How does AI-enabled applications take the lead at CES 2025?

In this episode of Fibocast, Jim Morrish, founding partner at Transforma Insights, engages in a dynamic conversation with Jim Engelson, Director of Business Development at Fibocom USA, right from the bustling show floor of CES 2025. Together, they delve into the forefront of Artificial Intelligence-enabled Internet of Things (AIoT) applications showcased at the event.

They explore the latest breakthroughs in AI and IoT integration, highlighting innovative technologies such as AI-powered robotics, smart retail solutions, and advanced smart home hubs. Listen now to find out how AIoT is revolutionising various industries by enabling real-time data processing, enhancing connectivity with 5G modules, and creating intelligent devices like cellular-connected shopping carts and smart lawnmowers. Additionally, they examine Fibocom's recent announcements, including the AI stack for seamless AI integration and the AI Buddy personal companion device.

With expert insights and forward-thinking analysis, this episode uncovers how AIoT is transforming the landscape of connectivity, driving efficiency, and opening up new business models.

Join us to explore the cutting-edge developments in AIoT and discover how these technologies are shaping the future of connectivity. 

Transcript

Jim Morrish: [00:00:00] Hi, everybody. Welcome to Fibocast the go to podcast for thought leaders, engineers, and tech enthusiasts in the telecommunication industry. In Fibocast, we deep dive into transformative technologies shaping the future of connectivity, from groundbreaking 5G communications to the unlimited potential of artificial intelligence and the Internet of Things.

We explore how these innovations are transforming the way we communicate, operate, and grow. My name is Jim Morrish. I'm a founding partner at Transforma Insights, a firm of industry analysts. And I'm the host of today's Fibocast on exploring how AI enabled applications have taken the lead at CES 2025.

Today, we are in fact at CES 2025 on the show floor, and we're here to explore the latest innovations with Jim Engelson. Hi, Jim. Could you please introduce yourself and tell us about your role at Fibocom? 

Jim Engelson: Jim, it's a pleasure to be here. [00:01:00] Thanks for the introduction. Yeah. My name's Jim Engelson.

I'm director of business development at Fibocom here in the US and as you said, excited to be here at CES on the show floor.

Jim Morrish: Thanks, Jim. So we're here at CES . Um, what's the most impressive tech that you've seen at CES 2025? 

Jim Engelson: Yeah, it's, you know, first things first CES, I think, just what impressed me the most is that it's fully back.

It's a big show, lots of visitors, lots of vendors, all the show's floor is full, so it's definitely back as a going concern. What kind of impressed me or caught my eye…  yesterday as I walked the floor there was, you know, a few things… robotics. I definitely saw a lot of robot applications, whether it be, you know, sort of personal or healthcare related or just trip, home related as well. Like I think I came across four or five [00:02:00] different kinds of smart robot pool cleaners, for example. Like I didn't know that was such a big thing, but it was, it was amazing. At the auto section, you know, the push and the focus along with autonomous driving. I mean, that's clearly evident.

And then, uh part of the reason why we're here today is to talk about AI. And, that is by far a clear message that everybody's sending. But what, sort of impressed me was AI pushing not only from, you know, we're in the IoT space, but More and more into the consumer world, quick example, I went to the Sony booth and you know, AI was very prevalent as they, you know, had sort of interactive game demo, you know, with, live, people on stage that then translated to a gaming screen, live.

It was quite amazing, actually. 

Jim Morrish: Yeah, definitely. Um, I mean there's a lot [00:03:00] here and I agree with a lot of what you've said. And it's similar to many of the things that I have noticed as I've told the show, a lot of AI, I mean, that really can't be overstated, particularly You know, of course, leading with NVIDIA and their desktop and laptop GPUs.

But, but many other things. I mean, Google TV using Gemini to aid interaction with the TV and Samsung using Vision AI too, so you can search, for example, for an actor. Yeah, so AI really is, um, diffusing into many situations now. And there's a lot more of it, clearly a great significant uptick on the horizon.

Um, It's an interesting development in smart glasses. Smart glasses have been with us for a long time, and it's always been one of those technologies that's never quite arrived. It seems like it might have taken another step to being something real and tangible, and that people can use and that both extends from the more wearable, daily wearable technologies with embedded cameras and GPT assistance, to the more [00:04:00] sophisticated cinema style, AR type solutions, and these things are going to be a boon for training and education in various situations like that, and of course in enterprise situations are very much an opportunity for onboard AI and 5G connectivity to get the most out of those glasses, so it's a great show. A lot's been happening and, and I think our view both is that CES is back with a vengeance, which is great. 

Jim Engelson: Yeah, absolutely. So maybe as we sort of shift gears on AI and focus back on IoT, you know, you see a lot in sort of what you do, what are some of, um, you know, some typical scenarios in the IoT world that, you know, are leveraging AI, that you're seeing.

Jim Morrish: Yeah, absolutely. So, so if we, if we look at the convergence of AI and, and IoT or as many would term that, and we're, we're adopting the term ourselves AIoT, um, I mean, the poster child for this really is CCTV weather is from my view where you have [00:05:00] on board AI on board the IoT device performing things like image recognition and analysis and what that means is that the CCTV device can send triggers and information and refined data Not just your big fat data streaks and what that does is it significantly reduces the bandwidth required to support the CCTV camera it enables more flexible placements of CCTV cameras.

 So for instance, you know, deploying these things on perimeter fences that are solar powered and that just need no wiring at all. Um, and this really opens up the market in terms of flexibility of deployment, the lower cost of the devices. And of course, you know, less people required for monitoring the feeds from these devices.

Um, I'm sure you've seen some examples too, Jim. 

Jim Engelson: Yeah. In fact your examples, it's, it's just. Great segue to,  one of the key sort of use cases that we're seeing initially with quality inspections on the production line. [00:06:00] So, you know, we have an AI powered,  industrial camera that's connected, it has, you know, 4G, 5G, Wi Fi connectivity built in, um, and, you know, these, these production line cameras are then, you know, in real time.

Analysing the production line for any defects,  anomalies and whatnot, and then making real time decisions, to then flag a product for removal or trigger some sort of event on the production line, maybe a stop to, correct things. But, um, yeah, like these, these AI driven cameras or CCTVs are really in the forefront right now.

And. You know, from a Fibocom perspective, we have an entire lineup of,  smart AI driven modules,  that support both, 4G, 5G, as well as Wi Fi, and also [00:07:00] include, powerful CPUs, GPUs, and computing power that then can make these sort of real time decisions,  on the edge,  just like we were talking about, so.

You know, they come in different flavors in terms of platform support, whether it be Linux, Android or Windows. So, they really support a vast array of opportunities for use case deployments. 

Jim Morrish: Yeah, I mean, certainly there's huge potential there, as you've outlined very eloquently. Focusing specifically on AI and IoT integration, can you comment on some of the key technologies, you know, such as, for instance, machine learning, machine vision, on device computation, large language model trading for industrial use and so on, and what about the impact on cellular connectivity models, customers, business models, and so on?

Jim Engelson: Yeah, yeah, yeah. [00:08:00] Definitely, it's, this is all impacting business models and cellular use case, you know, you kind of alluded to it earlier on that, that is the ability to make decisions on the edge,  and have a lot of that sort of,  data being dissected,  you know, at the edge. And then it's only the relevant data that needs to be sent back.

Has really, allowed for less data overall to be sent through the traffic, which means, you know, better business,  cost opportunities for the user,  you know, easier to get to the cloud,  easier to retain from the cloud, stuff like that. In terms of, you know, some of the key technologies, that we're seeing here, as you alluded to, like, One area that, um, we see is up and coming is sort of smart retail,  as an example.

So, there's sort of multiple opportunities here, [00:09:00] but you know, you can have an AI driven camera with cellular backup, but then, for example, when a customer enters,  say a supermarket,  this camera can do facial recognition thereby identifying the customer. And pulling up profiles in terms of, you know, shopping habits what typically they buy and then this could be communicated to, in store,  maybe, robotic, um, you know, assistance or any other sort of navigational assistance, um, that then could help guide the customer, to where they want to be or where they need to go. Um, you know, another example might be, um, AI powered shopping carts, that are also connected through cellular connectivity.

So, you know, the customer comes into the supermarket, starts [00:10:00] filling the cart with the goods. There's a real time sort of tally in terms of what the total bill is where they're at. And then when the customer is ready to check out and they have everything they need. They can execute that transaction right at the shopping cart level.

Um, you know, thereby sort of alleviating the need to go line up or queue up at the checkout. Allows the store to better utilize their employees and staff as well. 

Jim Morrish: Great. I mean, there's a lot of opportunity there. You know, for enhanced business models, um, and value added for consumers, even extending to cellular connected shopping carts, which I have to admit is something I had not heard about until now.

So AIoT will turn up in places that many of us just don't even know it's going to be there. Yeah. Um, so what's Fibcom's role in this fast evolving marketplace? Um, how's [00:11:00] Fibocom, you know, empowering customers to, with value added AI empowerment? 

Jim Engelson: Um, yeah. Actually since this week here at the show Fibocom has made two relevant announcements, um, surrounding AI.

Uh, first is the AI stack. Which is kind of a set of firmware code aimed at simplifying the ease of AI integration into a number of devices, you know, not only ours, but you know, customer devices. So this offers compatibility with, you know, existing AI frameworks like TensorFlow or Onyx and AI models for You know, whether it be audio or LLM, so forth.

So the whole intent here is to improve, you know, AI introduction and time to market [00:12:00] into devices. And then along with that, today we announced AI Buddy, which is a portable device that has one of our smart, cellular modules embedded in there. Uh, this works globally, but it has a screen, um, along with sort of natural voice recognition, integrated chat, GPT, um, so it's kind of a, like, personal AI companion that one can carry so whether you're traveling and you need some translation assistance or, you know, you need help finding a restaurant nearby.

Um, this can all execute that, you know, directly from the AI Buddy device, which runs our Fibocom AI stack. some other areas that were evolved in with AI, um, smart home it's [00:13:00] another popular topic. And actually we've seen a lot of it here at the show. Um, but Fibocom is very successful in the CPE market worldwide.

And we do a lot of business with our partners, the carriers around the world. And what we've recently done is add AI capability to our higher end 5G modules based on Qualcomm's new X75 chipset. And these are then embedded into CPEs for home networking. So these home CPEs not only now provide.

You know, home internet connectivity, but now, you know, carriers can leverage this AI functionality to then have it become like a smart home hub, that's connected. So, I mean, that's very cool. And then you know, lastly, we're [00:14:00] last but not least, we're also involved in things like smarts, robotic lawnmowers, which I've also seen a few here on the show.

So It's definitely an emerging market, whether it be, you know, pool cleaners, as I mentioned, or lawnmowers. Uh, but yeah, having an integrated AI cellular module built into these lawnmower robots, along with sort of higher end RTK GNSS type functionality where we can get sort of centimetre level, you know, location definition.

This allows for a sort of advanced, lawnmower cutting and whether it's just your backyard or a larger space like, you know, a thousand square meters, for example, like the, the opportunities are sort of endless here. Fibocom is very busy in the AI space. On that note, you know, [00:15:00] maybe I can throw one back at you and we've talked about sort of what we're seeing today around AI, but in your mind, what are you seeing coming in the next sort of three to five years?

You know, what's your vision here? 

Jim Morrish: Well, I think a lot of, if you focus specifically in the context of IoT. So an AIoT We've mentioned a lot of the benefits through this discussion. You've  outlined a lot of them and it is probably worth pulling that together, I think. So in general, there's an opportunity for improved performance.

So you can have devices with real time responses and also resilience in case of a wide area connection goes down there can be no compliance and privacy benefits in terms of the data staying on the device rather than being back hold off the device to cloud environments and also there are significant opportunities for reduced operating costs.

So things like, you know, reduced bandwidth utilization, and opportunities for pre-emptive maintenance on more complicated [00:16:00] equipment and many of the value added services that we've discussed I think during the session. So there's a lot of opportunity and it's, yeah, it's, it's developing very quickly.

We as a company, so Transforma Insights, we actually have put together some, you know, really very detailed forecasts of, of AIoT, assessing something like 270 applications. We've come up with a figure of about 9.1 billion AIoT devices in the marketplace in 2033, so that's a CAGR of about 20% from today.

So pretty significant and in terms of penetration of IoT, that's from about 9% today to 23% in 2033. And, you know, obviously led by things like, you know, tele-visions and natural language processing on smart speakers and just many devices and the robotics and the lawnmowers and the pool cleaners, all of these things, you know, they add together to become a really very significant opportunity.

In general, though, the use cases that are leading adoption, the [00:17:00] primary use cases that really justify putting AI on an IoT device, strike us as being basically image processing and analytics, and once you've got that, there's all sorts of other things you can do like exception and identification if people move into areas they shouldn't, but image processing and analytics is the core.

And the other core AIoT application really is natural language processing. And again, once you process natural language, there's a lot you can do with it, like translating it, for example. So what are your thoughts on the impact of AI and how that's changing IoT?

Jim Engelson: Yeah, you definitely struck a chord there with your insight.

And, you know, it rings true to what we're active today with. You know, we talked about the cameras and, whatnot, but, you know, going forward,  we talked a little bit about the business models and,  you know, some of the benefits that,  customers can get from AI, you know, things like, lowering data rates and the [00:18:00] amount of data usage and whatnot and the ability to push more, decisions being made at the edge.

And I think, you know, more and more, we're going to see more devices take advantage of that. So it's not just cameras like today, but, you know, tomorrow it's going to be more, sensors so, you know, we're going to kind of go down the,  the food chain, but just means more and more devices, as you alluded to that are going to be out there connected and leveraging AI.

So I think it's a huge opportunity. 

Jim Morrish: Absolutely. I definitely agree. Jim, this has been a fascinating conversation, and we could continue for a long time, I'm sure, but we are running out of time. Do you have any closing thoughts? 

Jim Engelson: Yeah, just one. First of all, I wanted to thank you for your time, and for sitting down with us.

It's been a great discussion. Really enjoyed your insights into [00:19:00] You know AI and the IoT world and you know, we're looking forward to seeing your predictions come true over the next couple years you know CES is a is really showcasing AI this year and I expect it to be even ten-fold next year So I want to thank you for bringing your insight To not only myself but our audience And for anybody who's interested,  learning more, feel free to reach out.

Not only you can, um, check out our website at Fibocom.com, but also reach out directly and we'd love to connect and share more about what Fibocom is doing in the AI space in your local market. 

Jim Morrish: Thank you, Jim, and thank you for your time. I've definitely enjoyed the chat and the insights even extending to connecting shopping carts, so that's a great example I'll be storing for reference in future. No doubt. And thanks to [00:20:00] everyone out there. Thank you for listening To hear more join the Fibocast, Fibicom's new podcast coming to iot-now.com 

Look out for more details on the Fibocast on social media and web channels.

Thank you everyone.