if you're, if you're plugged into the wall, it's a different situation than you have a coin cell battery and a door and window sensor. And you, you know, the customer expectation is it lives seven to ten years, very different. One of the things I will say though, even in some of those line powered applications, if you think of an electric meter and you'd say, wow, you have instant, you're connected to the grid, you have infinite power, there can often be either regulatory or kind of customer requirements that would say you can't pull, even in that situation, a lot of power. So for example, like a light switch in your home, you would think it's line powered. Why can't I just take all the power I want? A lot of those switches don't have what we would call a neutral running to them. So the switch kind of connects through the light bulb back to the switch to power. And if you start pulling current through that light bulb with a LED bulb, the capacitors will charge up and the light will flicker on and off. So even though you have power there, the end customer requirement may be you can only draw a tiny amount to prevent that sort of behavior. So generally, yes, but we even in line powered devices see the requirement to push the low power.
[00:14:45] Speaker B: The other aspect of innovation, Ross, that I was very fascinated by that Silicon Labs is driving is, you know, the basic applications like smart metering, where you can don't need someone to go to a meter to collect the data, whether it electricity, water, but also the add on use cases now that you're identifying, you know, on some of those things. Right. Which is a real sort of, you know, value add. Can you talk a little bit about that aspect and how do you drive that innovation at Silicon Labs, especially in your industrial and commercial IoT business?
[00:15:17] Speaker C: Yeah, I think it's been a real interesting journey to go on. You know, I think in my space in particular, when you're talking about kind of a business to business model, those customers need to see an roi. They need to see that this is returning something. I'm more efficient, I'm saving power, I'm reducing my labor costs. Whatever it is, there has to be some payback. And what I've seen over and over again in the industry is I can give you a couple of examples. You mentioned smart metering. So at the very beginning, the idea with smart metering is don't roll a truck. So don't send somebody out to the house to have to read the meter. Very clear value proposition. Also very clear roi, kind of fast roi. So it was, but it was one use case. It was, I'm going to read the meter and what you'll see now is those suppliers, those utilities that have adopted those early, what we call AMR automatic meter reading projects, they've migrated so now it's time to refresh those networks. And when they're talking about the networks now, they're like, yeah, we're going to do billing, billing's a given. But if I have a million electric meters and they're all connected in this wireless network, what else can I do with them? So what you'll see is one killer use case that sort of funds the network and then these add on use cases that will come with it. So in metering there's a couple of cool ones that they're doing now. There's this idea of they call grid edge intelligence. So my meter is just no longer a billing device. I want to monitor the health of the grid. So I want to see if there's a brownout happening in my leg of the grid. And now I can provide that data back to the utility or even things like trying to monitor the current signature from the meter into the home to tell the consumer these are the loads that are turning on and off in your home and how much money you're spending, let's say on your air conditioner versus your dishwasher. Those are real value added services. Those are real New use cases. But they all came once that first killer use case got entrenched. Now it's like, how can I use this network to do other things? And there's a couple other examples. Commercial lighting. So the whole idea with commercial lighting was kind of the cost savings by being able to wirelessly control your lights convenience, also sort of being able to adapt the lights to the condition. That's kind of the core use case. But then the lighting companies started to say, hey, I've got a wireless network that covers your entire building where people and all your critical infrastructure and your people are located. What else can I do with it? So now they're doing things like asset tracking, so you can have a Bluetooth tag that the lighting system will use to track assets and people through your facility. So again, one core initial use case. These other cool ones that fell out of that. One final example I'd give is in electronic shelf labels. So you know, the electronic shelf labels, the whole idea was you don't have to have someone go out and change pricing on the shelf. You can do it all in real time.
Once that network got established. Now there are other services on top of it. If you have someone doing click to cartoon, you have a shopper use Instacart or whatever your favorite app is. Now the app can talk to that system, and when it sees, hey, the next thing to pick is in this aisle, it can start flashing that tag to tell you, hey, I'm over here. Pick me. Now, those are all services and cool new use cases that came about because there was one killer use case. They got the network deployed. And then once you have the network now, people are finding all kind of other interesting things to do with it.
[00:19:17] Speaker B: Now that's very fascinating and actually building on that line of thinking, right?
Especially given the number of sensors that you deploy in, like, any industrial or commercial applications. Right. You're collecting a lot of data, and not just the data from that sensor, but across a much wider network. And given like the emergence of these large language models, how do you sort of integrate that in your product roadmap for the next one, two, three years to like, even build applications that were unthinkable, you know, even a couple of years back.
[00:19:46] Speaker C: Right. Yeah. I think there's the whole idea of, as you pointed out, if you have a million sensors pushing data to the cloud, how do you unlock value with that data? You know, the data is just data. If you're not mining insights, it's not particularly actionable or usable. So a lot of the first Push for, let's say machine learning, AI ML, kind of the newer generation of intelligence is in the data center and that's what gets all the attention. And I mean rightfully so. It's incredibly impressive. Massive data sets, a lot you can mine there. So you know, first and foremost those models aren't particularly valuable if they're not fed with lots of data. So part of our connecting all of these devices is building those data sets. So that's part of what we do. But if you look at, we talked about some battery powered devices or sometimes we'll be on networks that have very low bandwidth. You can't push a lot of data, you can't have a continuous stream going to the cloud if you're trying to get 10 years of battery life out of something. So there's also now a focus for intelligence at the edge.
And you know, the idea is you can do some local analysis, you can do some local discrimination before you would ever push data up to the cloud. And you know, a real common one most people can probably relate to is if you think of your voice assistant, that wake word is typically detected locally because you want that to be really quick. When the person says that you get the wake word, the subsequent phrases are sent to the cloud. So kind of what you do is you, you pre filtered the data so you're not streaming everything to the cloud and using all that bandwidth until you get the keyword. And now you send a very targeted packet to the cloud. You can think about that with other applications like maybe predictive maintenance. If you have a sensor, a lot of predictive maintenance is being retrofitted. So you get a battery powered device that has like a vibration sensor, you put it on the side of your rotating machinery and you start streaming data. Well, when it's battery powered you can't always be communicating. So you can do again, look for things. There's something that looks like it's an anomaly. Now let me start streaming data so you can kind of filter it down to only what's interesting. And more and more I think you're going to see that that kind of mega intelligence we like to focus on in the cloud, moving out into those endnote devices.
What we're doing in that front is really enabling special accelerators on our microcontrollers. So we have these AIML accelerators that can run if you're familiar kind of with the area. There's TensorFlow and there's a TensorFlow light. You can actually run TensorFlow light on our very small battery powered MCUs and customers are doing all kind of interesting applications with that. And examples like if you're a tool vendor, some of the cool things they're doing is detecting like you have an electric screwdriver and you're trying to drive a screw in. The screws all the way in. Now the bit is just rounding out, you know, the screw head, they can detect that and they can stop the drill. And that's, you know, very interesting intelligence, but it's happening on a very small battery powered device at the edge. Another example I would just give you is if you're thinking of building automation systems. Everyone's had the experience of if you have a motion sensor, a PIR sensor and you're sitting in the conference room and everyone's quiet for a little while and the lights go out, you have to wave your arms back on, right? It's not a great experience. But if you could use sound or video to detect that there are people in the room, that would be really interesting. But again, it's probably a battery powered device may have privacy concerns about streaming sound and data or sound and video to the cloud. So we've demonstrated examples on our devices where you can take a video image and you can people count in a room and never have that video leave from the device. So we think this Edge intelligence unlocks some applications but also will be a smarter way to kind of fill the cloud by kind of pre selecting some of the more interesting events and not sending everything.
[00:24:22] Speaker B: But that's actually very cool because this Edge intelligence concept that you're talking about, right, it solves two big concerns that people will have right around the battery. You know, of course if you're sending less data, it conserves the battery life for much longer. But more importantly, security, you're not sending a lot of data and which, you know, more and more people are concerned about. So that's, that's, you know, that and I think that innovation that you are driving will actually be, you know, really helpful and value added. There's, there's one more thing that I, you know, that you guys are working on, right? Especially with PI Fi 6 coming on and Matter, which can sort of integrate devices across multiple different vendors. How are you using that to, you know, to enhance your product portfolio and your, you know, your sort of markets that you reach out to and use cases?
[00:25:10] Speaker C: You know, I think WI Fi is, you know, we've, we've just gone through announcing a new family of WI Fi products for us and with a lot of the themes you've heard Me talk about, you know, we've brought those into our WI fi portfolio. So low power, high security, you know, very robust networking stacks. You know, all those things we brought into WI Fi. You know, WI fi is incredibly interesting. One of there's different reasons why people will choose WI fi in their products. One could be you just have a fat data pipe and you know, WI fi can give you the bandwidth you need. So if you're thinking of video cameras or you know, streaming video, the kind of like a media stick or something like that, you really have to have WI Fi. So some, some are driven by the application just demanding a lot of bandwidth. But the other thing WI fi brings to the table is it's ubiquitous. So, you know, if you are a device manufacturer, you can assume that the likelihood that someone has WI FI in their home is really high. So you don't have to worry about another gateway that you're adding to your system. You can assume kind of the infrastructure is there and that's a pool as well for wifi in some applications. You know, WI fi is not without its challenges. So it can tend to be power hungry. So it's hard on batteries or in battery powered applications, you know, and it wasn't really initially designed to have hundreds or thousands of devices connecting to an access point. So we've really kind of tackled many of those challenges. So, you know, our WI fi devices are meant to run off a battery. So if you're building like a WI fi door lock, you know, we have extremely good battery life that you can unlock with that, no pun intended.
Also we have the security built into it and then on top of it, how to make WI FI play in these ecosystems mentioned matter. And you know, so there's matter over thread and matter over WI fi and we support both. So now the, you know, the vision is regardless of what ecosystem you choose in your house, be it Amazon or Google or even your isp, gives you a gateway, those devices will be able to integrate in, you know, in a simpler way. But I would say, you know, what we brought to the table and continue to bring to the table there is because we have this long history of building these mesh networks. You know, a lot of the players in that space are coming from a kind of a cloud data set, sort of cloud kind of infrastructure or kind of a gateway mindset. And a lot of the things you can do there don't play when you're talking a small battery powered device. So a lot of the low power innovations there, you know, we've driven into the standard to make sure it can meet those customer use cases.
[00:28:12] Speaker B: No, that's very helpful. Switching gears a little bit. Want to talk about, you know, your personal journey at.
You have been a veteran at Silicon Labs, I think, you know, 25 years. He joined 1999. And over those years, you have taken a wide variety of roles, marketing to now the, you know, SPP of this industrial business unit. Can you talk about, you know, your own sort of career path and your roles that led to the current position and how that has shaped sort of your approach to towards leadership and innovation?
[00:28:43] Speaker C: Okay, sure. So, you know, my background is I got a bachelor's in Double E, which I enjoyed. And then I got really interested in embedded computing. So I got a master's in computer systems engineering, kind of doing robotics at the time. And I went to work for a test and measurement company here in Austin, National Instruments. And I was writing at a very low level, embedded code. And I really enjoyed that. And you know, I like the kind of the immediacy of feedback on code. So that was that kind of what drew me to it. But what I found myself missing was, you know, utilizing all that Double E background that I had. And, you know, what I noticed was the computer science people I was working with really didn't understand a lot of the math and the kind of the electrical innards of what we were building as test and measurement equipment. I thought, okay, this is kind of something unique I maybe can bring to the table. So I moved into a role there where I started to oversee product development of the hardware and software. And so that background was really valuable because I could talk and have the respect of both kind of groups because I had been in it. So it allowed me to sort of kind of be a player coach and sort of lead by example. And, you know, what I found was I was more and more liking the analog aspects of it. So, you know, when Silicon Labs started, I was like, wow, here's a really interesting and super talented group of people for me to come and join. We had about 100 people when I joined one product, like $2 million in revenue. And I joined as an applications engineer. And what attracted me to the role was I was becoming more interested in the customer and the business side. But I wasn't ready to give up kind of my engineering chops. So I like applications engineering because you were kind of this interface between the customer and the R and D team. And it allowed me to learn more about our products, to learn more about what the customer applications were, but still stay pretty technical. And I, as I went through that, I was working with the marketing teams very closely and I started to get interesting, interested in the marketing side. So we did an acquisition of a company and, you know, I raised my hand and say, hey, I'd, I'd really like to go and help run the marketing team. And that was the basis of kind of the start of our wireless technologies.
So I went and ran kind of the business and marketing aspects of that team for a couple of years. And again, what has always helped me, we're very technology focused, we're very engineering culture. So, you know, being able to bridge that gap between the business and the customer and the R and D teams has been, I think, a skill that served me well. And when I started to run that marketing side again, I liked it. But I wanted to also still kind of stay closer to the product definition, the engineering side. So I got an opportunity to then lead a product line. So that would be marketing applications, R and D. And in our test team, it was a relatively small product, but it was a great place to kind of cut my, cut my teeth. And, you know, I've kind of continued that journey of now stepping into the IoT business about six years ago and in this role. And for me as kind of a general manager, I think that general part has been really important of, you know, I'm not a domain expert on the most of what we do from an analog perspective at the circuit level, but I know enough and I've been steeped enough in the products and kind of the underlying technology that it's allowed me to kind of ask very good, probing questions to build kind of respect with those teams and then, you know, kind of being able to stitch that all together for the entire product and sort of our entire product life cycle. Because I've done almost all those steps. It's just given me a good perspective. And so that's, that's kind of served me well through my career. And the final thing I'll say is, you know, when I talk to, like, younger engineers, they're like, well, wow, you must have planned that all out. It just, you know, when I explain it like that, it's like, well, you had a master plan. And I was like, no, not at all. I still don't know what I want to be when I grow up. But what I found was, you know, always staying kind of curious, always being looking for other areas to learn about and grow my skills. That's kind of been the master plan. And, you know, I've just been fortunate to have done it at such a great place at Silicon Labs. And, you know, just things played out well for me.
[00:33:33] Speaker B: Perfect. Ross, this was a really good conversation and I thank you for your time. Really appreciate the insight, especially about the Silicon Labs and the innovation that you're driving and about your personal journey as well. So with that, we'd like to, you know, close this podcast. And thanks again.
[00:33:48] Speaker C: Thank you for the time. I really enjoyed it.
[00:33:55] Speaker A: Thanks for listening to INA Insights. Please visit Aina AI for more podcasts, publications and events on developments shaping the industrial and industrial technology sector.
