Enormous potential for savings: How AI is optimizing agriculture

In complex situations, artificial intelligence proves to be surprisingly helpful time and again. Using an AI system specifically designed for agricultural production, a research group from various Chinese universities has succeeded in drastically reducing problematic end products from nitrogen fertilization.

These are not exactly in short supply. For example, around 120 million tons of nitrogen are released into the water as nitrate or into the atmosphere as ammonia. Neither is really good.

One causes high costs in the treatment of drinking water. The other increases the greenhouse effect. The conditions for the formation of the harmful end products vary, depending on the weather, the season, the soil or the crop rotation.

It is almost impossible to estimate exactly when and in what exact quantities fertilization has the greatest effect. This is where machine learning comes into play. Fed with countless data from farms, the use of fertilizers could be significantly improved.

For example, the production of ammonia alone could be reduced by more than one third. Instead of the more than 4 million tons per year, the quantity could be limited to less than 3 million tons.

While emissions from wheat cultivation could be reduced by up to 28 percent, emissions from rice could be reduced by as much as 56 percent. Mind you, without sacrificing yield.

This would not only have a positive impact on environmental protection. It would also save considerable money on the use of fertilizers. Not to mention the energy saved: The production of nitrogen fertilizer is said to be responsible for more than 1 percent of global energy consumption.

Then all the AI has to do is not consume more than the saved production when calculating the optimal use of fertilizer and it works out for everyone.