Hiroaki Asakura, a farmer in Aichi Prefecture, central Japan, learned how to be a green thumb from his father.

This meant knowing the seasons for planting crops and what pesticides to apply and when. But climate change has made reliance on instinct a dangerous game.

To deal with today's reality, Japanese firms have begun supplying AI-based agricultural pest diagnosis solutions to farmers like Asakura via smartphone apps, in a bid to boost crop yields and reduce the amounts of pesticides used in cultivation.

At a time when the agricultural industry is facing increasing labor shortages because of fewer farmers amid Japan's aging population, companies like Mirai Vegetable Garden based in Nagoya, western Japan, are making life easier with forecasts for pest solutions by machine learning on big data.

Others such as Nihon Nohyaku Co. and NTT Data CCS Corp. -- agriculture and system integrator companies respectively -- have jointly developed an AI app to identify types of crop diseases and suggest the appropriate pesticide controls.

The timely application of pesticides based on outbreak forecasts is expected to help farmers who might find themselves at a loss to make good decisions when the temperature drops later than in the average year, for example, due to the effects of global warming.

Mirai's app, which learns from past records covering about 1 million pest outbreaks and weather data, came in handy for Asakura, 50, on his farm in the city of Toyohashi, when he decided to "spray pesticides earlier than usual" on his broccoli field at the start of November. It indicated the likelihood of black rot increasing after three days if he waited.

Black rot outbreaks, which produce brown discoloration and decay in the leaves of fruits and vegetables, usually occur in hot and humid conditions. Rain was forecast after the next three days, followed by warmer temperatures.

Although Asakura learned the ins and outs of farming from his father, there have been times when he has been late spraying pesticides and failed to prevent diseases from occurring.

"It's too late if you spray after an outbreak occurs. Lately, the temperatures have risen higher than before, so I can't cope anymore simply based on experience and intuition," he said.

The app supports a total of 10 different crops, including broccoli, strawberries, tomatoes, squash and radish. It also includes a function that allows farmers who have a pest outbreak to share information via the app so that neighboring farmers are forewarned.

Mirai began offering its app in October. JA Toyohashi, a Japan Agricultural Cooperatives member, and JA Zen-Noh Gunma, which is part of the National Federation of Agriculture Co-operative Associations, are among the local farmers organizations that have introduced it on a trial basis.

Mirai Vegetable Garden President Tomofumi Hatakeyama referred to one case of vegetables being damaged by climate factors. According to him, there has been an outbreak of a disease normally seen in warmer regions occurring in Tsumagoi village in eastern Japan's Gunma Prefecture -- known as a major cabbage-growing area with relatively low temperatures suitable for production of the vegetable.

"AI can help us grow crops now in response to climate change," Hatakeyama stressed.

In 2019, Nihon Nohyaku and NTT Data CCS began offering an AI app that identifies pests and weed species from photos taken with a smartphone. It can diagnose more than 1,100 pests and weeds in a total of 25 crops, including rice and cucumbers.

Available for free, the app has been downloaded some 190,000 times. "It allows experienced farmers and agricultural advisors to share their knowledge," an NTT Data CCS official said.

A survey by the Ministry of Agriculture, Forestry and Fisheries has found that the spring of 2024 saw a large outbreak of fruit tree stink bugs, which cause damage to a wide range of fruits including apples, pears and peaches.

Stink bugs are normally seen in autumn in Japan, but climate factors apparently enabled many of them to survive the winter. This prompted municipal governments in at least 25 of Japan's 47 prefectures to issue advisories, far exceeding previous years' numbers.

Source - https://mainichi.jp