Transcription

Cellular, for the people, may have finally arrived in the form of this innocuous little chip. A first-of-its-kind, ultra-low-power, single-chip cellular board with a complete open-source software stack. Not built for phones, but instead for everything else. Hailing from Belgium, you won't find this on AliExpress, and it costs almost nothing to run. But promises are easy, so today we're going to put it to the test. Because Walter brings GPS and cellular together, I'm turning this board into a real-world tracker and seeing if it could actually be used to lojack your kids, your pets, maybe even your car. We'll test this in my building and expand outward from there into my neighborhood, city, and beyond to find out exactly what this little guy can do. Walter is a multi-radio system on a module that pairs the ESP32-S3 with a Sequence Monarch 2 LTE-M NB-IoT modem. You get Wi-Fi, Bluetooth, cellular, and GPS, all on a board smaller than a matchbox. Walter combines all these radio techniques in a really small module that's also completely certified, CE-FCC, so you can go to market faster. Now on this channel, we talk about LoRa quite a bit, but this is a different modality with slightly different application and use cases. This is for real internet and not just broadcast packets. With a SIM card, you can roam across countries, access licensed spectrum, and push firmware updates, which is something LoRa just can't do. All right, let's break down the different wireless options you'll see in the IoT world and why Walter's choice of LTE-M and NB-IoT actually matters. First up is LoRa and Sigfox. These are ultra-low power, long-range protocols, but they come with a catch, which is they're pretty slow. And we're talking kilobytes per second for LoRa, and in Sigfox's case, just a few hundred bits per second. And that's fine for things like temperature pings once an hour, but you're not doing real-time control or firmware updates over these. Now let's look at NB-IoT. This sits right in the sweet spot for low-bandwidth sensor data. You get better penetration. It can reach through walls, even underground, but the trade-off is latency. It can take two seconds to hear back from the network, which brings us to LTE-M, which is where Walter really shines. You still get low power draw with modes like PCM and EDRX, but now you're moving up to hundreds of kilobytes per second, and that opens the door for two-way communication, faster cloud sync, even voice or video if you really need it. And for reference, here's 5G. Blazing fast, ultra-low latency, but totally impractical for low-powered embedded systems. It would be like putting a jet engine on a bicycle. Now what does latency mean in practice? It's the delay between when your device sends data and when it gets a response. If you're just logging weather data once a day, latency doesn't matter. But if you're controlling a valve, tracking movement, or expecting a real-time alert, latency can become critical. With Sigfox or NB-IoT, you could wait several seconds or even minutes with weak coverage. LTE-M brings that down to under 200 milliseconds, making it fast enough for real-time control. Let's take a closer look at the board itself. Here's what's packed into this tiny powerhouse. At the heart, you've got an ESP32-S3 microcontroller with native support for Wi-Fi and Bluetooth low energy. Right next to it sits the Sequence Monarch 2 GMO2SP modem, which handles NB-IoT and LTE-M with GPS baked in. The board exposes all I-O pins along castellated edges for easy soldering or integration in your own PCB. There's also a U.FL connector for LTE and GPS antennas, allowing you to use high-efficiency external antennas depending on your deployment. Power-wise, it's got an ultra-high efficiency DC-DC converter and onboard MOSFET power switching. This lets you cut power to peripherals like sensors on command, extending battery life dramatically. And no LEDs means minimal passive drain and no unnecessary components. This board is stripped down for endurance in the field. The LTE antenna supports a wide global frequency range from 698 megahertz all the way to 3 gigahertz. So it works across carriers and continents. WALTR ships with the TaoGlass antenna. These antennas are guaranteed to be available for at least 10 years, which is critical if you're building products meant to last. The GPS antenna is where it gets even more interesting. WALTR uses a passive omnidirectional GPS antenna, which means your device doesn't have to be perfectly positioned to get a fix. Ideal for trackers on bikes, boats, parcels, or gear that moves. Ceramic antennas are directional. They need to face the sky. But omnidirectional antennas receive signals from all angles, which is a big win for real world deployments. And here's the hidden genius. WALTR avoids active GPS antennas entirely. Why? Because their power consumption is unpredictable. With one active antenna, it maybe consumes 60 milliamps, while another active antenna only consumes 7 milliamps, which is a huge difference. So that's why we've chosen to support passive antennas only for the GNSS. In short, predictable performance, better battery life, and a design that respects the reality of how these devices are used in the wild. WALTR comes ready to connect straight out of the box and includes a Soracom IoT SIM card designed specifically for machine-to-machine or M-to-M communications. That means it's optimized for data only, ultra-low power devices like WALTR, and not phones. When I powered it up here in Miami, the board locked onto GPS and connected to the LTE-M network and uploaded to the demo server all automatically. And that's how seamless M-to-M is supposed to be. Now, I know you all loathe monthly subscriptions, and same here. That's why WUNST stands out. They offer a flat one-time payment model, around $10, gets you 10 years of service, and 500 megabytes of LTE-M or MDIoT data, with global coverage in over 100 countries. No recurring fees, no surprise bills, just one and done, hence the name. It's ideal if you're deploying something like a weather station, GPS tracker, or remote sensor that only pings data occasionally. And since WUNST supports PSM and EDRX, it pairs perfectly with WALTR's ultra-low power modes, giving you legit multi-year runtime off a single battery. But if you want more control over your fleet or global data coverage with a clean dashboard, check out Hologram's Hyper-E UICC IoT SIM card. You can order a SIM directly from their site for just $3 USD. No contracts, no setup fees. It comes in a smart card form and works worldwide on LTE-M and NB-IoT. Once it arrives, just insert the nano-SIM into your WALTR, register the SIM ID on Hologram's dashboard, and activate it. In most cases, APN settings will auto-configure, but if not, you can set them manually using the WALTR modem library. Just remember, if you're aiming for multi-year battery life, make sure that your SIM provider supports PSM and EDRX. That's how WALTR sips power between transmissions instead of guzzling it. The included Soracom SIM gets you started instantly, but third-party options like Hologram give you flexibility, scalability, and global reach, all for a few bucks. Here's where WALTR really shines though, which is on the software stack, because it's not locked into some walled garden like some other boards. So how does WALTR compare to something like the Rack5010 with a Questel BG95? Simple. WALTR gives you more control, less power draw, and zero lock-in. Where the Rack5010 ties you to their proprietary RUI toolchain and limits you to Bluetooth Low Energy with a Nordic chip, WALTR runs on an ESP32-S3 with both Wi-Fi and Bluetooth, plus a Sequence Monarch 2 modem tuned for ultra-low power. It's got more compute, more memory, and gives you full access to all I.O. pins. WALTR is fully open-source, both on the hardware and software fronts. Now to be fair, the Rack5010 has its perks too. I like the fact that it comes with a JST battery connector and a solar panel port right out of the box. So similar boards in concept, with slightly different variations. WALTR consumes just 9.5 microamps in deep sleep with the modem in PSM mode. That's elite performance. And WALTR is optimized for the 99% of time your device isn't doing anything. So who's using it? Everyone from scientific buoys in the North Sea to e-bikes to horse pregnancy monitors. One customer connects horses to the internet to predict when they'll give birth. Okay, test number one, in the apartment. So I'm going to test if this guy can pick up a signal from my apartment here in Brickell. So I'm putting the WALTR just outside my window because the GPS needs a line of sight to the sky in order to acquire a fix. And I have it plugged into USB-C into my computer, and then I'm using any generic MCU terminal. In this case, I'm using the flasher.meshtastic web application, which can actually read USB serial device data. And we can see here already it's connected to the network, but it's trying to acquire a GPS fix, which usually takes a little bit longer. Let's see if we can get anything here. And you can see it's saying, started GNSS fix. So this is the output log from the WALTR. I'm actually using the meshtastic flasher. It works right in the browser over USB serial. It's connected here. This is hologram, which is showing the SIM data, the usage, the cell tower, the location based on the cell tower. And then this is the WALTR QuickSpot demo. And you can see that's my node right there. And on the map, it is right in Miami. It immediately shows up pretty straightforward, right in Miami. And then if you want to prove that out, come over here. And I am in Miami, and this is the node. And I propped it up like that because we want that antenna to be facing the sky because that has the GPS location. That rectangle right there is the LTE antenna. That circular one over there is the GPS antenna, which we also want to be able to get a signal so that we can provide our longitude and latitude to these other applications over here. For test two, we're hitting the road. I'm powering this up in my car, and we're going to see if WALTR can actually track my location in real time. Okay, so it says disconnected from the network, connected to the network. So we quickly connected to the network. That's the LTE network. And the next step is we're trying to get a GNSS fix. So that always takes a second, right? And also keep in mind that the GPS passive antenna is in the car, so it is being blocked by a sheet of glass. I don't know what the effect of that is necessarily. We shall find out shortly. We got a GPS fix in about four seconds, and that is with the obstruction of glass. So that's looking pretty good. All right, so now we're over in hologram. Let's see, last connection 33 seconds ago. So this is our active session. Okay, so I can see Verizon Wireless, LTE. We have the tower number here. So I think hologram doesn't know anything about the GPS, actually. So it's able to give us a approximate location based on the tower in use. So let's go over to the QuickSpot demo, which I think actually makes use of the GPS coordinates. Okay, so we see my node, node number one, Dataslayer. Oh, here we go. Here we go. Okay, so now we have a very granular map. Oh, and look, we can actually see, this is amazing, because we're moving right now, right? So we can see the little pin moving the little pin moving in real time here. See that big body of water right there? That is right over there. But here's the brutal one, test three, a thick concrete stairwell smack in the middle of the building. No windows, no line of sight. This is where networks die. So in the stairwell, we were able to connect to the LTE network fairly rapidly, confirming that NB-IoT and LTE-M have pretty good penetration capabilities. But we didn't ever acquire a GPS fix or any longitude-latitude coordinates, which kind of was to be expected, because the GPS antenna needs a line of sight to the sky, even though it's omnidirectional. So you can imagine how this affects certain applications. But even having the LTE connectivity gives you some sense of where it is based on the tower ID, and that deep penetration could be really useful. So on balance, I've been very impressed with the Walter, and particularly for urban settings or settings where you have an ample number of cell towers. This really opens up a whole new set of possibilities in terms of remote monitoring. Now, if you're thinking, this is awesome, but what if I need something totally off-grid, no cell towers, no SIM cards, you're not alone. And that's exactly why I built the Houdini M1, a completely off-grid dog tracker using Meshtastic, built for your escape artists and anywhere cell service just doesn't reach. If you want one, I've got links down below. You can either grab it from my Etsy store or build it yourself with a step-by-step guide. Ever wonder what kind of innovation big tech doesn't want you to know about? Check out this next video. Thanks.