Annie: Yeah, I can start. So, I think that's really application-dependent. We've built the software ourselves, to provide the information, both to smartphones and to standard laptops — all sorts of screen sizes. The thing here is that this really comes to the core of being able to adapt to use cases, and what we've done in terms of adapt to use cases. It's not easy to create one user interface that is adapted to a variety of use cases. There are some places you can start, in a free and open source-type presentation, with graphics-type libraries out there for JavaScript, and that might be a good place to start. But there are also cloud services, where you can basically connect up IoT devices. But again, that's much more generic, and not adapted to use cases, and if you want to provide your services or create something for a specific market, I really highly recommend trying to work on the user interface, and put lots of thinking into that.

Bob: Yeah. We realize that, and Digi has a cloud service that connects to our devices. It's called Digi Remote Manager®. And it displays, in table format, different gateways and things, and says whether they're up or down. And what we've realized, in the embedded space, it's exactly what Annie says. You need a different user interface. So we focused on the APIs that come out of that device management solution, Digi Remote Manager, and we've recently built a complete stack of, I want to say JSON software, that gives iPad developers or Android developers, for the iPhone, a reference implementation that shows exactly how to connect to our devices, and then how to build that user interface.

Like Annie says, the user interface is all-important, and we don't want to show a table of devices. Our first example is a smart city application, where we show buses moving around a city. But we recognize that every application is different. You want to show a field where you can click on the different stations and see what's going on in each one of them, or some of our customers who control brake pressure on a truck want to see a representation of the truck and check the brakes. So, we do provide that. We have a Digi XBee® mobile app, and a reference implementation that lets our customers build that user interface as part of the complete solution. It's just another area where the Internet of Things is a challenge, but it's what we really need to do to make it valuable for the customers.

Bob: Yeah, that's a great question. We are looking very closely at satellite, with some of our technology. Lots of networks are going up there. Starlink is probably the one that's most well-known. And we are looking at our different cellular modules that can provide direct-to-cellular technology in the future. We're also looking at some of our modules that are sub-gigahertz, that can provide a connection to a low earth orbit satellite, very easily, because there's nothing in the way. Some of the private networking solutions that we've designed can go to a satellite as well. So, the answer is nothing available now, but in places where you're beyond the range of a cellular, we think satellite's a good solution, and we're working on some things that you'll see in the future.

Bob: You know, that's a very specific application, but we do, believe it or not, have a success story on our website, that talks about aquaculture and fish farms. That one is in Latin America. So, yes, we have a customer who is doing that.

Annie: Can I start there?

Bob: Yeah. Go ahead, Annie.

Annie: We've been on that journey. And I would say low orbit satellite, or DigiMesh, or, I think we're talking about the right communication for the right application. I don't think it is an either/or. I think it's both. It depends on the application. Many of our DigiMesh networks are now connected to Starlink. Starlink is basically connected to the gateway, and is the most popular kind of backhaul to the Internet for our IoT networks. Part of those networks will also contain some mobile, Cat M1 type of connection, using the XBee 3 Cellular module. So it's a multi-communication, of a mixed network.

Bob: Yeah. And the nice thing about XBee is that we have a standard form factor and standard interface to all the XBees, and all the different technologies and protocols. So, we also recommend the best protocol for the use case. And it's about the data rates and the distances and the density, and passing, where the signal goes on the way back to the base station. So, with XBee, you don't have to make a choice right up front. You can stay with a standard technology, a standard form factor and standard software with XBee, and then select the best protocol for each application.

Annie: Absolutely.

Annie: Yeah, we've done quite a few of those, and I think I mentioned this in my presentation. I think extremely important to build competencies in radio planning. Radio planning is important. It's really not too complicated, but you need to put some thinking into how you create your mesh networks. And then they scale amazingly, but you have to remember that you have to keep within six hops. You have to keep within the specifications or the kind of the potential and limitations of a network like this. But if you do, you can scale very well.

Bob: Yeah, and from our viewpoint, over the course of the 20 years that we've been doing this, we see a transition from traditional product development cycles to a more of a proof of concept. And that's why we focus with our SparkFun partnership, is that we build a kit where you can open it up and hook up sensors, connect to a network, and right out of the box, build a proof of concept. Then you can show an end customer, "Hey this is what the system's going to look like."

And some of the newer technologies, like Arduino and Raspberry Pi, have made that possible. So, instead of spending five years developing a product and bringing it to production and launching it and hope it meets somebody's requirements that you saw five years ago, you can very quickly put together a modular solution, put it in front of somebody, and say, "Is this something you can use?" Then you can tweak it, and then use that same technology you're doing a proof of concept with to scale straight into production with XBees and different things. So, that's really a shift that we've seen recently.

Annie: I think as Bob said, getting these six reference farms that we had initially, and working with them to really understand their use cases, and then translate that into what we would call development iterations, not sort of having a big waterfall development for like three years, but really rapid iterations of development, that really solved a lot of the initial challenges that we had with scaling and how we create the systems, and are they maintainable, and a lot of these details that had to fall into place. When we got to that point, I mean, it was really around, for the first two to three years we had10 to 12 different iterations in hardware development. Really rapid development.

Bob: Yeah, and the nice thing about DigiMesh in particular is it's a simplified networking protocol, so you don't have to worry as much about the complexities of networking, and you can concentrate on that application.