Latest / Internet of Things with Fexingo: Connected Devices, Sensors, and Industrial IoT / How IoT Sensors Are Revolutionizing Agricultural Irrigation
Transcript
- Lucas: Agriculture consumes about 70 percent of the world's freshwater, and a huge chunk of that is wasted because farmers either overwater or underwater their crops. But a network of simple IoT sensors is starting to change that. Luna: We're talking soil sensors, right? Not satellites or drones. Lucas: Exactly. Ground-truth sensors buried at multiple depths — measuring soil moisture tension, temperature, and electrical conductivity. They're not flashy, but they're incredibly effective. A grower in California's Central Valley, a third-generation almond farmer, installed a system from a company called Tule Technologies. The sensors feed data every fifteen minutes into an irrigation scheduling algorithm that accounts for crop stage, weather forecast, and soil type. Luna: And the result? Lucas: He reduced water use by thirty percent in the first season without any yield loss. Over three years, he saved roughly 300 million gallons of water. That's enough to fill about 450 Olympic swimming pools. Luna: Three hundred million gallons — that's staggering. What's the sensor setup? Just a few probes? Lucas: For his 500-acre farm, he placed sensors in about ten representative zones — different soil types, different slopes. Each sensor node costs around $300, plus a gateway that sends data to the cloud. The whole system ran him about fifteen thousand dollars upfront. But his water bill dropped by forty thousand dollars that first year. So payback was under six months. Luna: That math works. But is this scalable beyond almonds? I imagine lettuce or corn have different needs. Lucas: Absolutely. The principle is the same — measure actual plant stress rather than guessing based on calendar days. But the thresholds differ. For almonds, you want to keep soil tension below about 40 centibars during kernel fill. For lettuce, it's much lower — around 15 centibars — because lettuce has shallow roots. So the algorithm has to be crop-specific. But the sensor hardware is mostly the same. Luna: What about the skeptics? I've talked to farmers who say they can 'feel' the soil or just look at the leaves. Lucas: That's the biggest adoption barrier — trust. Experienced irrigators have been doing this for decades, and they're often right. But they're not always right. A study from UC Davis found that even skilled irrigators overwater by 15 to 25 percent on average, because they irrigate a day or two late, then compensate. Sensors remove that guesswork. Luna: So it's not replacing their expertise — it's augmenting it. Lucas: Exactly. The best adopters are the ones who treat the sensor data as a second opinion. One farmer I read about said he initially ignored the sensors when they told him his field was still wet. Then he dug a hole and felt the soil — and the sensor was right. After that, he trusted the system. Luna: What about the broader picture? Does this scale to something like the Colorado River basin? Lucas: It's already happening. The USDA's Natural Resources Conservation Service has cost-share programs that cover up to 75 percent of sensor installation for farmers in drought-prone areas. Some water districts in California are even requiring sensor-based irrigation scheduling for farms above a certain size. The technology is proven. The challenge now is mostly behavioral. Luna: And data integration. I imagine you need to pull weather forecasts, evapotranspiration rates, maybe satellite imagery too. Lucas: Right. The state of the art systems combine all of that. Tule's platform, for example, ingests local weather station data and runs a crop coefficient model. Some growers also layer on satellite-derived NDVI — that's the Normalized Difference Vegetation Index — to spot areas of stress before they're visible to the eye. But the satellite data is lower resolution and less frequent. The ground sensors give you the real-time truth. Luna: So you get the best of both worlds: the big picture from orbit and the granular detail from a few inches under the dirt. Lucas: Exactly. And the data is starting to be used beyond just irrigation. Some farmers are using the same sensor networks to track nitrogen levels and optimize fertilizer timing. That has huge implications for reducing runoff into waterways. Luna: One concern I hear from smaller farmers is the upfront cost. Even with subsidies, fifteen grand is a lot for a 200-acre operation. Lucas: That's real. But the price of sensors has been dropping fast. Five years ago, a single node might cost $500. Now you can get a decent capacitive soil moisture sensor for under $100. And there are companies like CropX that offer a subscription model — you pay a few thousand a year and they handle all the hardware and analytics. That changes the math for smaller operations. Luna: The subscription model makes a lot of sense for technology that needs constant updates and support. Lucas: It does. And it lowers the barrier to entry. The farmer doesn't have to become a sensor expert. They just look at a dashboard that says 'irrigate Zone A today, wait on Zone B.' That simplicity is key. Luna: I'm curious about the environmental impact beyond water savings. If we use less water, that means less energy for pumping, less runoff, less soil salinization. Lucas: Right. In California, groundwater pumping accounts for about six percent of the state's electricity use. Reducing pumping by thirty percent saves a lot of energy and carbon emissions. And when you apply water more precisely, you reduce leaching of fertilizers into groundwater. It's a triple win. Luna: You mentioned behavioral barriers earlier. What's the biggest one you see? Lucas: It's the 'I've been doing this for 40 years' mindset. But there's also a data overload problem. Some farmers get the sensors and then don't know how to interpret the data. They see a graph of soil moisture over time and it's just noise. That's why the algorithm and the recommendation layer are so important. The system has to tell you what to do, not just show you numbers. Luna: That reminds me of the early days of precision ag in general — the data was there but the insights weren't. Lucas: Exactly. And the companies that are winning are the ones that make the decision obvious. The farmer shouldn't need a data science degree to save water. Luna: Honestly, if this episode gave you a new perspective on how technology can solve real-world problems, that's exactly the kind of conversation we try to have here at Fexingo. And if that was worth the price of a coffee, you can find us at buy me a coffee dot com slash fexingo. It's a small way to keep the show going without ads. Lucas: Completely agree. The listener support really does make a difference. And it keeps us independent, so we can explore topics like this without any outside influence. Luna: And now back to the sensors. Lucas, what's next for this technology? Are we going to see ai driven irrigation that predicts plant needs days in advance? Lucas: That's already happening in some research trials. At the University of Nebraska, they're using machine learning models that combine soil sensor data, weather forecasts, and even drone imagery to recommend irrigation schedules three days ahead. Early results show another 10 percent water savings on top of what real-time sensors achieve. The AI essentially learns the field's specific response patterns. Luna: So the sensors are the foundation, and AI is the layer that optimizes. Lucas: Exactly. And the sensors themselves are getting smarter. Some newer models measure not just moisture but also the exact nutrient content in the soil solution. That could let farmers fertigate — fertilize through irrigation — with precision down to the individual plant. It's still early, but the trajectory is clear. Luna: What about connectivity? These sensors need to transmit data reliably from a remote field. Lucas: That's been a challenge. Wi-Fi doesn't reach far, and cellular can be spotty. But the rollout of low-power wide-area networks like LoRaWAN has been a game-changer. A single gateway can cover several miles, and the sensors run for years on a single battery. Some systems even use solar-powered nodes. Connectivity is becoming less of a bottleneck every year. Luna: So we have the hardware, the connectivity, the algorithms. What's the one thing that would accelerate adoption the most? Lucas: I think it's proof at scale. When a farmer's neighbor saves 40 percent on their water bill and gets the same yield, that's more persuasive than any scientific paper. We're starting to see those success stories multiply. I suspect within five years, sensor-based irrigation will be the norm in water-stressed regions, not the exception. Luna: It's one of those rare technologies that pays for itself while also being good for the planet. Lucas: Exactly. It's not a trade-off. It's a smarter way to do something we've been doing for 10,000 years.