Subscribe
Jagdev Singh, a cotton farmer, has faced three consecutive years of hardship. Since 2021, severe pink bollworm (PBW) infestations have significantly impacted his income from his 0.6-hectare cotton field. However, this year brings a glimmer of hope. Singh is one of the 18 farmers chosen for a groundbreaking pilot project by the Central Institute for Cotton Research (CICR) under the Indian Council of Agricultural Research (ICAR).
This project utilises cutting-edge artificial intelligence (AI) for real-time pest monitoring. By deploying AI-powered pheromone traps, the CICR aims to empower farmers like Singh with the information needed to make timely pest management decisions. This initiative marks a first in India, pioneering the use of AI for pest control in any crop.
The pilot programme focuses on three major cotton-growing districts in Punjab: Muktsar, Bhatinda and Mansa. A total of 18 fields, six from each district, have been selected for the project. The innovative approach has the potential to significantly reduce economic losses for farmers grappling with PBW infestations.
This project will be implemented in Rajasthan and Haryana after the pilot in Punjab is successful.
PBW (Pectinophora gossypiella) destroys parts of the developing cotton fruit, such as the square (flower bud) and the boll (rounded sac of seeds with cotton fibres). It has posed a serious threat to the cotton crop and farmers’ livelihoods.
In the last three years, the cotton crop across north Indian states, Punjab, Haryana and Rajasthan, reported a severe PBW attack and even the genetically modified pest-resistant variety called Bt Cotton (Bollgard II seed) has been falling prey to the pest it was created to resist.
On average, it causes at least 20-25 per cent yield loss. A recent series by Down To Earth (DTE) detailed PBW’s cotton curse in India’s North Zone.
Pheromone traps, which contain the pheromone gossyplure, a chemical emitted by female moths to attract male moths, have long been used to reduce crop pests. Farmers typically install five pheromone traps per hectare and are expected to monitor the activity of male moths to determine whether it exceeds the economic threshold level, or ETL.
ETL is the insect population level or the extent of crop damage at which the crop value exceeds the cost of pest control.
Farmers, on the other hand, find it difficult to conduct this monitoring on a regular basis, especially when they also have other Kharif crops to cultivate simultaneously.
Take Singh as an example. “Using conventional traps, I never learned the exact timing of the attacks, which hour of the day the crop is most vulnerable to PBW attack, or the number of pests. It was difficult to manually monitor the crop every day. For example, suppose we go to the field and check the crop every three days and do not find any infestation and then we do not go for another two days, thinking the crop is fine, but when we return on the sixth day, we discover that the majority of the crop has vanished,” the farmer said.
In the new digital approach, he receives hourly crop updates via his mobile phone. A camera fixed in the trap takes regular pictures of the moths that stick to the trap due to the pheromones’ lure. These images are then transmitted in real time to a remote server in the cloud.
The insects’ images are analysed using a machine learning algorithm that has been trained to identify and count PBWs caught in traps. Anybody with a username and password, from ICAR scientists to farmers and agriculture extension officers, will be able to access pest alert information in real time on their mobile phones.
“We previously observed that farmers rarely count and monitor pests caught in traps. The process is time consuming and farmers also have other crops to cultivate. The new system enables timely pest management advice to cotton growers, ensuring efficient control and keeping damage below economic threshold levels,” YG Prasad, director at CICR told DTE.
Crop conditions are assessed based on scientific data collection and analysis and farmers are issued advisories.
Singh has been remotely monitoring the crop and has not visited the field in the last three days.
“Three days ago, I got to know in the evening that there was an attack of PBW on my crop and I immediately visited the field and I could save my crop,” he said.
This data will also be shared with agriculture state department officials, taluka officials and village level extension functionaries, who will use it to create mass campaigns and distribute advisories to their farmer database, according to Prasad.