*The Burn-In Contributor Nathan Orr contacted AgriTech startup ConserWater for an opportunity to correspond directly with CEO Aadith Moorthy about his company and his vision of a more water-conscious agriculture industry. Disclaimer: The Burn-In is not endorsed by ConserWater and is not paid to promote their material or products.
Salt water may cover roughly 70% of our world’s surface, but fresh water is exponentially more finite, with less than 1% of Earth’s total fresh water readily accessible. And as humanity continues to consume water at increasingly high rates, we’re now facing freshwater shortages across the globe.
In some locations, the adverse effects of water scarcity have already become a painful reality. One such example is Mexico city, where fresh water is estimated to run dry within the next 30 years due to inadequate supply infrastructure and a rapidly growing population. Meanwhile, in Cape Town, South Africa, access to running water is expected to halt as soon as 2019.
Recent studies show that global water usage has been overly wasteful for decades, yet many of us are still oblivious of how dire the situation has become. One leading contributor to water scarcity and waste has been the agricultural industry, where both issues of low access and overuse have been commonplace for decades.
Artificially Intelligent Watering
ConserWater’s website greets users with an explanation that 2/3rds of the world’s fresh water is used within the agriculture industry, which is why droughts have such disastrous impacts on farmers. But with the emergence of improved agricultural technologies, it no longer has to be this way.
As a Caltech graduate with a background in Artificial Intelligence, in October 2017 Moorthy launched ConserWater’s online dashboard—a cost-effective A.I. monitoring system that tracks water distribution data in the field to fine-tune watering processes and achieve optimal soil moisture for crop growth, while at the same time minimizing the consumption of excess water.
The software uses a combination of dynamic information, including satellite and historical weather data, alongside A.I. learning to best predict watering needs. This technology is also scalable in cost and size, encouraging farmers of all kinds to reduce waste and bring watering operations closer to peak efficiency.
ConserWater Motives and Goals
When asked about the inspiration behind ConserWater, Moorthy recalled a story.
“A couple of years ago I was traveling through a small village in India, just passing by, and we overheard that a farmer there had recently committed suicide. I inquired around further and learned that the reason he took his own life was because he couldn’t grow sufficient crops. That really got to me.”
Recent statistics show that in the U.S. alone, the suicide rate of agriculture workers are five times higher than the national average, and double that of military veterans.
Moorthy went on to explain, “I started thinking more about how, as human beings, we all need food to survive, but the farmer who grows our food needs water to make that happen first. The thing is, what many of these farmers are most passionate about isn’t just working, it’s being able to provide. But due to climate change, droughts are becoming more common, and the farms have been the first to suffer.
“I wanted to see if I could apply my technical skills to take on this problem, so now we use A.I. to help farmers maximize their plant productivity. Growing more with fewer resources.”
It’s easy to see how the more accurate water usage is, the better farms perform. Especially when considering both the cost of water and the energy used to transport it.
“The most direct cost after water is electricity. You have to pump the water somehow, right? And that could be water from a simple canal, or getting groundwater from deep underground. Nowadays, in the central valley of California, it’s not uncommon to see wells that go down 1,500 ft. You can do the math and realize people are spending hundreds of dollars for every acre they water just in terms of energy use.
“After that, there’s the more esoteric waste. For example, it’s really common to see leaks in drip irrigation. It uses small plastic pipes, and it’s very common for those to break. But ConserWater’s A.I. learns to identify and point out these damaged areas, rather than making farmers manually inspect their pipelines.”
Water leaks are a significant contributor to unnecessary water waste worldwide. For reference, billions of liters of water are lost daily in England alone due to faulty and neglected industrial piping.
The Changing State of Farm Technology
When considering the advantages ConserWater presents for the agricultural industry, it’s a wonder why no technologies like it have come out before.
From Moorthy’s perspective, it all comes down to broader accessibility of information. “We believe a lot of the data needed to make ConserWater happen hasn’t been available beforehand. Everything about A.I. comes down to good data.”
With the continued growth of IoT technology, data collection and machine learning methods have accelerated like never before. But while systems are gathering exponentially more data, applying it is a whole different story.
“Of course, even when you have data it’s hard to get actionable insights from it. Our whole idea was we had to think outside the box. Why I say that is, beforehand, none of the typical machine learning techniques out there worked with our data sets. You mostly hear about data usage being applied in image recognition, automatically finding cats in pictures and things like that, but that sort of thing is way different than what’s needed for agriculture.
“We had to create a new approach to machine learning algorithms in the context of agricultural data to get usable information from it. What we’re dealing with isn’t so much about the classification of data, our problem is more about telling you how many resources you need right now, which is extremely different from finding a cat on a screen.”
Now that the technological hurdles have been overcome, some remaining barriers for implementation of this type of A.I. empowered agriculture relate to age and region. “We’ve found older farmers are sometimes more hesitant to adopt new technology, but younger generations are becoming increasingly open to innovation––especially when it promotes sustainability.”
ConserWater A.I. in Action
ConserWater’s sensor-free use of A.I. makes it unique from typically physical IoT sensors. Instead, it relies on satellite observations to inform farmers––including their watering systems––on where water is best allocated.
“Let me give you an analogy for how one of our satellites works. To start, think of how a microwave heats food. It uses microwaves, as the name suggests, which excite water molecules in your food, so they start vibrating at the molecular level. You feel these vibrations on the outside as heat. Eventually, this warms the rest of the food.
“In a similar manner, water molecules throughout the earth are emitting their own waves all the time at a particular frequency. A satellite can detect these waves in the ground, and from them determine soil moisture.”
From the farmer’s end, the only equipment needed is a desktop or smartphone to access their data. They can also receive text or email notifications to stay current with the status of their operations. Additionally, ConserWater has partnered with multiple irrigation controller companies to manufacture optional IoT-powered units that apply ConserWater API in the field to completely automate watering processes. True to the nature of artificial intelligence, the system will then autonomously apply the data results from those watering methods to study and improve upon itself.
“ConserWater uses reinforcement learning and is always getting better with time. As farmers water locations, we continuously gather data that then teaches it further. For example, imagine a farm where there’s a very slight incline which is small enough for the farmer not to notice. That slight incline would have a notable effect on how the water there is absorbed into the soil.
“With our system, after you water, it uses its satellites to see that the area wasn’t sufficiently watered for some reason. It will then tell itself to do something like increase the watering there by 10% and see what happens next time. It studies those results to measure different solutions, working to optimize while minimizing resources.”
Farming into the Future
Commenting on what’s next for ConserWater, Moorthy explained that their current objective is to “continue developing partnerships with firms and agricultural controller developers to make this technology accessible to farmers at the grassroots level all around the world.”
Looking ahead at five years from now, the company hopes to continue to expand their reach beyond developing watering systems alone.
“We want to eventually be able to provide an end-to-end farm system, from start to finish. From telling when the most optimal time to plant seeds are, to what conditions are best for growing those seeds, to the best time for harvesting, to how much yield there will be. Essentially, we want ConserWater to be an all-encompassing service for everything the farmer needs.”
Water conservation efforts are rising to the forefront of global attention, and companies like ConserWater are shaping their future. What are your thoughts on the recent innovations in farming technology, and predictions for the developments still heading our way?