loader TNT
How AI is helping Indian Cotton Farmers reduce pesticide use

How AI is helping Indian Cotton Farmers reduce pesticide use

A
by Amelia Scott — 1 month ago in Artificial Intelligence 2 min. read
1044

You will find over 5.8 million cotton farmers in India in accordance with this nation’s Textile Ministry. Each yearthey face significant losses due to insects attacking their plants.

In 2017, farmers in the state of Maharashtra confronted a reduction of $15,000 crores ($2.1 billion) since 50 percent of their crop was under assault from insects.

Because of this, over 55 percent of pesticide in India goes towards farming. On the other hand, the incorrect use of those chemicals can harm the crop or decrease the quality.

That is why Wadhwani AI, a research institute in India, began hunting for solutions from 2018 to help farmers save their plants with insights from artificial intelligence.

Conceptualization and data collection

The notion of the AI version was to ascertain just how many pests are observed in that region and send an advisory on pesticide use. But, there was no previous field information to train the model. So the team had to create a particular app to accumulate information.

Farmers already use pheromone traps to capture pests and forecast if there may be a bigger assault.

The information collection program requested farmers to shoot photographs of pests captured in these traps on a white sheet of newspaper.

The first target was to recognize unique bollworms, which poses a significant threat to cotton plants.

Jerome White, the senior writer in Wadhwani Artificial Intelligence, told me the team invested the first few seasons observing and collecting that the data to enhance the model.

The group needed to be certain that the model properly identities the kind and amount of pests from the image to provide precise advice to the farmers.

The first data collection began in 2018 in Maharasthra, and a year ago, the group deployed an early variant of the model which has been trained and supported with over 28,000 pictures.
Also read: AI has become so Simple and Less Price that anyone can use it

Deployment

Since this program was used mostly on low-end telephones, researchers needed to compress the version out of 268MB into 5MB. They then utilized PyTorch Mobile to set up it to a program that worked offline.

The version currently assesses images delivered by farmers and based on rules determined by farming governments of India. It is currently deployed in many areas of three Indian nations, Gujarat, Maharashtra, and Telangana.

Next steps

At summer time experiment in Maharashtra, 150 farmers utilized the machine and detected 25% profit in the harvest.

Due to these outcomes, Maharashtra and Telangana state government has agreed to extend the project in the following cotton growing season (June-November).

Together with that, researchers are working together with all the Better Cotton Initiative (BCI), a international non-profit that appears after the enhancement of farmers, climbing this job internationally.

Amelia Scott

Amelia is a content manager of The Next Tech. She also includes the characteristics of her log in a fun way so readers will know what to expect from her work.

Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments

Copyright © 2018-2020 The Next Tech. All Rights Reserved.