In 2019, nearly 690 million people around the world went hungry. The pandemic could make that number grow by another 95 million people. But a series of new reports argues that it’s still possible to meet the U.N.’s Sustainable Development Goal of ending hunger by 2030—and lays out what’s needed to make that happen.

The amount of aid given for food security and nutrition each year will need to double, researchers say, and the poorest countries will also need to invest more, adding up to an extra $33 billion spent per year. But if that investment goes toward the right approaches, they say, the problem is solvable. The researchers, from a coalition of organizations called Ceres2030, used a machine learning tool to analyze more than half a million reports and studies and find the types of interventions that governments should prioritize and how much they would cost to implement. The analysis took three years.

“If we break it down into discrete problems, then yes, I think we absolutely do have a chance to achieve zero hunger by 2030,” says Jaron Porciello, a data scientist at Cornell University and primary investigator and co-director for Ceres2030. The researchers looked at key interventions for rural areas in the poorest countries, particularly sub-Saharan Africa, where people are most likely to be hungry. Then they wrote 10 research papers looking at different solutions.

One research team, for example, studied climate-resilient crops, such as maize that can survive droughts—a key technology when farmers are facing increasing threats from extreme weather. But the crops are only likely to really help when combined with other interventions. “If you don’t support farmers with opportunities to empower them through training, particularly through extension services, then those technologies are actually not going to be successful,” Porciello says. Another study looked at ways to prevent food waste before food is delivered to markets. A third study looked at how to train young people in new skills so that they can earn more money in the food system.

Donor governments can use the data to help decide which programs to support. Donors want “concrete, specific solutions that are actionable,” she says. “So I think we bring together really high-quality evidence, and then we bring together a cost model that puts these things together so that donors have more information about not only specific interventions, but then actually what those interventions might cost.” Each year, around $9 billion should be spent on farms, training farmers and improving crops and livestock feed. Another $2 billion should go to solutions to prevent food waste in the supply chain, and $3 billion should go to training rural youth and helping those living in extreme poverty get good jobs.

Read the rest of Adele Peters’ article here at Fast Company