OpenIDEO is an open innovation platform. Join our global community to solve big challenges for social good. Sign Up / Login or Learn more

ConserWater: Artificial Intelligence-based Precision Irrigation for Every Farmer

Conserwater is the world's first Artificial Intelligence that predicts exactly how much water farmers need to give crops anywhere globally.

Photo of Aadith

Written by

How does your idea apply new technologies to make agriculture and water systems more resilient in the face of climate threats?

ConserWater uses patent pending methods to combine data ranging from NASA satellite data to weather, with geospatial deep learning to predict irrigation water requirements. It can increase any farm's water efficiency up to 30%. Agricultural regions use nearly 2/3 of the world's fresh water. Non wonder when drought strikes, it is the farmer who suffers the most. With global warming, the prevalence of droughts has increased, and ConserWater offers a cheap and effective solution for adaptation.

What fraction of the world’s fresh water is used for agriculture? The answer is 2/3. No wonder, agricultural regions of the world are often the worst affected by droughts. When drought strikes, only rich farmers can afford expensive water saving technology such as soil moisture sensor array, while the average farmer is left to suffer! This no longer has to be the case in the 21st century. Introducing ConserWater: the world’s first Artificial Intelligence that predicts exactly how much water farmers need to give to their crops at any location around the world at any time.

ConserWater uses patent pending technology to synthesize a large variety of data from satellite imagery to historical weather to come up with accurate irrigation predictions. We also have pilot projects underway throughout the world to understand exactly what our diverse array of customers, ranging from subsistence farmers to utility level users, wish to see in this product. In the near future, you will no longer see farmers mindlessly flood their fields with water: Rather, they will start using Artificial Intelligence to engage in precision irrigation.

See conserwater.com today!

What early, lightweight experiment might you try out in your own community to find out if the idea will meet your expectations?

We have been running pilot programs primarily in Ghana and India, to assess what the small farmers of the world would wish to see. These have helped us refine the product, and add new features that were requested by these users. The product was well received and enabled these farmers to reduce their water use. In our own community, we are in the process of reaching out to local farmers and utilities to evaluate ConserWater for their large scale use, as a predictive tool for large regions.

We are excited to learn about ideas at diverse stages of development. At what stage would you currently classify your idea as? (multiple choice)

  • Full Scale Roll-Out: I have completed a pilot and analyzed the results. I am ready to expand the pilot significantly.

How do you plan to grow or scale this idea?

As for the technological scaling, several countries worldwide have already been machine learned, and are supported by the ConserWater AI. Our initial market segment will be the customers of our licensing partnerships (farm management and sprinkler companies) and the enterprise web app (agricultural corporation and utilities). This will allow us to generate revenue which we can invest in the long term to reach individual small farmers worldwide through global partnerships with local NGOs.

What skills, input or guidance from the OpenIDEO community would be most helpful in building out or refining your idea?

Guidance on marketing strategies would be very useful at this point in our journey. Our main focus now is to bring our products to large scale users, so that we can validate our business model and plan.

This idea emerged from:

It was November 2015. I (Aadith Moorthy, Founder) was passing through some Indian villages, and I saw a funeral procession. A farmer had committed suicide. I asked a villager what happened and he replied “The monsoon failed. We cannot grow crops. We are suffering”. It was then that it occurred to me that water security was the secret to improving farmers lives! I then started thinking how I could use my technical skills to improve their lives and came up with the initial idea for ConserWater!

Tell us about your work experience:

I (Aadith Moorthy, founder) have significant experience in Artificial Intelligence and Machine Learning from my California Institute of Technology (Caltech) Computer Science background and several previous projects in Machine Learning, from protein modeling with deep learning, to game play with reinforcement. Our web developer, has been making web apps since he was 10. Our COO is a 30-year industry veteran, especially in machine learning, and running startups. Finally, our mentor, Mr. Vaibhav Lodha, has been a serial entrepreneur in the water and AgTech space for over a decade, bringing significant experience on working with farmers and governmental organizations.

Attachments (1)

IJOEAR-SEP-2016-34-2.pdf

A scientific journal publication of the merits of the ConserWater Artificial Intelligence.

15 comments

Join the conversation:

Comment
Photo of namrata
Team

Hi Aadith,

I like the technology and intention it has. I am curious to know about the problems you are facing while determination of the technology for the small farmers. Can you list 5 major problem you faced or still are. I can share you my email id if you want to discuss it further.

Regards

Photo of John Hall
Team

Asdith,
I like the machine learning approach. I will forward your information to our local farmers in Pennsylvania. I would like to stay in touch with you and learn more about your work.
John

Photo of Nick
Team

Well done Aadith, India and Ghana are some of the most popular emerging markets from an economic perspective.

Photo of Nick
Team

Have you considered the availability of access to mobile devices as well as a consistent internet connection? I would imagine most individuals who would benefit most from this concept are limited in their resources.

Photo of Aadith
Team

Dear Nick,
Thank you for your excellent question. As it stands now, the teledensity in several developing countries is already very close to 100%, so this is not as severe an issue as many of us living in developed countries think. Furthermore, it is only part of long term vision to reach individual small farmers worldwide through global NGO partnerships. We believe, from estimates, that almost every small farmer will have a smartphone within 2-3 years from now, following the current trends of adoption. And this is the time frame we have for our long term vision. Our short term plan is to focus on enterprise level users, who are much easier to reach than individual small farmers who constitute a very fragmented market globally.

Photo of Catherine Griffin
Team

Aadith, congratulations on being one of our featured contributions! Please consider applying for the Climate Ventures 2.0 program: https://www.f6s.com/applyclimateventures2.0/apply

Photo of Aadith
Team

Hi Catherine, Thanks for your comment.
In fact, I only found out about openideo because I wanted to apply to the Climate Ventures 2.0. I have already put in ConserWater Technologies' application, since the deadline used to be Mar 16

Photo of Mitul Sarkar
Team

I would assume that each farmer inputs the location of his/her farmland into the app, to get predictions?
What is the approximate resolution (surface area) of the AI's prediction to each farmer? Or is it based on postal code, etc? Thanks.

Photo of Aadith
Team

Currently, to make it easy for small farmers to use, our baseline smartphone apps directly capture the farmer's location based on the GPS of the smartphone. This is usually pretty accurate around the world to an accuracy of a few tens of meters.
Our enterprise web app lets a user select the area of his choice on an interactive map.
The AI prediction itself is pretty localized to the location: we actually do a new prediction every time a farmer asks for one.
Please try our smartphone apps out at conserwater.com!

Photo of Mitul Sarkar
Team

Hi, Aadith, thanks for the answers.
My humble 2 cents:
1. For places where smartphones are still not common enough, your systems might be able to use text-to-speech generation, so even an illiterate farmer can get the information from the AI via a voice call when the farmer identifies his province, district, village, etc.

Photo of Aadith
Team

Yes, Thanks for your comment Mitul.
We are already working on getting speech recognition integrated into our services. Especially in parts of Africa, farmers may speak languages that have no written form, so this is a necessity!

Photo of Lee Trawick
Team

Aadith  are you at Caltech? I noticed NASA and Pasadena in your app and had a hunch you were at the top tech uni in the world. I was just visiting campus this weekend for Hacktech.

Photo of Aadith
Team

Yes good observation. However, I don't give you too much credit for it because I said that in the experience section!

Photo of Catherine Griffin
Team

Aadith, great post and congratulations on your success so far. Have you raised any capital yet for this project? Do you expect farmers and utilities to pay for this?

Photo of Aadith
Team

Hi Catherine, Thanks for your comment. We are completely bootstrapped right now: we want to wait till we get all our products out and start getting more feedback from our potential large customers.
As I mentioned in our business model and short term plan, we don't expect small farmers to pay for this: they can use our completely FREE ad supported baseline smartphone apps. For the utilities, our value proposition is very simple: using ConserWater you can minimize the amount of water wasted (up to 30%), by pinpointing how much water needs to be used for irrigation in each area. This may seem like a far-removed problem for us city-dwellers, but is actually a major issue in the world. For example, there were large scale riots in South India just as recent as a few months ago because the state governmental utility boards could not agree over how much water should be sent to each states farmers.
More broadly, this value can also be extended to our licensing partnerships, which are also paid: using ConserWater, a sprinkler company can create AI-powered sprinklers that are very water effective yet cheap, to bring to their end customers.