IoT from Rhinos to Rockets: How LoRaWAN Powers 125 Million Connected Devices

Ruth: Mm-hmm.

Alper: That GSM and its derivatives require. So just like WiFi, anyone can go buy LoRaWAN Gateway and plug it into their local area network. And then they built their IoT network based on LoRaWAN and it's long range with the ability to penetrate through the walls and other obstacles. So when people deploy LoRaWAN in open space where the extreme cases, there's a perfect line of sight. The range of communication between an IoT sensor and the LoRaWAN gateway can be as large as 600 kilometres.

Ruth: Wow.

Alper: Yeah, it's really mind-boggling. Right? That's exactly why we have LoRa Alliance member companies placing LoRaWAN gateways on the low-earth orbiting satellites.

Ruth: Mm-hmm.

Alper: To pick data from sensors on the ground, such as Planet and Lacuna Space, to commercial LEO operators, members of LoRa Alliance have deployed this technology to provide planet-wide coverage.

Obviously, coming to an enterprise, they don't typically need a 600-kilometre range, right? What they need is a campus-wide coverage or building-wide coverage. When you constrain the signal inside buildings, it uses power or more technically, its sensitivity to penetrate through the walls. That's why with very few indoor LoRaWAN gateways, one is able to cover a whole building, a whole campus. In a typical building, a LoRaWAN gateway placed on a floor provides coverage for two floors up and two floors down, and penetrating through multiple walls on the same floor. Typically not only an indoor coverage is provided, but the signals go outside and they also provide outside coverage as well.

Ruth: Mm-hmm.

Alper: And coming back to your question, enterprise, they tend to prefer building and owning and operating their own private network, much like they do it with WiFi. So they own the gateways and then they decide what devices connect and what devices cannot. Same thing is applicable to LoRaWAN as well.

They easily and at a very low cost build a campus wide network. And then they provision devices of their own. And even though the signals can physically go outside, they don't let others eavesdrop or connect because we have full cryptographic security protecting these private networks from any unauthorised access.

Ruth: Terrific. And you mentioned when we talked before that you are the most deployed LPWAN outside of China. How do you compare with NB-IoT or Mioty?

Alper: Sure. So when LoRa started back in 2015, there was already couple technologies in the so-called LPWAN area. LPWAN stands for low-power wide-area networking, the mix of providing long range while consuming little power. That's the category that we have been operating.

So before LoRaWAN came along, there was already a few other technologies, most notably Sigfox. Then after 2015, other technologies started emerging. Seeing that there's a real market for LPWAN and seeing that LoRaWAN is gaining so much momentum, the market opportunity became clear to the other players as well, and that's when the competing technologies started emerging.

And within this 10 years of time, LoRaWAN clearly took the pole position. Sigfox having had a two year head start, meaning that they had started deploying two years before LoRaWAN did. They had a pretty good nationwide networks deployed in several countries, but given that technology was a closed technology, not an open one like having an alliance behind it.

Ruth: Mm-hmm.

Alper: The technology was owned and operated by a single company that prevented it from scaling. So today there are Sigfox Networks, but they have lost their momentum and they're far behind LoRaWAN. And then 3GPP, obviously, typically 3GPP and the GSM industry, they're pretty reactive.

Unless they see a huge opportunity, they don't mobilise until they see someone else's capitalising on such opportunities. Well, that was the case, for example, back in the days with WiMAX and LTE. So those were two competing 4G technologies, WiMAX coming from outside GSM.

Ruth: Mm-hmm.

Alper: And seeing that it was already gaining momentum. Then the 3GPP reacted by building LTE, a competing technology. Those two technologies, one from GSM, one from WiMAX Forum, they were on a head-on collision course. And for that one had to win. There was no room for two equivalent technologies. And that's what happened with the LTE winning. Now, the similar story started unfolding with NB-IoT.

But NB-IoT and LoRaWAN are not really on a head-on collision 'cause they're, the functionality wise, they're not equivalent. NB-IoT is using a licensed band. It's only available to operators. So you or your colleagues or anyone else cannot deploy it freely on their own like one can do with WiFi or LoRaWAN.

Ruth: Okay.

Alper: And it is nowhere near as low power as LoRaWAN is, because LoRaWAN was built from scratch for efficiency. Whereas NB-IoT is a retrofitted version of 4G LTE. So can never be as simplified as low cost and as low power as LoRaWAN is. And hence, even though NB-IoT did have a spike in the past, it started losing steam both internationally and even in China.

Operators like DoCoMo and AT&T pulled the plug on NB-IoT. Those services are discontinued. Internationally, momentum has been significantly down. Now, I do mention China because China has a special position here. The government has endorsed NB-IoT back in 2019 as the wide area IoT network.

That's why majority, like 99% of NB-IoT connections are in China. And even that has been slowing down, basically yielding the floor to Cat-1 bis, another technology, which is more power hungry and also costlier than NB-IoT. So the momentum has been slowing down in China and it's been pretty much, I would say, fizzled away internationally.

As such, discounting the number of devices in China, in the international markets, LoRaWAN has more devices connected and also much faster growth than NB-IoT.

Ruth: Mm-hmm.

Alper: In China again, the growth of LoRaWAN is higher than NB-IoT, but the sheer numbers of their deployments is still more than LoRaWAN in China.

Ruth: Yeah, it's interesting. I imagine being a decision maker or business decision maker, it's always really hard to keep track of all the different standards and all the different technologies that at the beginning of the innovation phase are all competing against each other until one steps ahead, or makes a real business case or real interesting use cases where you can then translate your own needs to what solution fits best.

Alper: Right, right. Exactly. And then there are a few other technologies as well that has been around in the past, I mean, since the beginning, like, Wi-SUN has been around, which is mostly used for electric metering and street lighting. So, compared to LoRaWAN, I mean, LoRaWAN is serving basically all verticals.

It's not just one or two use cases. Practically any use case of IoT can be served by LoRaWAN

Ruth: Mm-hmm.

Alper: with two exceptions being one is like if a use case requires transmitting megabytes and gigabytes of data per sensor per day, that is not for LoRaWAN. Because we're using unlicensed band and long range, we cannot transmit such big data.

So if we're talking about connecting a camera, it's not for LoRaWAN or connecting a car that tends to transmit megabytes and gigabytes of data that's not for LoRaWAN. And the other one is, if the use case requires a truly real-time communication, meaning that guaranteed delivery under milliseconds, at the millisecond’s granularity, say you are controlling a robotic surgical arm da Vinci arm, right?

That you press a button from one continent, and the patient is operated in an operating room in another continent, then you need guaranteed latency. Then definitely LoRaWAN is not meant for that. Or if you're controlling a drone that you press a button and you want that drone to make a turn in split second, guaranteed delivery for those kind of things.

One needs a licensed band technology like NB-IoT or LTE-M.

Ruth: Mm-hmm.

Alper: And then there were even other technologies that popped up and disappeared quickly, like RPMA and other technology that came into LPWAN, but we don't hear about it anymore. And then there's WirePas, and then the latest entry is Matter.

Now when we look at Matter, the kind of features the Matter brings to the table is basically superseded by anything we already have with LoRaWAN all the way from being low power, long range, providing high density networks. In terms of the density, we see the highest density of networks when we place base stations high up as high as, like I said before, 600 kilometres.

When you mount a base station that high, the coverage it provides go all the way up to 3000 kilometres. Then millions of devices fall under that coverage, looking at the sky. That is the kind of density we're talking about. And for that density we already have a modulation technique supported by LoRaWAN called LR-FHSS, for that level of scalability, that's only, we only face that kind of challenge.

For the satellite networks, this type of density is not required on the ground and for any density one has on the ground. LoRa modulation with the LoRaWAN is already sufficient, and that's what our utility operators are already proving in the field.

Ruth: We have so much to unpack right now. Maybe we should dive into use cases now and make it more tangible. You've mentioned the verticals, smart cities, agriculture, utilities.

Tell us about your most transformative use cases you have seen in the past. What has really stood out for you?

Alper: Yeah, so we have so many use cases every day. We are discovering one. That's the beauty of this technology being highly accessible, you know, from all regions of the world and all walks of life from, you know, students doing projects to multinationals. So, what I find interesting, and I think people would find interesting is, with this technology, one example is, in the smart buildings or facility management, given its low cost, low power, high penetration nature.

We see tremendous amount of momentum. For example, in North America, more than half of the Starbucks stores have deployed LoRaWAN for cold chain monitoring. So when people are picking their favourite drink from Starbucks, there's a pretty good chance they're already inside the LoRaWAN network. But that's a private network built by Starbucks for monitoring the temperature sensors they have in more than 10 places in a typical Starbucks store. The motivation behind that was to save time because up until they connected their sensors wirelessly, one had to spend 25 minutes per store per day to check on those temperature sensors. Now, those 25 minutes are going towards making their customers happier.

Ruth: Mm-hmm.

Alper: That's a saving now jumping from a Starbucks coffee shop in United States to, to say Africa. Right. So I'm gonna take you through a little tour here.

Ruth: Yeah. That's nice.

Alper: We've got more than 35 National Parks, wildlife Parks across 10 nations have deployed LoRaWAN for helping basically the planet.

What they do is they track endangered species like black rhinos and elephants that has LoRaWAN connected trackers. And through that they can monitor their movement to check up on their health and also behaviour which can reveal if they are getting sick or if they're under attack or being chased by poachers. So that has significantly reduced the poaching activity in those parks. On top of that, they also track the ranger vehicles and the rangers themselves to make sure they're patrolling the region, which also further reduces the poaching activity and even the ranger dogs do have trackers and there are video clips of IoT dashboards showing how the poachers get busted by the dogs chasing them.

You can see them all on IoT dashboards and tracking the speed of the vehicle to prevent overpopulation and overspeeding, things like that. Another truly life changing use case is the panic buttons. So we have a member company in LoRa Alliance and they built a panic button solution using LoRaWAN.

With panic button, what one needs is a long battery lifetime. So it's not dead when you need it. And also the ability for the signals to go through the wall, so every corner of a facility and outside can be covered. So that member has deployed more than 700,000 panic buttons across 15,000 locations in North America again.

And they're mostly used at schools, hospitals, and hotels. And there were even incidents where these panic buttons were engaged, which dispatched the law enforcement to arrive at the scene under two minutes to save lives. Literally lives have been saved by the LoRaWAN connected panic buttons.

Ruth: Mm-hmm.

Alper: So whether it's like saving people or planet and all the business cases like operating oil refineries, steel manufacturing sites connecting more than 5 million water meters across France using a nationwide network by Orange. Those are the typical cases that everyone has heard.

What is really special with our cases that we have this happening already unfolding, none of these are theory. These are real hard numbers in the field and touching lives all across the globe.

Ruth: That's incredible. I find it really inspiring how tech can help solve the critical problems that we are facing as humanity and the planet in general, and especially the anti-poaching and environmental use cases I find always very interesting.

We had a guest on the show last year. He was building with chemical sensors and not an internet of things, but an internet of trees. The sensors help prevent wildfires before they start.

Alper: Probably the same, could be right. Dryad

Ruth: Yes. Thank you. We had the CEO of Dryad on the show last year and it was incredible what they were doing with the chemical sensors and how much area of a forest they can cover just by these small, tiny things they can put on trees.

Alper: Exactly. So, you know, as you must have already covered in the show, just briefly, what is really special with their solution is it detects the ignition, unlike the other solutions that requires the fire to grow a point that there's a visible smoke or fire that can be detected by cameras or satellites and all that.

This does catch the fire at the ignition within the first two minutes, and now once it catches, what is it gonna do? It has to signal that. Now, that's where LoRa comes into the picture. Long range nature helps the sensors be embedded in the forest and the data being picked by long range communication, whether it's done terrestrially or also they do have direct to satellite communication as well, using LoRaWAN again, so obviously.

LoRaWAN is part of their solution, but it's a critical part of it by providing the long range. 'Cause those sensors have to communicate with receivers that are far outside the forest.

Ruth: Yes.

Alper: And obviously, I mean, the nature is acting up, right? I mean there's all these nature disasters are increasing in number and intensity.

Fire, flood, earthquakes and hurricanes and all that. Now, yeah, LoRa is doing its part already. And again, it's the same thing being long range, low power, low cost. That's the enabler. That's why you don't see ZigBee or Bluetooth or NB-IoT based solution in any one of these cases.

'Cause where these things are happening, there's no way you can build a wide area network using a short range radio, like WiFi, ZigBee, Z-Wave, Bluetooth. So that's where we stand out. And the market or the people is noticing that already, and leveraging it in many ways that is helping people and businesses as well.

Ruth: What would you think are the biggest integration challenges businesses face?

Alper: Well, I mean the IoT itself is really challenging. It's been around for almost four decades

Ruth: Hmm.

Alper: Obviously we are operating inside the IoT realm and we are affected by that as well. And IoT at first appears like a huge opportunity, like billions of devices, right? But when someone starts looking little closer, they realise that billion device market is so fragmented across regions, across use cases, and even adjacent use cases end up requiring very distinct efforts.

So, you know, once you look into IoT, you go like, okay, tracking looks interesting. Then you look into tracking, oh, there is people tracking equipment, tracking, other asset tracking. Then you go like, okay, how about animal tracking? Then there's cat tracking, dog tracking, cow tracking. These are all also distinct, and then all of a sudden that billion unit market becomes like, say, a million unit market. That's where the challenge begins. The second challenge is for cracking that market, you need to integrate several things like a chip, a device, a network, an application. And then you need to have support, you need to have a sales channel, marketing, et cetera. And that's another challenge.

And that's why cracking all these markets have been taking so long overall in IoT. And that's why the things don't happen overnight with LoRaWAN either, but I can tell you this, I mean, the IoT market has been slow progressing, but LoRaWAN has been accelerating that. So when we reach 125 million devices with 20 percent growth year over year, we are far ahead of the typical IoT growth for any other IoT technology.

So we are pushing the envelope here, which is not only accelerating the market for LoRaWAN using IoT cases, but also the rest is also coming along. 'Cause a device built for LoRaWAN may be used for other technologies as well. So we are basically rising the whole market for everyone. But yeah, the biggest challenge is integrating all these components and also the market is still somewhat in an education mode for the IoT.

The prospects understanding their needs, the prospects understanding the possibilities, what can be done, what cannot be done. And then the IoT companies understanding the verticals. So there's this two-way walk towards each other. And it's really a boutique activity. Many companies envision that they'll build a product and then it'll be so horizontal. It'll be used by several millions of devices. Unfortunately, it requires a lot of boutique activity to crack one market and replicate it to the other. But obviously we are again, on an exponential growth year. It is only getting better and faster. But I'm just telling why it took us so long to get here, nearly 40 years, but the next ten, five, even couple years would look much different, much brighter, and they'll look much faster than the past 40 years.

Ruth: You have mentioned talking about the future, I suppose that also AI and edge computing is shaping LoRaWAN as well. How are you adapting to that?

Alper: Yeah, so one of the reasons IoT will accelerate putting aside anything I said is AI, right? Because

Ruth: Hmm.

Alper: It obviously has tremendous potential where you're still at the very beginning, right? And the AI is starting with digital data by using the data that's already available digitally.

Now, the next challenge for AI would be to sense the physical world and command the physical world. That's where LoRaWAN comes into the picture. We are providing the foundation of massive IoT. We are the glue between the digital world and the physical world.

Ruth: The data the sensors are collecting, right?

Alper: Effectively what we're building is the digital nervous system of the planet.

Ruth: Yeah.

Alper: Sensing what's happening, turning it into digital data, and transmitting that to the brain. AI and AI collections, and they do whatever they do. And if they need to command something on the physical world, again, the digital nervous system, LoRaWAN, carries it back to the physical world.

Like opening a valve or closing a door or ringing a bell, right? So that is the big picture. Now, more specifically, there are LoRaWAN networks that leverages edge computing, meaning that LoRaWAN gateway and the LoRa core network, which manages the network of multiple gateways.

There are several products that combine them in a single box physical unit and also have the data processing. So all of the IoT is completely carried out on customer premise data not leaving, and that happens for a couple reasons. One, the site might be off the grid. There might not be any internet or a reliable internet.

That's why all the decision and processing needs to take place on site or two for data privacy. Certain organisations require data not to physically leave the site. So that's why we have edge computing and that's where we also see the AI marrying these edge computing to do full processing at the site.

So that's where physically LoRaWAN and AI touches each other. And in fact, there's even a deeper integration. We see several use cases where the AI machine learning is running on the sensor itself. So we have a class of sensors that processes high amount of data, like image processing, or vibration processing.

There are such sensors. The vibration sensors are used for predictive maintenance. The image processing is used for things like detecting people or counting people, but because the device doesn't want to transmit all this big data, it does a processing on the device, like counting how many people are in a retail store or detecting if the vibration of a rotating equipment has deviated from its normal, which is an alarm situation.

Right? And then conveying that using LoRaWAN. So, yeah, we have this type of sensors that has AI running on the sensor and communicating the outcome of the sensing using LoRa. So we have all sorts of integration already. And like I said, both LoRa and AI will accelerate by collaborating, feeding to each other, taking advantage of data flows with respect to the overall digital transformation.

Ruth: I like your nervous system analogy. That's really cool. You have mentioned public versus private networks. How would you advise, how should enterprises decide what?

Alper: So, we have multiple public networks all around the world. Some of them are fully nationwide. Like in France, there's Orange providing nationwide network. In Switzerland, Swisscom in Netherlands, KPN

Ruth: Mm-hmm.

Alper: In Austria, A1 Telekom Austria and then we also have public network operators that build, that has citywide networks already deployed, and that can bring networks on demand. So when an enterprise needs LoRaWAN connectivity, they should first see if there's already a public network operating in their country, providing the coverage they need.

Sometimes physically the network is already there and sometimes it is not there, but there is a public network they can contact and get an on demand network built for them to start with, which gets shared by the others. So it is still a public network. If there's none of these options, that enterprise would build its own LoRaWAN network unlike cellular 'cause with the cellular network, that's not an option. 'Cause that enterprise wouldn't have a licence. It would be too costly to build 4G 5G network. So that's a non-starter. But with LoRaWAN, there is this low cost, high accessible nature.

So if there's no public network option, then they can build their own network.

Ruth: Mm-hmm.

Alper: Now. If there's a public network, obviously they can make a judgement like, should I build my own network or should I use a public network? There's a cost, there's an availability, and it's very parametric, right?

But then there's one other element. So if the devices of the enterprise is always on campus, obviously building a private network can serve that. But if some of those devices are outside their campus like going around the nation or other countries. Obviously that enterprise wouldn't build a nationwide network, and that's where actually public and private collaboration comes into the picture.

So an enterprise can build its own network, a private network to connect its own devices and also work with a public operator and set up a roaming agreement

Ruth: Mm-hmm.

Alper: such that the devices when they leave their campus. They can roam into the public network and stay connected. So, with LoRaWAN, one can build public, private, and even community networks.

So the ability to support three distinct type of networks is very unique with LoRaWAN, no other technology can do that, but on top of that, we have the ability to integrate them together so a device can roam from a community network to a private network to public network. And that is definitely not seen anywhere else with another technology either.

And there are multiple examples of these. For example, in Romania, a company called Adeunis, they have their own private network for connecting water meters, but they also have roaming with Helium network, a community network. Such that if the data packets are received by Helium Network, those are also forwarded to the Adeunis core network.

So that is an example of a private, public roaming. Similarly in France, Veolia with 5 million water meters, they have their own private network, but some of those water meters are also roaming into the public Orange network.

Ruth: Mm-hmm.

Alper: So we have this collaboration, collaborative technology, and also collaborative ecosystem.

Ruth: And it's open source as well, right?

Alper: Well, these are open standard, meaning that all of the work we do with the LoRa Alliance as we build the standard interoperable specifications are publicly available. Anyone can download them, read them, and implement them. On top of that, we also provide an open source implementation. Our specification so the device makers and solution makers can use open source implementation as well.

So it's an open standard open source implementation, and it's an open alliance. Anyone can join LoRa Alliance. This is the membership organisation. Global, nonprofit technology association. And we have members from all around the world, all the way from universities to operators, to system integrators, device makers.

So I would highly encourage the audience to check it out, lora-alliance.org, see our membership, see the benefits, whether you want to follow the technology closely or certify your devices. Or you have the ambition to steer the technology to join this journey as we are revolutionising the telecommunication industry.

Everyone is welcome to join us. Like you said earlier, we have more than 350 members and the numbers are growing fast. It's a highly collaborative environment. All these companies are bonded together under the mission of making massive IoT reality and we are on a great track right now it's very exciting.

Ruth: It's really clear that this long range, low power networks are enabling complex mission critical applications.

It's really fantastic. So thank you Alper, for coming on the show today and sharing your expertise. It has been truly insightful. Thank you very much for being here today.

Alper: Thank you very much, Ruth. Thank you for your questions and having me this chance to explain what we do in the industry.

Ruth: Until next time, keep curious and keep innovating.