Skip to playerSkip to main content
  • 22 hours ago
A machine‑learning model learns how soybeans differ depending on where they’re grown – helping identify whether they come from illegally deforested land.

Category

🗞
News
Transcript
00:00Can this bean expose a crime? British scientists have found a way to trace where soybeans come from
00:07with surprising accuracy. Brazil, the US and Argentina are the world's biggest producers,
00:14but this new technology goes far beyond identifying the country. It can pinpoint a
00:20soybean's origin within 200 square kilometers. But why does that matter?
00:25A lot of the soy we import into the UK comes from illegally deforested areas. Soybean is the third
00:35largest commodity responsible for deforestation on the planet. New EU and international laws now
00:42require soybeans to come from deforestation-free areas. To help determine that, the scientists
00:49developed a machine learning model. It's trained in the unique characteristics of soybeans.
00:55Samples can differ quite a lot depending on where they're grown.
00:59And this is primarily due to influences from the soil and from the geology,
01:04and also from the atmosphere. So things like climate patterns, altitude and elevation,
01:11even distance from the coast. By comparing these traits,
01:14the model can estimate where a bean of unknown origin likely came from. In the future, this could
01:19help customs officials verify imports and expose illegal deforestation.
01:24That's it.
Comments

Recommended