Latest / Internet of Things with Fexingo: Connected Devices, Sensors, and Industrial IoT / How IoT Sensors Are Transforming Wine Fermentation Monitoring
Transcript
- Lucas: So picture this: you're a winemaker in Napa Valley, and it's harvest season. You've got fifty thousand gallons of Cabernet Sauvignon sitting in stainless steel tanks, each one fermenting at its own pace. If the temperature in one tank spikes just two degrees overnight — you could lose the whole batch. Luna: And traditionally, you'd have someone walking the rows of tanks with a handheld thermometer, checking maybe twice a shift. Lucas: Right. And that works — until it doesn't. But here's what's changing: a growing number of wineries are deploying IoT sensor networks that monitor fermentation in real time. We're talking temperature, pH, specific gravity — every variable that matters — transmitted wirelessly every few minutes. Luna: I've actually read about a few Napa estates doing this. One that stood out was a mid-size producer — I think they had about 150 acres under vine. They installed something like 200 sensors across their fermentation hall. Lucas: That's the exact case I want to focus on. So this is a family-owned winery — let's call it 'Stonebridge Vineyards' for the sake of discussion — they've been operating since the 1970s. In 2024, they decided to retrofit their entire fermentation operation with a mesh network of sensors. Luna: And what drove that decision? They must have had a specific problem. Lucas: They did. In the 2023 harvest, they lost two full tanks — roughly 30,000 bottles' worth — because a cooling valve failed overnight and the night crew didn't catch it until morning. That's about half a million dollars in potential revenue, gone. Luna: Ouch. So the IoT investment was essentially an insurance policy. Lucas: Exactly. And the numbers bear that out. They spent roughly $40,000 on the sensor hardware, installation, and the first year of the software platform. In their first harvest with the system, they reduced batch loss by 18 percent. That's not just the catastrophic losses — it's also catching subtle deviations that would have degraded quality. Lucas: Quick honest thing — a handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that's literally what funds making this many episodes. No ads, no sponsors — just people who find value in these conversations. If today's tech talk gave you something usable, that's the way to keep it going. Luna: Yeah, and we really appreciate those who do. It makes a difference. Lucas: So back to the sensors — the really interesting part is the tech stack they chose. Stonebridge went with LoRaWAN — that's low-power wide-area networking — rather than Wi-Fi or cellular. Luna: Why LoRaWAN? I'd have thought Wi-Fi would be simpler since the fermentation hall is probably within range of the main building's network. Lucas: Good question. The issue is density. A fermentation hall is essentially a metal box filled with large stainless steel tanks. Wi-Fi signals bounce off those surfaces, creating dead zones. LoRaWAN, on the other hand, penetrates much better through metal and concrete. Plus, each sensor runs for years on a single coin cell battery — no need to run power or change batteries mid-harvest. Luna: That makes sense. And the sensors themselves — are they off-the-shelf or custom? Lucas: Mostly off-the-shelf, but with specialized probes. For temperature, they used standard DS18B20 digital sensors — the same kind you'd find in a home weather station, but with a stainless steel sheath rated for food contact. The pH sensors are a bit more specialized — they need to withstand the acidic environment of fermenting wine and be cleaned with hot caustic solutions between batches. Lucas: The specific gravity measurement is the trickiest. Traditionally, winemakers take a sample and float a hydrometer in it — it's manual and takes time. Stonebridge used a vibrating tube densitometer, which is essentially a tiny tuning fork that changes frequency based on the density of the liquid. That gets transmitted every 15 minutes. Luna: So the winemaker can watch the fermentation curve in real time on a tablet. That must change how they work. Lucas: It does, and not everyone loves it. The head winemaker at Stonebridge — a guy named Carlos who's been making wine for 30 years — told me that at first, he felt like the sensors were second-guessing him. He'd taste the wine and think it was ready for the next step, but the data would say the specific gravity was still dropping. He learned to trust the numbers. Luna: There's always that tension between craft and data. But I imagine the data becomes another tool, not a replacement. Lucas: Exactly. And the really powerful use case isn't just monitoring — it's prediction. Stonebridge fed two years of sensor data into a machine learning model that can now predict when a fermentation is going to deviate from the ideal curve. They get an alert six hours before a problem becomes visible to the human eye. Luna: Six hours — that's enough time to adjust temperature, add nutrients, or even intervene physically. Lucas: Right. And that's where the ROI compounds. The $40,000 upfront cost was recouped in the first year just from the batch loss reduction. But the real value is in quality consistency. When you can keep every tank within a tighter window, the overall wine is more uniform. That matters for a brand that sells at $50 a bottle. Luna: Are other wineries adopting this? Or is Stonebridge an outlier? Lucas: It's spreading fast. I've heard of similar deployments in Bordeaux, in Mendoza, even in Oregon. The technology cost has dropped about 40 percent in the last three years as LoRaWAN chipsets have become commoditized. A basic temperature sensor module now costs under $10 in volume. Luna: So the barrier to entry is getting lower. I could see small boutique wineries starting with just a handful of sensors on their most critical tanks. Lucas: Absolutely. And that's the pattern we see in industrial IoT across the board — start with a high-value pain point, prove the ROI, then expand. Stonebridge is now adding sensors to their barrel aging cave — humidity and temperature — and they're experimenting with near-infrared spectroscopy to measure sugar and alcohol content without taking a sample. Luna: That's wild — no more pulling a sample and running to the lab. The sensor just tells you. Lucas: Exactly. And that's the direction of travel. The interesting question for the broader industry is: how much of winemaking becomes automated? Some people worry that we'll lose the art. But I think Carlos put it best — he said, 'The sensors don't make the wine. They just let me sleep through the night.' Luna: That's a good line. I wonder if we'll see that kind of IoT adoption in other fermented products — beer, cheese, even bread. Lucas: We already are. There are breweries using the exact same LoRaWAN sensors for fermentation control. And some artisanal cheese makers are monitoring humidity and temperature in aging rooms with IoT. The principles are the same — any biological process that needs tight environmental control is a candidate. Luna: So what's the one takeaway for someone listening who works in a different industry? What can they learn from Stonebridge? Lucas: I'd say: find the single point of catastrophic failure in your process. For Stonebridge, it was an undetected cooling valve failure. The IoT solution didn't try to digitize everything at once — it focused on that one vulnerability. Once they proved it worked, they expanded. That's the playbook. Luna: Smart. Start small, solve a real pain point, and let the data speak for itself. Lucas: Exactly. And in the meantime, the wine is better. I'll drink to that.