Home Correction app This 17-year-old created an AI-powered app to help farmers around the world

This 17-year-old created an AI-powered app to help farmers around the world


A few years ago, I saw a clip of the brilliant astrophysicist, Neil deGrasse Tyson, explaining how people could tell global warming was actually happening without reading a single academic study or sorting through difficult scientific explanations.

“You can take inspiration,” Dr. Tyson said, “from the rest of the animal and plant kingdom… Nature already has the answer to what everyone here is arguing about.

As the climate changes, plants and animals are forced to move uphill and north. Stresses on animals and plants increase as temperatures rise, soils become drier and other climatological effects occur. As a particular ecosystem becomes less habitable for native species, it can often become much more habitable for invasive species.

The problem of invasive species is particularly critical for farmers.

Mr. Anniyan is a rice farmer in Kuttanad, a district of Kerala, India. Rice is a difficult and resource-intensive crop, requiring precise timing of paddy flooding and careful weeding to ensure rice plants get the necessary nutrients early in their development cycle.

Kuttanad lies below sea level, which means that non-native species and flood water often settle in Kerala’s beloved ‘rice bowl’ region.

Several years ago, after a long period of heavy rain, Mr. Anniyan noticed a new plant growing in one of his paddy fields that looked like a frilly water lily. He had no way of knowing what the lily was, but he dutifully pulled it out and threw it away. The next morning, as he walked to his paddy field, he was surprised to find that more of the paddy field was now covered by the pernicious weed.

After weeks of a fruitless battle against the species, giant salvinia (Salvinia molesta), an invasive water fern native to South America, Mr. Anniyan ended up losing his entire rice crop. The fronds of the giant Salvinia illuminated the immature rice plants and smothered the exchange of gases between the atmosphere and the surface of the water.

Mr. Anniyan’s grandson, a thirteen-year-old boy named Nathan Elias, who grew up in suburban Austin, Texas, witnessed the tragedy during a summer visit to his grandfather’s rice paddies. father and decided to do something to help.

“Helping” for Nathan meant using machine learning algorithms to create a software application called invasive AI which allows a farmer to upload a smartphone photograph of an unknown plant, insect or animal, and uses artificial intelligence to identify the species.

Each photograph is geotagged with the coordinates of the site, so that InvasiveAI not only identifies potentially invasive species, but crosses the observation with those of users from other areas. By collecting data from a network of users, Nathan’s InvasiveAI is further able to make predictions – using data regarding local weather patterns and climate trends – of the likely current range of the invasive and its future rate of spread.

Armed with this information, local farmers are made aware of the likelihood of future infestations so they can prepare to combat invasive species before they suffer too much economic damage.

Nathan has worked on InvasiveAI throughout high school and has just started his senior year at Academy of Liberal Arts and Sciences (LASA), a magnet school in the Austin area. He told me he was now starting to think about college applications and gritted his teeth trying to prepare for the rigors of the admissions process.

If the person in charge of admissions at Stanford University School of Computing reads this please let me know as I believe I have found a very good candidate for your school.

InvasiveAI received a round of applause from Texas State Agriculturalist Ms. Lynn Seman, and several University of Texas agronomy professors called it “hugely impressive.” InvasiveAI is now used by the Texas Master Naturalist Program, a citizen scientist organization associated with Texas A&M University, in addition to being deployed to a user base of more than 6,000 agricultural workers. The AI ​​models used in InvasiveAI have been used to detect and project the spread of over 10,000 unique cases of invasive species growth.

There are a few similar apps available that use AI to identify various plant and animal species. However, these apps do not focus on invasive species or make predictions about range or spread.

There is also a program called EDDMapS which is associated with the Center for Invasive Species and Ecosystem Health at the University of Georgia, which focuses on invasive species. EDDMapS is also interesting, but it relies on manual recognition of invasive species rather than AI. A farmer or citizen scientist takes a photo and submits a report to EDDMapS and one of the 900 experts from the United States and Canada associated with the University of Georgia project analyzes the photo and renders a verdict on whether the species photographed is invasive or not. In other words, it’s basically a web-enabled manual reporting system, so it’s not as scalable as a system using artificial intelligence.

(By the way, if building a sophisticated AI-powered computer program to help farmers wasn’t enough for a high school student, Nathan also received a provisional patent for his work and protected the codebase.)

Nathan told me he’s looking forward to his help helping farmers like his grandfather avoid the economic and psychological stress of losing crops to invasive species and hopes to compete for a grant so he can buy cloud computing space for InvasiveAI rather than having to spend time building your own server to host the app.

Nathan’s goal is to continue to build the capabilities of InvasiveAI to create a platform that will enable farmers around the world to cohesively address invasive species, with the right support and assistance.

To this end, Nathan has started rolling out InvasiveAI to family farmers in Kerala, who now use the app to detect and visualize the spread of over 200 unique types of plants, animals, insects and pathogens. invasive in and around their rice fields.

Nathan’s grandfather explained that “…before, we [farmers] only worried about the rains. Now these rains are heavier, causing frequent flooding, and [they] bring invasive plants from higher regions to [Kuttanad]. With technologies like my grandson’s, we can start identifying and combating these species very early.

It’s easy to feel overwhelmed at times by the daunting challenges facing our civilization in this post-climatic world. Talking to this young engineer and entrepreneur gave me a small glimmer of hope.

Smart investors take note.


The original article incorrectly states that images submitted to EDDMapS are sent to University of Georgia agronomists. The Center for Invasive Species and Ecosystem Health at the University of Georgia has a network of more than 900 experts and regulators across the United States and Canada who monitor and verify the existence of invasive species at a certain location.

According to Chuck Bargeron, Director and Senior Public Service Associate at the Centre, “These experts and regulators conduct actual field visits to confirm the identity of the reported organism and decide on appropriate management actions. While the AI ​​is a great tool to help sort reports, it’s not a substitute for having boots in the field. »