· AI is future of agriculture due to aging population, says top grower
· Grower says traditional methods still beat AI on return on investment for now
Shen Haiyan hasn’t made up her mind whether artificial intelligence is a friend or foe.
Shen, 47, is one of the top strawberry farmers in China and sells millions of seedlings a year to farms. She is in Kunming with her all-female team to participate in a competition where traditional growers will challenge data scientists to see who can come up with the most cost-efficient techniques to improve agricultural yield.
“I still remember the first time I saw strawberries grow from my lots,” said Shen, nicknamed “Strawberry Queen” for her exploits. “I felt so happy that I couldn’t help bursting into laughter.”
It is a good business. Industry sales of strawberries have increased through the years as rising incomes prompted more families to seek a healthier diet. But like all of agriculture, it is also a capricious business, beholden to the vagaries of nature. Shen has weathered three major natural disasters — drought in 2012, floods in 2013, and snowstorms in 2018.
But Shen is feeling confident and good about herself on this recent summer day in Kunming. Her 24-year-old daughter is a member of her team taking part in the Smart Agriculture Competition organized by e-commerce company Pinduoduo and China Agricultural University.
The competition will run through October, with four teams of young scientists from top universities pitting their skills against four groups of traditional growers. The judges will consider the economic value created and the reliability, scalability, and technical merit of the solutions when choosing the winners. The winning team will receive research funding, academic and commercial support, and have their technology implemented at farms in Yunnan with the help of Pinduoduo.
“Improving agriculture in China requires a comprehensive approach, from increasing market access, to streamlining the supply chain, to raising productivity,” said Xin Yi Lim, Executive Director for Sustainability and Agricultural Impact at Pinduoduo. “Our goal in organizing this competition is to get the best brains in technology and agriculture to develop practical and commercially viable solutions for the industry.”
Winning the competition will require keeping costs low while boosting yields. In this regard, Shen said the traditional growers still maintain a cost advantage at this point. She said that being on the ground allows them to detect and control pest infestation and diseases much quicker than the remote monitoring by the AI teams.
Her scientist counterparts politely disagree. Lin Sen, leader of one of the AI teams, said their experience in setting up fully automated greenhouses stands them in good stead. Lin said his team would employ plant-growth models and machine-learning algorithms to develop the most suitable solutions for growing strawberries. “I don’t think we will do worse than the farmer teams,” he said.
The biggest winner, though, is likely to be agriculture. AI is already increasingly used around the world in farming. Examples include improving harvest quality, monitoring weather and soil conditions, detecting and controlling diseases and pests, and creating forecasting models to increase productivity.
That could help solve some of the entrenched problems plaguing agriculture in China. China must support more than 20% of the world’s population, with about 7–9% of the world’s arable land. But nearly 98% of farmers in China cultivate farms smaller than 2 hectares, making it difficult to standardize growing processes and achieve economies of scale. The rural workforce is also dwindling and aging, as youths head for the cities to find more stable and higher-paying jobs.
At the Smart Agriculture Competition in Kunming, the interaction of scientists and growers is already reaping benefits.
Shen’s daughter, Sun Yuqing, is due to pursue graduate studies in economics at Northeastern University and is determined to take this opportunity to learn from the AI scientists. She hopes to combine AI with traditional agricultural techniques in the future.
“I hope our company will become the first to apply AI in strawberry growing,” said Sun, who has a degree in plant protection from Anhui Agriculture University and did an exchange at Colorado State University. “Sometimes, I go by their greenhouses and see they do save a lot of human labor by using sensors and automation.”
For Shen, she hopes AI will be powerful enough to free humans from the backbreaking farm work by the time her daughter takes over the family business. “Maybe in 30 years, after our generation born in the 1960s and 1970s have retired,” she said.
“We rely on experience, and our way of data collection relies on estimation, so it is imprecise,” Shen said. “Their data-driven approach could give us some guidance on how we can improve.”