“We're generating new crops five times faster.”

Like many food executives, Jeremy Bunch is concerned about the impact of climate change on his business.

“Weather and climate are probably our biggest risk,” said the president of US flour company Shepherd's Grain.

The Idaho-based company sources its wheat from farmers in the Pacific Northwest of the United States.

As weather patterns become more unpredictable, Bunch said, “you need to have a Plan B and a Plan C in case Plan A fails.”

To power these plans, Bunch's company is now using an AI-powered software system called ClimateAi.

ClimateAi aims to use current and historical data, such as satellite imagery, temperature and rainfall records, and combine it with future predictions to provide farmers with the most accurate, locally tailored weather forecasts possible, from one hour to six months ahead.

It then gives precise advice on when to plant and harvest specific crops and predicts yields.

Shepherd's Grain only started using ClimateAi last year, but already most of its 40-plus farmers have been mentored on the app.

“They're starting to consider using ClimateAi to help plan crop management decisions for wheat, the primary crop grown in the region,” Bunch says.

“By predicting the weather, growers can decide which crops to plant. The platform knows when to plant and when the crops will flower and start producing seeds.”

One of the biggest problems facing the seed industry is getting climate-resilient seeds to market faster and cheaper, says Himanshu Gupta, chief executive officer of San Francisco-based ClimateEye.

“By the time seed companies do this — 10 to 15 years from now — the environment has already changed,” Gupta said. “We're racing against time to launch new seed varieties.”

ClimateAi helps these companies see how specific test seeds perform in specific regions and provinces, he says, “so seed companies can understand the best places to grow the seeds.”

A study published in the journal Nature last year warned that if climate change causes large numbers of crops to fail simultaneously around the world, the consequences could be disastrous.

“Synchronized harvest failures across major crop-producing regions pose a threat to global food security,” said the report, led by climate scientist Kai Kornhuber of Columbia University's Lamont-Doherty Earth Observatory.

The warning comes as the world's population is expected to reach 10 billion by 2050, from the current 8 billion, according to the United Nations.

As the world's population continues to grow and pressures on crops increase, could AI be the key to developing new varieties that can better cope with extreme climates?

In the Tanzanian city of Arusha, David Gerena, an agricultural scientist at the International Center for Tropical Agriculture, is leading a project called “Artemis.”

Funded by the Bill & Melinda Gates Foundation, the project is using AI to help breed more resistant crops. Specifically, AI is helping speed up a task called phenotyping.

This involves observing characteristics such as how many flowers, pods and leaves a plant has, and visually studying new crop varieties.

“Traditionally, it takes about 10 years to develop a new crop variety,” Gerena explains, “but given the speed of climate change, this time frame is no longer viable.”

Gerena adds that phenotyping efforts have traditionally relied on human eyes, “but humans have not consistently done this with the precision needed to make subtle but important plant selections.”

“Temperatures in the field can exceed 30 degrees. It's just tiring, and fatigue affects the quality of the data.”

Instead, growers participating in the project use a smartphone app to take photos of their crops, which a trained AI can then instantly analyze, record and report on.

“A computer can tirelessly count the flowers and pods on every plant, every day,” Gerena says. “This is really important because the number of flowers on legumes correlates with the number of pods, which directly affects yield.”

“The data is so complex that it's hard to understand what's going on, but with AI we can make sense of that complex data, pick up patterns, show us where resources are needed and provide recommendations.

“Our plant breeders estimate that with better data from AI computer vision, they could potentially shorten breeding cycles to just a few years.”

In North Carolina, Avalo, an agricultural technology, or “agtech,” company, is working to create climate-resilient crops by using AI to study crop genetics.

“Our process starts with crop genomic data, including sequences from different varieties,” says Rebecca White, Avalo's chief operating officer.

“For example, different varieties of tomatoes have slight differences in their genomes that result in different traits, such as different flavors or pesticide resistance profiles. Our machine learning programs can capture these small differences in different varieties and tell us which genomes are important for which traits.”

Using their technology, they were able to produce broccoli that matured in a greenhouse in 37 days instead of the usual 45 to 60 days, White said.

“Broccoli produced on that timescale gets an additional growth cycle, saving carbon emissions and improving environmental impact.”

Avalo, which partners with companies in Asia and North America, is also working to improve frost tolerance in rice and drought tolerance in potatoes.

“Our core technology identifies the genetic basis of complex traits with minimal training and, through sequencing and predictive analytics, enables us to quickly and inexpensively evaluate and model new plant varieties,” White says.

“We are creating new varieties of diverse crops that can be developed five times faster and at a fraction of the cost of traditional breeding.”

But while AI can help mitigate the effects of climate-related weather and make crops more resilient, there are many challenges to using AI in agriculture, said Kate E. Jones, professor of ecology and biodiversity at University College London.

“The effectiveness of AI in ensuring food security also depends on addressing challenges such as data quality, accessibility of the technology, and more. We need to recognize that AI is one tool among many in a comprehensive strategy for sustainable and resilient agriculture.”