In this episode of the IoT For All Podcast, Robert Hamblet, CEO of Teal, and Rob Tiffany, Chief Product Officer at Red Bison, join Ryan Chacon to discuss cellular IoT adoption best practices from a buyer’s perspective. Robert talks about eSIM technology and emphasizes that flexibility and preserving options are crucial for utilizing eSIM technology effectively. They also refer to the possibilities with iSIM and touch upon the importance of making informed decisions about choosing the right IoT components. The podcast provides an insightful conversation about eSIM, iSIM, and the future direction of cellular IoT solutions.
About Robert Hamblet
Robert Hamblet is the Founder, CEO, & President of TEAL, a global networking company headquartered in Seattle, WA. Teal is the first US-based eSIM platform to be certified by the GSMA providing a cloud-native, Credentialing-as-a-Service platform that provides intelligent connectivity and networking solutions for IoT device and network operators. Prior to founding Teal, Robert developed some of the industry’s earliest eSIM platforms for several multinational connected car manufacturers.
Interested in connecting with Robert? Reach out on LinkedIn!
About Rob Tiffany
A Top Voice in IoT, 5G, and Digital Twin AI, Rob Tiffany is the Founder and CEO at Digital Insights, an organization providing strategic advisory services on emerging technologies to leaders in industry and the military. Rob has held global leadership roles at Ericsson, Hitachi, and Microsoft. As Vice President and Head of IoT Strategy at Ericsson, he drove 5G connection management with the IoT Accelerator and participated in global M&A activities. As Chief Technology Officer at Hitachi, he received the Presidential “Product of the Year” award for designing the Lumada Industrial IoT platform which landed in Gartner’s “Leaders” Magic Quadrant. Spending most of his career at Microsoft, Rob was Director and Global Technology Lead for the Azure IoT cloud platform. Prior to Microsoft, he co-founded NetPerceptor developing one of the industry’s earliest Enterprise Mobility Management (EMM) platforms for smartphones. As an author and speaker, Rob is a frequently sought-after source globally. He’s been featured in Wired, Forbes, Fierce Wireless, Inc. Magazine, Dataconomy, Thinkers360, Onalytica, Mobile World Live, Techonomy, and SXSW.
Interested in connecting with Rob? Reach out on LinkedIn!
About TEAL
TEAL’s patented, GSMA-certified eSIM technology connects any compatible device to any data network worldwide. With more network operator agreements than any other connectivity provider, TEAL gives businesses everywhere the flexibility and control to remotely switch between networks, ensuring the highest level of reliability and performance for any internet of things (IoT) deployment. TEAL supports applications across many industries including mobility, robotics, drones, industrial IoT, railways, and healthcare.
About Red Bison
Red Bison designs, builds, and manages high-speed networks with an integrated Edge Cloud and Building Intelligence Platform for commercial and large residential real estate. Their services empower building owners to streamline their operations, strengthen security measures, and ultimately increase overall property value.
Key Questions and Topics from this Episode:
(00:18) Introduction to Robert Hamblet and Rob Tiffany
(00:50) Understanding cellular IoT solutions
(02:07) Choosing the right connectivity
(04:27) The role of developers in IoT solutions
(05:05) The impact of network congestion
(09:38) The evolution of cellular connectivity
(15:20) The promise of eSIM and iSIM
(20:00) Scaling cellular IoT solutions
(36:34) The future of cellular IoT
(42:31) Learn more and follow up
Transcript:
– [Ryan] Welcome Robbie and Rob the IoT For All Podcast again. Thanks for being here.
– [Rob] Thanks for having us.
– [Robby] Thanks for having me.
– [Ryan] Great to have both of you. You’ve both been on before kind of individual episodes a couple times. So excited to have both of you together. I think the topics we have planned are very relevant right now.
Before we do that, I wanted to just toss it around and have you give a quick introduction to, about yourselves and the companies you’re with, just for our audience who might not be as familiar. Robbie, you want to kick things off.
– [Robby] My name’s Robbie Hamblet. I am one of the co-founders, and I’m the CEO of TEAL Communications, which is an eSIM technology company based out of Seattle, Washington.
– [Rob] Yeah, Rob Tiffany. I’m a Chief Product Officer at a company called Red Bison, which is in Kirkland, Washington, and it’s in the proptech space, so IoT, edge, and commercial real estate. What a combo.
– [Ryan] Today, we wanted to talk a lot about adoption best practices, particularly relating to cellular IoT solutions. Framing this kind of from a buyer’s perspective to help all those people out there looking to better understand, learn, and adopt solutions. Let’s go ahead and start off, Robbie, maybe we can kick this off with just an overview of cellular IoT solutions, like what they are, what that means when we say cellular IoT solutions versus other solutions and kind of things like that.
– [Robby] Yeah. Cellular has been used mostly in like consumer devices for phones and tablets and to a lesser extent laptops. And for IoT solutions, cellular is popular because of its ability to be an actual wide area network. So a connection means you have an IP address, and it also means that you are able to be mobile.
So other technologies maybe don’t maintain a high level of throughput as you’re physically moving the modem or the device around. So cellular has been used in any kind of outdoor mobile application where there isn’t a dedicated connection to a base station or something like that. That’s where cellular has been most popularly applied.
– [Ryan] When we’re talking about utilizing cellular for a solution, how do you best determine the type of connectivity that is well suited for a solution? Because I know like before we even jumped on, Rob, you were talking about the work you’re doing in buildings and skyscrapers and how cellular is not the best option for that.
And Robbie, you even agreed, so just how do you, how should people be thinking about evaluating the types of connectivity or when cellular is the right option, maybe when it’s not, besides just the fact that being outside versus inside. Are there other factors that are involved there?
– [Rob] Certainly as Robbie mentioned, mobility is a big deal. If you’re moving around, going around the world, that kind of thing, you’ve got to have cellular. I think over the years, I think both of us, we’ve, we get pretty pragmatic about things, and you’re like what’s the use case that drives, instead of just saying, you should always use cellular or you should always use whatever. It’s use case dependent. After years of doing fun science experiments and stuff in IoT, people have gotten pretty darn pragmatic and cost sensitive depending on the scenario. And so like when I talk about doing IoT in a building, if there’s already a network there inside the building, which there probably is, there’s ethernet inside the walls, right, and Wi-Fi and stuff like that, I’m likely just going to hop on and piggyback on that network, right? I need an IP network, just like Robbie was talking about. And so if I’m indoors, and there’s a network, there’s a good chance I’m going to hop on that one if I can. And it has a lot to do with just cost and ease, right?
– [Robby] Yeah. Use what’s there. Best network is the one you can actually see and connect to. But yeah, definitely consider congestion too because there’s been a lot of recent dialogue around Wi-Fi and the system of repeaters and channels that Wi-Fi use is just not really well designed for bulk mass device adoption.
Buildings could get a lot smarter, but they don’t necessarily have the ability to access a radio spectrum that isn’t already congested, I mean, Rob, I don’t know, have you seen, have you done any kind of deployments on things other than Wi-Fi indoors?
– [Rob] Definitely done Bluetooth and Thread and stuff like that in the past. You bring up a good point. I always like to put myself in the shoes of the developer who’s actually building all these solutions. And developers, can be lazy or they want to take the path of least resistance. And so if you can give them an IP address right from their thing, their device, whatever, all the way to the destination, they’re going to love you. As soon as you say, oh, actually here’s some kind of weird protocol, and you’re going to have to use a gateway and a translator.
– [Robby] And there’s no SSH.
– [Rob] Yeah. And people are like, oh, I don’t know, is there an easier one? And there’s a lot of stuff out there, you said LoRaWAN and a bunch of others that you know just lots of them are not IP based, and so it makes it harder for the developer.
I also think it back to, when you think about congestion, we need to put the onus on a lot of developers out there to be efficient because congestion, there’s this infrastructure, like Teal’s building, a lot of other companies are building, and then there’s these developers building solutions that are riding on top of that infrastructure to deliver whatever the solution is.
I remember having flashbacks to, if you go back to early 2000s, there was this big hype around SOAP and XML as a way to send data across the internet in a heterogeneous way, which we didn’t have before. It turns out it was the most bloated, fat way to send data and call methods on other computers ever imagined.
And then you had things like REST come along and JSON, which were smaller, then you had compression. Then you’ve got all these other cool binary formats. And the reason I say that, not just get too geeky, but developers and the choices they use when building these solutions plays a giant role in how all this plays out.
You could have perfectly great connectivity inside your house, but it could seem congested because a bunch of guys built some stuff with the wrong protocols of, that kind of stuff. That’s a thing.
– [Robby] It’s really interesting like when you are so like, and there, there is a also interesting point there of sometimes there are like QOS limits when you’re in a network, and you’re using a network operator. So like cellular will have Verizon, AT&T, T-Mobile, they all publish like guidelines for what a machine device should be using, and they’re less sensitive to a CAT-1 device, but if you’ve ever, back in the day, we used to like wait really eagerly for the latest phone update. And it was like why can’t the carrier just let this phone update happen? It’s they’re going through all these tests. A lot of those tests are like network performance related. They’re trying to make sure that the device is going to behave on the network. It’s not going to open a TCP socket every second. It’s going to, it’s going to wait its turn in line. And there is like a really interesting thing I saw that Netflix was getting behind this week related to network congestion. But when you program your own network, right, it’s my Wi-Fi network, I don’t have some bureaucratic device performance behavior team testing my devices I as the developer in that scenario, although I definitely did not develop these solutions, I have to look at what’s going on in the network, and I have, I almost become the bad guy.
Just because it’s a free lunch doesn’t mean you should eat as much of it as you can necessarily. And things like LoRaWAN and Wi-Fi 2.4 and Thread and some of the other, Bluetooth, that we’ve talked about like, they still need best practices even if there isn’t some network provider that’s like enforcing that best practices. That used to be like that was the biggest annoyance, but now it’s smart the way they did that. If everybody was just going hog wild, those devices would slow, or those networks would slow down a lot.
– [Rob] I mean, you know what the thing that’s eating the most bandwidth in your house of your Wi-Fi is actually your smart TVs now. The TVs are all a bunch of collection of apps that are streaming, and the streaming is just eating everything in sight.
– [Robby] Can I go on record as saying that like I, maybe I’m the first with this conspiracy theory on the internet, but if you have a Fire TV, in the last few updates, they have been like running the programs in the background while the TV’s off, and they’re getting ad revenue from that, and they have no way to validate to the end user, or sorry, to the customer of theirs which is the advertising platform, they have no way to validate that somebody’s actually watching that ad, whether the TV is off or not, right? But by auto-playing ads, they’re getting the views, right? Do, anybody else think about that?
– [Rob] The same people who built those little things that you’d run on your PC during the dot-com era where you’re looking at ads all day, and they were going to pay you money.
– [Robby] I remember getting a couple $5 gift cards in like fifth grade, just signing up for Progressive insurance quotes over and over, and I couldn’t even drive yet.
– [Ryan] Let me get back to something you mentioned earlier, Robbie, you were talking about, we mentioned cellular, and you were talking about, you were alluding to where cellular is now versus where it was before and what it’s enabled. So can you give just an update for our audience of where we are just in the general scheme of things with cellular connectivity in the IoT space, like where we are versus maybe where we’ve come, the landscape, the differences between kind of the different types that are out there that people should be paying attention to.
– [Robby] You’re starting to see networks prepare for the migration, it’s not even a hard migration necessarily, but they’re getting ready to ship their 5G standalone networks. And so much of the last five, even 10 years, I don’t even know, has been based around LTE as a backbone and 5G kind of stapled on top in what’s called 5G non-standalone. And so Cat M, NB-IoT, those are very much 4G standards. And there’s now with the networks getting ready on the infrastructure side for standalone 5G, where they don’t maintain 4G connectivity on certain subscriptions. There is a readiness for 5G RedCap, and the newer evolutions of Cat M and NB-IoT.
– [Rob] What the heck is that RedCap thing all about?
– [Robby] I think driven by like bill of materials. There’s such a bastardization right now in the lower cellular module market where like you’re doing NB-IoT, but you maintain a 2G license and a 2G subscription at the same time because you need to have 2G to do SMS, which is what a lot of protocols are still based around, and that’s what you’re using. RedCap’s just reduce capability. It also means that the networks can deliver less features to a lower cost for the types of devices that don’t need them.
So things like NB-IoT, which has been a disaster in many markets, they came with a lot more features than were required because they were based around a 4G ecosystem that had to have that backwards compatibility.
– [Rob] If you go way, way back, a lot of this IoT on cellular was 2G or GPRS.
– [Robby] And still so many use cases. They literally just need eight bytes to send a GPS header via UDP to a cloud somewhere. And that’s all they care about. That’s all they’re doing. And I think with RedCap, it’s really focusing on devices that are asset trackers or energy meters that are just reporting a voltage or something periodically. You don’t need all the overhead of like a CAT-1 connection or a Cat M connection.
– [Rob] Because if you think about where we been with cellular and IoT, if you go back to just like, when I used to work at Ericsson, and Ericsson worked with AT&T a long, long time ago, when you first started having the very first connected cars, and I don’t mean connected like it’s high tech, it meant, oh, hey, you can have Wi-Fi inside your car through this LTE connection. And so they used to, we used to run all that stuff out of our Plano, Dallas office for Ericsson with these data centers for all these cars in the United States because LTE was the first time that it was really good enough where people could share Wi-Fi inside a car. And it was interesting, but then it’s just over time, so much of it back to pragmatism and price, the cost when it’s a car, I always think about whatever that IoT bill of materials is needs to be less than 1% or whatever of the total cost of the thing that you’re monitoring. And so it’s easier to have a more expensive solution when it’s a connected car, but if it’s a little tiny thing you’re tracking, it’s got to cost pennies or people won’t do it.
– [Robby] Yeah, it’s still not going to cost pennies in cellular, unfortunately, because of how the licenses go to actually put them and include them in a modem.
But yeah, it’s, there’s also a lot of pressure. It’s happening right at the perfect time like with the Quectel ban from the U.S. markets, like that reduce, they’re not outright banned, I don’t want to say anything out, like actually, it’s actually incorrect, but there’s a lot of concern over Chinese vendors, like Quectel, and they’re banned from certain applications, and there’s concern that they’d be banned from further applications. That’s going to create a lot less competition in the market, higher prices for cellular.
It’s been always frustrating to me as a fan of cellular technology, even in consumer devices like laptops, it’s a $250 upgrade or something crazy like that. It should be ubiquitous. It should be just like that’s your outdoor connection. Wi-Fi is your indoor, there’s probably a whole thread we could go down and maybe Ryan has a question for us to prompt it about Multipath TCP.
– [Rob] He’s been waiting to ask about that all day.
– [Robby] But like just a hybrid type solution, how it can use all these different applications because the best network is the one you can see and being pragmatic about do I pay 50 cents to a dollar for a LoRaWAN module or 20 to 50 dollars for a cellular module, and then you have a subscription type on top of that. If they’re all doing the same thing, it’s just how much dev work are you creating for yourself using a more proprietary solution.
– [Ryan] I did want to ask about the benefits of being able to, of what cellular brings to increasing adoption and helping increase adoption solutions. Before we do that, I think it’s important to talk about a couple things that people are hearing a lot about, especially the non-technical audience, when they hear, when they talk about eSIMs and different kinds of I guess like true eSIM versus eSIM versus MNO eSIM, MVNO eSIM, like what, yeah, so when you’re hearing that as a potential buyer, it can seem like a lot and not, to understand the details between them, I think that’s sometimes overwhelming. What are, can you just like high level talk about the different types and then the benefits and limitations of each just so that our audience who is hearing this out there in their conversations around solutions will just understand what is being communicated to them.
– [Robby] Yeah, so I think this being a question for me, like I am the CEO of an eSIM company, and it’s still, I understand how confusing it is for people because eSIM is a concept of a reprogrammable network identity. That’s really what it comes down to is can you change the network identity?
What people think they’re getting and what they actually get is so different sometimes. So because you slap an eSIM logo on something doesn’t mean that it’s going to actually support what you thought eSIM was built to support because the standards still preserve like a good amount of carrier autonomy with the approaches to the market.
So today you get an eSIM either from one of the eSIM tech vendors, and that would be like a Telus or an IDEMIA or a Teal or a GND, one of those kind of names. And then you go out, and you build your different operator connections into that, or you’re getting it from a carrier, and a carrier eSIM is not going to do the same things as an eSIM that is yours, that you purchased and a platform that you’re running. It’s also similar to what we were talking about with who’s the operator. Sometimes people think about who’s the eSIM operator. It’s really like who is the eSIM management platform. And if it’s managed by your MNO, or it’s managed by somebody not you, you don’t get to have a say of what that device is going to do. You don’t have as much flexibility. You might be a little less insecure and a little more confident about the chances of a roaming change in the market affecting how your device works because you might think that carrier or trust that carrier has your back, but the MNO, to the most, to 99% of use cases, the solution that’s being provided by an MNO or an MVNO for eSIM is not the real expectation of the customer, but they just check in the box of oh it’s got an eSIM in it. I can change it. How you’re going to be able to change that is where it becomes a pragmatic issue. Like, how can you actually get Verizon to install an AT&T profile into your card? It’s never ever going to happen, and they’ve even made that clear.
– [Rob] Choice.
– [Robby] So is eSIM just about freedom, is it, should we make a movie about it?
– [Rob] Yeah, absolutely. You should have a painted face and like freedom. Yeah.
– [Robby] Yeah, it’s like open internet. It is, it should be like more of a foundational technology. And some of the new standards do open it up more because today why you have this problem of you get a chip from somebody, and you don’t own it really at the end of the day. You can’t force Verizon to change anything in your Verizon global eSIM. You can’t force an MVNO to change anything in that MVNO eSIM. It’s up to them at the end of the day.
The new standard, SGP.32, going to use an acronym there, but the new GSMA standard for eSIM does allow you to change what platform you’re managing that eSIM with. So it’s going to create a lot more competition. It’s going to leave a lot of uncertainty. The EU really doesn’t like it right now, which is why it’s been delayed a little bit, or not why, but it’s one of the reasons, and it’s going to create a lot more responsibility on the devices to use the technology appropriately. But the new standard I think will make it so that most eSIMs behave the same. It’s just a command to change it to something else instead of being locked into somewhere.
– [Rob] I feel like so much of this is just about education. It seems like the consumer is just not educated enough to know that there’s a difference.
– [Ryan] Yeah, absolutely. Which I guess brings me to the next point is, now that we’ve given an overview of what eSIM is, how is it leading or how is it contributing to IoT adoption across industries and what are some of the issues that you’ve seen with scaling cellular solutions in the IoT space?
– [Robby] I’ll let Rob take the latter half of that and just general cellular scale problems, but with with eSIM, what it’s been mainly trying to solve is like your access ability. Do you have the rights to access a network? And if those rights change, do you have an ability to get around that? Because things happen, regulations create new compliance issues, like Germany, Brazil, we’ve talked about all these examples a lot, Turkey, UAE, China, they all have evolving data regulatory legislation. And if your device is planning to go global, and most IoT solutions think about the internet as a global concept and not something that’s completely local to one area, like on-prem or even a country, eSIM can open up a lot of borders literally and figuratively as far as where you’re going with that device and where you’re able to stay in compliance with local internet regulations.
– [Ryan] So it’s basically enabling more use cases and solutions to exist because of that flexibility.
– [Robby] Yeah. The other thing I’ll add is like the new satellite technology is very much eSIM capable. So really 17 NTN networks after decades and decades of not even duplex type satellite networks, but the ones that essentially relied on a trust based authentication mechanism of I’m not going to receive any data from something that wouldn’t benefit from sending me that data. That’s literally how like satellites were like I’m just going to authenticate everything that’s coming to me because why send me something I can’t do anything with. Now they’re using cellular authentication. I think at some point we might even see a cellular, sorry, not cellular, eSIM based Wi-Fi authentication, but there’s just not great tooling available. I think iSIM will change the story for that, and the SIM card will exist as a separate enclave of memory. Much like your TPM chip on your laptop exists to validate hardware, the iSIM, eSIM environment would exist to validate your network identities separate from the actual operating system. And yeah, I think the second half of that question from Ryan, Rob, was around just in general historical issues with scaling. I laid out like the network access problem, but I’m sure you have a lot more like application level experience too.
– [Rob] Before I jump into that though, you said something interesting just about how the SIM may be on a laptop or whatever starting to be the identifier as opposed to a TPM chip. It, a blast from the past kids out here. So back when I was at Microsoft forever, if you remember, we had Windows XP and then we were working on this thing called Longhorn. Ultimately, it was Vista or whatever, but there was the time, whatever it was, whatever it was, it was that short-lived operating system.
But I was on a team, and we were actually prototyping and everything the idea of changing all the identity and everything for how Windows worked to SIM cards actually. Back in the early 2000s, probably 2004 or 5, something like that is when we were working on it. It just made sense. It’s, hey, here’s this global thing that everybody uses to identify themselves that’s not tied to some proprietary Microsoft way to do it or a whoever way to do it.
– [Robby] It might sound really silly, but if you think about how like Java cards are still used in banking for chips. It’s all the same Arm. It’s an, it’s a specialized Arm SecurCore, it’s hardened so nobody can physically possess that chip and get access to the actual root security keys. It’s, a lot of development effort has gone into how to make these little Java cards secure. And there really isn’t an alternative. Passports use the same technology. Digital driver’s licenses use the same technology. It’s just a different form factor of it. And, yeah, so eSIM, we talk a lot about like its applications to networks, and they don’t call it a SIM card when it’s in your banking card or if it’s in your passport or whatever, but it is, the technology is, it’s a little Java chip.
– [Rob] I still have my Java ring somewhere laying around here in the office from one of those old JavaOne conferences. But yeah, scaling a solution with cellular. First thing, like Robbie said, is just access, choice. You need to be the, the customer needs to be in charge or the provider and or the customer needs to be in charge. If you need to switch carriers because where your deployment is doesn’t have good connectivity with one of the operators, and you need to make a switch to try something else, you need to be able to do that. Boy, if there was one reason why to use eSIM and have that kind of choice, that’s a big one that you face when you’re actually doing an IoT project, and you’re deploying it, and it’s a cellular one. You can imagine hitting a wall. It’s like, oh, we went with this operator and because we can’t change, we’re stuck and our project’s stuck because where we happen to be, there’s not enough connectivity, and I’d love to try somebody else’s.
Long before we were doing this IoT stuff, obviously we had the mobile revolution, right? And the smartphone revolution, and rugged enterprise smartphones were popping in different, had different modems and stuff for mobile operators, for people like FedEx and UPS when they’re take, out there in their trucks delivering stuff, and they did that specifically to make sure they could always have connectivity. And I think the same thing applies to IoT. And so being able to just digitally go to a dashboard on the web and say, I’m going to switch to this guy and maybe have better luck, that’s a game changer. People used to have to have expensive hardware with supporting multiple carriers and everything all in one piece of, and that was nuts.
– [Robby] My wife’s BMW, it’s a BMW i3, that little electric one, like from 2016, is no longer technically capable of accessing the AT&T network despite the fact that there are LTE bands in the thing because it was shipped with a soldered 3G subscription, and that, there’s no changing that. Even though AT&T has all the reason in the world to change that, there’s no technical way to do that because they never developed a platform to make that happen so. It wasn’t embedded SIM, it just had no smarts behind it.
– [Rob] The other thing with scaling is, I know in the past we’ve talked about 5G. One of the things around using the things like NB-IoT or LTE, one of the promises around 5G was this idea of a million concurrently connected devices within a kilometer radius of a cell tower. Obviously, you’re using small amounts of data to do that. It’s not like big volumes of streaming video. But that’s a game changer. I think back to all the smart city projects over the years that worked or didn’t work or whatever and the thinking that, let’s just say, that they were on track to succeed, like the bureaucracy of City Hall, and they actually had money this time to do the project. They would’ve hit, they would’ve hit bottlenecks really quickly that they didn’t know about because prior technologies couldn’t support enough concurrent streams of data. And so that’s, that was always my favorite thing about 5G, aside from faster cat videos, was just the mass of devices that you can have simultaneously because you can’t do a smart city thing if you don’t have that. But a lot of people didn’t know any better. They just assumed it would work. It also goes back to scalability comes from the developer too. The way you send data, how tiny your data is, all that kind of stuff. Your backend system, how long are you hanging onto a connection before you let go to let the next one. All those things play a role in scalability.
– [Ryan] We’ve kind of addressed the scalability issues, things that maybe were leading to success, how eSIM is playing a role in all this. But if we’re thinking about just generally putting together a strategy, how, like if a company is looking at how can we be successful in the deployment, how can a company basically deliver that complete strategy for scaling a cellular IoT solution and what are the things that somebody listening to this needs to make sure they’re taking into account when they get started on that journey?
– [Robby] You got to keep as many options open as possible. Like we were talking about like if you don’t have an IP address, and you don’t support SSH, like there’s just a huge limit to what you can hire somebody to change in your devices that are now shipped out in the field. You might be able to physically change something by rolling trucks, but you want preserve as much optionality because you want to get it right, but you also know you’ll probably find things that you need to change going forward. And that kind of applies to all hardware. Don’t get stuck down a proprietary path with either a specific module designed for a specific operator. It’s plenty easy now to go out, it used to be buy a different iPhone for each carrier, right? Now, you don’t have to do that. So, if you’re building cellular, make sure you’re building with as many bands as possible if your bill of materials can support a more expensive chip because there are slightly more optimized costs if you can reduce complexity. But don’t get burned by reducing complexity to the point where you can’t be agile to change things when the unforeseeable unforeseens happen because there’s plenty of things like you can get warned about roaming, you can get warned about the number of TCP connections you’re having per minute, per hour. Things that you can foresee being changed by an operator, but there’s plenty of things that you wouldn’t necessarily see like maybe the LTE networks get retired faster than you expected for your favorite carrier. Yeah, optionality is big. And then if you want to adopt new technologies, if you pick the most open stack that you could, that would be a good thing.
So, I like to use the example, and maybe this is going to be one of the questions, Ryan, was like, gaming is a big thing on PCs and it, on hardware and in a lot of ways paid the bill for the AI accelerators that NVIDIA is able to build today, right? The constant pursuit of like faster, better quality video game solutions drives a lot of innovation. And like VR is one way that people are using, I’m really going to VR, Rob. I’m really going there. But like people complain about mice that click or keyboards that type 20 milliseconds of input lag between the mouse and the screen. And I can’t say I’ve ever measured or cared about like the interface devices I use on my trackpad or on my laptop. Like typing into Microsoft Word, I don’t think I really care on how fast the keyboard’s going to register. But like signal latency between like a display and the computer that it’s connected to over HDMI or DisplayPort can easily be another 50 milliseconds on top of that. Real 5G solutions are 10 milliseconds or less. Literally less time to get to the tower wirelessly than you’re plugged in USB mouse is to your computer. Like just to talk about the data there, that’s crazy. The amount of IoT applications that can benefit from a sub 10 millisecond link speed to compute resources that are not on that device, that means you don’t any longer have to build expensive compute into the device that’s going to take battery and increase the bill of materials, the devices become more like thin clients, right? They’re essentially just network access and whatever sensor they’re connected to. They don’t need to run the vision model on the actual device, right? You don’t have to have everything calculated on the hardware itself, you can have it offloaded to a tower. Those things are going to massively reduce the bill of materials, increase the accessibility of those technologies for the consumers that want them. And it’s something that you can build around if you are very flexible about what type of subscription you’re going to equip your device with because certainly any of the cards you’re buying today from a carrier, it might have the right logo on it, but it’s not going to have the right infrastructure behind it. It’s not equipped to change to those kind of features. So, I know everybody’s been talking about 5G and the edge and everything for a long time, but certainly those are going to start having performance and cost advantages. So if you’re designing something today, preserving optionality, not just to fix your own mistakes but to upgrade things in the future, that’s a big, I’m bringing it all the way home to your question, that’s a big point of advice.
– [Rob] You bring up such a good point though. There’s the radio access network that’s gotten really hot and fast with 5G, which the, which led us to saying, oh, maybe we need to do edge in, at the cell tower base band, stuff right there, in that little shack that’s, and that drove a lot of this edge. Telecom was really late to the party on the edge. Obviously manufacturing was way ahead of them. But the edge is hard because then you’ve got to manage distributed edge nodes all over these base stations, which is a lot harder than just having this kind of one-to-many go into a cloud. And then the problem with the one-to-many with a cloud that most people use today is you have that rockin fast RAN going from your device to the tower, and then you don’t know what kind of backhaul network that tower has and that operator has, getting back to the open internet, then you can’t control the speed of the internet.
I love it whenever I, I’m such a telecom nerd, I’m always posting, hey, I found, here’s the speed test for this particular location, and it really rocked. But likewise, I post it when I go, wow, I’ve got all these bars of mid band 5G, and this app can barely load and nothing’s happening, and why is that? And the average person has no idea, and it’s because the backhaul network sucks from that tower back through that whole pathway.
– [Robby] Maybe we should go on the record for signal bar reform because like it used to be, the signal bars on our phones and devices, like it’s designed around a time when signal was the biggest problem, right? Now, it’s all about link speed and like ingestion. Can we stop having five bars and a buffering icon, right? Can we stop looking like we have full connectivity and throughput and not actually having it.
– [Rob] Again, it all comes down to money. All the telcos are leveraged to their eyeballs right now in debt, like you wouldn’t believe. And so part of the thing is while they spent all this money and they got Spectrum to roll out 5G, the one thing they also needed to do concurrently was speed up the rest of their backhaul network. As I like to remind people all the time, wireless is made of wires. If you don’t have better, faster pipes under the ground all over the place, it won’t make any difference how fast it is from your device to the cell tower.
– [Ryan] So let me ask you a question to wrap up a little bit here. This is going to air at the start of 2024. What do you feel like is, what do you feel like the future looks like for the cellular IoT solution space in 2024? going forward? Where are things going? And one thing I’d like you guys to talk about because I know it’s come up in conversations here is people are starting to get familiar with eSIM, but now people are starting to use iSIM. How is that going to potentially fit into the landscape from what we’ve already been talking about?
– [Robby] My advice to people on iSIM has been wait for the new eSIM standard, which sounds super confusing. Super confusing. Like I was saying, iSIM is a form factor. iSIM is not a new concept. It is literally just saying the memory is now in your modem instead of being an embedded chip, instead of being a plastic card, but it’s still going to run an eSIM operating system, which is really confusing. Almost as confusing as a new standard being called a consumer eSIM for IoT, which is very silly because the two previous standards are consumer or IoT. Now, it’s consumer for IoT, which like, who’s it for?
So iSIM, I’m telling everybody if you’re using iSIM today, there is no, there is no ability, there’s no tooling, there’s a new thing, there’s another thing called SGP41, which is even further than 32. 41 is the, like you’re able to load, side load at a factory what network identities live on that chip. 32 is like a type of eSIM. But you need some system to manage an iSIM. Otherwise, you’re just going to be, not just like embedding something into your device that now can go nowhere, but like etching something into your SoC that now cannot go anywhere. So if you take iSIM from an MVNO, you are stuck with that MVNO now forever. You don’t even have the option of desoldering that equipment. You don’t have the option of removing the plastic card.
– [Rob] When you think about is cellular IoT going to grow, are there any headwinds to that? At the very beginning of the call, Rob, you mentioned Quectel and all that kind of stuff. Are we having a reduced number of module vendors, therefore making it harder to get or driving up pricing or what do you think?
– [Robby] I think with RedCap, the cost will come down across everybody’s product line. So that means that it, everybody will be happy that there’s still reduced competition. But I also think that you’ll continue to see Chinese vendors used in non sensitive applications while there’s maybe not on an entity exemplist or something. Again, I don’t want to say specifically that, I know for a fact Quectel is banned from certain applications because I’m just reading the headlines of concern over using Chinese vendors in these different use cases but.
– [Ryan] So Robbie, let me ask you then, how do you, if I’m, again, listening to this and trying to understand, who should I work with? How do I decide between the different providers to technologies, things like that, what’s the, how can people be successful in the New Year with making those decisions?
– [Robby] It would be timing a product release and getting advanced pricing on the new modules that will be a lot cheaper. That’s a big key, I think next year. It’s not getting burned on the or getting cut by the bleeding edge of things like iSIM that don’t have any software support behind them today. And it’s looking at partners that preserve as much optionality as possible. So things that are open, open standards win, things that create open ecosystems win in the long run. Obviously, I’m steering TEAL towards openness and being able to bring any type of cellular connection you would want, even your own, into this type of platform. And that’s an alternative to going down a closed system solution.
– [Rob] Yeah, what he said. And don’t paint yourself in a corner. When you’re thinking about the whole solution, I can’t tell you, this is an the oldest time product. People start off their solution building it with Wi-Fi just because they’re prototyping, and it works, and then they halfway deploy something, and they realized, oh, wow, I needed it to be cellular.
– [Robby] Even within cellular, think of all the people that deployed NB-IoT solutions, and they needed MQTT. Oops. Oops, make sure your cellular subscription even supports the application datagrams that you’re using. There’s a whole industry now between conversion layers where companies are losing money because they’re spending it converting lightweight M2M into MQTT because AWS, how they designed it, doesn’t support the datagrams that NB-IoT supports, but they were sold on the power savings of NB-IoT and now they, they could swap those devices if they were using a platform like TEAL and something open, but now they can’t so.
– [Ryan] You guys give a lot of good advice for people who are looking to work with organizations to help them pick the right components for a solution, whether it’s the connectivity of the hardware, the software, whatever it is, there’s a lot of questions that need to be asked and also understood by the companies they work with before it’s, it probably makes sense to even deploy anything.
– [Rob] After all these years, it’s apparent to me that the cloud stuff, the platforms, that’s the easy part. It just is. It’s the easiest part of the whole solution. A whole lot of companies made you think that was the core of the IoT solution. It’s the easiest part of all. The stuff, the embedded stuff on the devices, the communication, that turns out to be a lot more difficult to nail. That’s where you see people just painting themselves in a corner and hitting a brick wall. So yeah, you gotta really focus on that.
– [Ryan] For our audience who wants to, connect with either of you after the fact and learn more about the, your companies and what you’re all doing, what’s the best way they can reach out?
– [Robby] For me, either LinkedIn or robbie@teal.io
– [Rob] You can find me on Twitter, X, @RobTiffany or LinkedIn or rtiffany@redbison.com.
– [Ryan] Appreciate it, guys, as always, and we’ll have to do this again soon.
– [Robby] Thank you, Ryan. Thanks Rob.
Industry News
