IoT and Agriculture: A Natural Combination

IoT and Agriculture: A Natural Combination

Solid examples of IoT applicability are sometimes hard to articulate. But agriculture provides a use case that is quite concrete, even if in rural environments.

By Andrew Brust for Big on Data | December 16, 2016 -- 20:14 GMT (20:14 GMT) | Topic: Big Data Analytics

We all know that the Internet of Things (IoT) represents game-changing transformation in the industrial application of technology...or at least we think we do. Perhaps more accurately, we sense that denying IoT's significance would be foolishly contrarian. And while we might have an innate sense of why IoT is so important, we also might come up short when pressed to describe specific applications of IoT technology.

That's why I love learning about real use cases, especially ones that veer off the path of domains like preventive maintenance, which are almost cliché. Recently, I had the chance to learn of one such IoT application. It's rich, complex and it satisfyingly broadened how I think about IoT. Specifically, I spoke with the Daniel Koppel, Co-Founder and CEO of Israeli company Prospera, which focuses on the application of IoT in agriculture.

The use case
Prospera, a company founded about 2 years ago by a team of computer scientists and agronomists, has built some very interesting technology that centers around monitoring crop growth, in order to optimize it. While farmers have long had some data -- like weather readings and low-resolution satellite images -- available to them, it turns out not to be enough. And even if it were, weather data from a government weather station -- which might be 30km away from the actual growing area -- doesn't deliver the "hyper-local" climate data that is crucial.

When you grow in volume, though, the geographic dispersal of your farmland makes it difficult to go around and collect that data manually -- and the rural settings for that farmland make the electrical and network connectivity, that had been necessary to collect that data, hard to come by.

It's different now
But now low-cost sensors can obtain temperature and humidity data; and low-cost cameras can measure light/radiation and gather valuable images. The devices can communicate over WiFi or 3G mobile data technology and can often run on solar power. This approach has been making technology with great efficacy in indoor agriculture, increasingly applicable in outdoor settings too.

Prospera does not view itself as a sensor company though, but rather as a data company. And not just one that helps customers collect data and act on it, but one that builds data intelligence and thus domain expertise.

In other words, there is an element of crowd-sourcing here: although granular data is kept private, all data (which, in aggregate, amounts to hundreds of thousands of readings per day) benefits the construction, testing and accuracy of predictive models. These models help track the correlation between specific values in the collected data, crop growth and output. Understanding those correlations, and making predictions based upon them, is where Prospera hits its value proposition sweet spot.

The vision thing
Beyond predictive applications, there are prescriptive applications too. Computer vision/imaging has serious applicability in this domain, as the capture of images combined with pattern recognition technology can help detect crop disease and, on an automated basis, dispatch personnel to address it. It can also help alert farmers to where they need to prune and harvest. So not only is the data collection made more economical, but the methodical analysis of the collected data, and the dispatch of responsive action, is made more feasible and economical as well.

While expense was once an issue, Prospera's Koppel says that "sensors are commodities" now. In fact, the company says that three conditions in the market have combined to make its technology so effective and efficacious: the neural network technology behind the machine learning has become much better; the sensor hardware has become much cheaper and, because of greater mainstream appreciation for Big Data and machine learning, market readiness has crossed the chasm too.

Why Israel?
Israel is a high-tech country, well-populated with venture-funded tech startups. Top-notch technical universities like Technion and Hebrew University (Koppel's alma mater), as well as byproducts of, and veterans from, tech research in the country's defense forces, provide much of the raw material for such commercial, entrepreneurial activity.

Israel also has a history as an agricultural society, centered around Kibbutzim and Moshavs, both of which are collective agricultural settlements, the former sometimes likened to communes. Further, because much of Israel is desert, irrigation techniques and other technological optimizations have been part of its agricultural approach since the country's founding.

Put all of this together and you have a place that, industrially and culturally, is predisposed to growing its crops in a scientifically-influenced fashion.

How far beyond?
While kibbutzim and moshavs in Israel have served as test labs for Prospera's technology, the company has customers who have deployed the technology in Europe and Mexico. The US is next -- and for some customers there, Prospera expects to collect data not just from stationary sensors over terrestrial wireless data connections, but via drone and over Satellite data links as well.

Some current customers implement Prospera's technology in parts of their farms, to compare data-driven farming with those of more traditional methods, and results have been good. Prospera says that customers have used the company's technology to discover problems in irrigation, ward off disease and reduce yield volatility.

Prospera's technology has even allowed farmers to make course corrections in their growing techniques, in order to maximize output in the current growing cycle, and not just apply lessons learned to the next cycle.

All roads lead to analytics
As fascinating and as far-flung as Prospera's IoT use case may seem to some of us, it ultimately comes back to the mainstream of BI and Big Data: collecting data and analyzing it. In fact, Prospera's software delivers rather familiar-looking dashboards on computers and mobile devices, just like the technology with which we may be more familiar.

Ultimately, that may be the most valuable lesson of all. The stuff we already know -- the OLAP cubes, the MapReduce jobs, the streaming data processing and the D3 visualizations -- can be thought of and implemented in very specific and very impressive industrial use cases. The technologies needn't be relegated to isolated discussions of their own rigors. In fact, when we think of the technology in applied capacities, we provide a lot more value, and we help Big Data and IoT move past their hype cycle quasi-paralysis.

We need more companies like Prospera, that combine tech with domain expertise, cultural idiosyncrasies and a lot of imagination. That's how this field will get to the next level. The value of vision goes beyond data captured from image sensors.