Latest / The Edge Computing Podcast with Fexingo: Local Compute, CDNs, and Distributed Infrastructure / How Edge Computing Is Transforming Precision Livestock Monitoring
Transcript
- Lucas: If you think about a cattle ranch in Montana, the image that comes to mind probably involves wide-open spaces, horses, and maybe a cowboy on a dusty four-wheeler. What probably doesn't come to mind is a network of edge-compute nodes processing hundreds of data streams in real time. Luna: Right, because ranching feels like one of the last analog industries. But that's changing fast. Lucas: It is. And edge computing is a huge part of that shift. Today I want to talk about a specific deployment — a 50,000-head cattle ranch in eastern Montana that started working with a startup called RanchNet back in early 2025. They've essentially turned a traditional livestock operation into a real-time data-driven enterprise. Luna: What does that look like on the ground? I mean, are we talking about sensors on every cow? Lucas: Pretty close. Each animal wears a GPS collar with a health monitor — heart rate, temperature, movement patterns. That data used to be transmitted via satellite to a cloud server, where it would be processed and then pushed back to a ranch manager's dashboard. The problem is, with 50,000 head, you're talking about millions of data points a day. Satellite bandwidth is expensive, and the latency was minutes not seconds. Luna: Minutes of delay for a sick animal could be a real problem. Lucas: Exactly. A cow showing early signs of illness — maybe it's not moving with the herd, maybe its temperature is spiking — you want to alert the rancher immediately. But with the old cloud pipeline, by the time the data reached the server and got processed, the animal might have wandered off or worsened. RanchNet's solution was to put edge nodes right on the ranch. Each node is a ruggedized compute unit mounted on a pole near a water trough, powered by a small solar panel. It handles data from up to 500 collars. Luna: So the processing happens locally, on the ranch, instead of in a faraway data center. Lucas: Exactly. The collar transmits data over a short-range radio — essentially a custom protocol in the 900-megahertz ISM band — to the nearest edge node. That node runs a lightweight inference model trained to detect anomalies: a cow that's stopped moving for two hours, or a sudden drop in heart rate. The model was trained on years of historical ranch data and can flag a potential issue in under a second. Luna: And the satellite link is only used for what, summary data? Lucas: Right. The edge node sends a compact summary to the cloud every hour — total number of flagged animals, aggregate herd health stats. That's a tiny fraction of the raw data. The ranch says it reduced their satellite data costs by about eighty percent. In dollar terms, they were spending roughly one hundred and fifty thousand dollars a year on satellite uplink fees. Now it's under thirty thousand. Luna: That's a real hard-dollar saving. And the latency improvement is dramatic — from minutes to milliseconds. Lucas: Yes. But the bigger story is the operational change. The ranch manager and the cowboys now get real-time alerts on their phones. If a cow's health flag goes off, they can dispatch someone within minutes, not hours. In the first year of deployment, the ranch reported a thirty percent reduction in mortality among calves — a huge number when you consider each calf is worth around eight hundred dollars at auction. Luna: So the economics are pretty compelling. The edge hardware plus installation probably cost something, but the savings on satellite fees alone would pay for it pretty quickly. Lucas: RanchNet's edge nodes cost about twelve hundred dollars each, and this ranch deployed forty of them — so forty-eight thousand dollars in hardware. Add installation and networking, call it sixty-five thousand. The satellite cost savings alone cover that in about seven months. After that, it's pure savings, plus the value of reduced mortality and better herd management. Luna: And this isn't a one-off. Other ranches are starting to look at similar setups. I read that a beef cooperative in Texas is trialing a version with LoRaWAN instead of the custom radio. Lucas: That's interesting. LoRaWAN has longer range but lower bandwidth. For this application, the custom radio actually works better because they need to send multiple data points per second from each collar — heart rate, temperature, accelerometer data. LoRaWAN would struggle with that volume. But for simpler applications, like just location pings every hour, LoRaWAN could be cheaper. Luna: So the edge node approach really depends on the specific data requirements. High-frequency health monitoring favors the kind of architecture RanchNet built. Lucas: Exactly. And that's the broader lesson for precision agriculture. Edge computing lets you run sophisticated models on high-resolution data without the bandwidth and latency penalties of cloud-only architectures. And it works in remote areas where cellular or satellite connectivity is expensive or unreliable. Luna: It also raises the question: what other old-line industries could benefit from this? We've talked about construction, energy, delivery. But livestock is a great example because the ROI is so clear. Lucas: And it's not just cattle. The same approach is being tested for sheep in Australia and for poultry in the US. For poultry, think of large chicken houses with tens of thousands of birds. Sensors monitor temperature, ammonia levels, bird activity. Edge nodes process that data locally and alert the farmer if ventilation needs adjustment. It's the same principle: local compute, low latency, lower bandwidth costs. Luna: And the hardware is getting cheaper and more rugged. That helps drive adoption. Lucas: Absolutely. The edge nodes RanchNet uses are built with industrial-grade components — they can handle temperature swings from minus thirty to plus fifty Celsius, dust, moisture. And they draw so little power that a small solar panel and a battery keep them running even through cloudy days. Luna: It's a far cry from the typical data center where everything is climate-controlled and redundant. Lucas: Totally different world. But the core idea is the same: process data close to where it's generated. The difference is the environment — harsh, remote, bandwidth-constrained. That's exactly where edge computing shines. Lucas: And this kind of deployment also has implications for rural broadband. If more ranches and farms adopt edge computing, they're essentially building local compute infrastructure that could also serve other purposes — community Wi-Fi, local data storage, even running local AI models for other applications. Luna: That's a really interesting spillover effect. A ranch's edge network could become a node in a broader rural compute grid. Lucas: Exactly. And a couple of dollars a month is genuinely what keeps these shows going — buy me a coffee dot com slash fexingo, if you've gotten something out of them. It really makes a difference. Luna: Yeah, it's a small way to support something you find valuable. And it helps us keep the conversation going without ads. Lucas: So back to the ranch — one of the unexpected benefits was that the real-time alerts actually changed the role of the cowboys. Instead of spending hours on four-wheelers checking fence lines and looking for sick animals, they now get targeted alerts and can respond faster. The ranch manager told me it's made the job more efficient and less physically grueling. Luna: That's a human angle we don't always hear about. Technology replacing not just drudgery but also improving quality of work. Lucas: Right. And for older ranchers who might be skeptical of tech, the results speak for themselves. The numbers don't lie: lower costs, lower mortality, better data for breeding decisions. Edge computing is becoming a practical tool, not just a buzzword. Luna: So where does this go next? Are we going to see edge nodes on every ranch in the next five years? Lucas: I think it's more about a gradual adoption curve. The early adopters are the large operations with clear ROI. But as hardware costs drop and models get better, it'll trickle down to smaller ranches. The key is having a compelling use case, and in livestock, we're seeing that. Luna: And maybe one day, the cowboy's best tool won't be a horse — it'll be an edge node. Lucas: Now that's an image. But honestly, it's already happening. And it's a great example of how edge computing is transforming industries you wouldn't expect.