Google on Thursday announced that Google One, its premium cloud storage service, will soon let you back up your device to the cloud from…
A mere two hundred years ago, our planet was home to less than one billion people. Today, the Earth’s population count stands at over seven billion. To further put this rapid growth into perspective, 6.5% of all people ever born are alive at this very moment.
With Earth’s rapidly increasing population, the growing need for food is becoming a serious concern. According to the World Bank, the human population will hit a staggering nine billion people by 2050.
That means that to keep sustaining human life on Earth, the world will need to produce fifty percent more food, which is a problem considering that, because of climate change, we are estimated to lose twenty five percent of our current crop yields before that time.
“The land, biodiversity, oceans, forests, and other forms of natural capital are being depleted at unprecedented rates. Unless we change how we grow our food and manage our natural capital, food security — especially for the world’s poorest — will be at risk,” it reads on the World Bank website.
Fortunately, there are many great minds at work that are seeking solutions to combat this problem.
Machine learning comes to the rescue
One such company, namely Descartes Labs, is currently using machine learning to examine various satellite images to predict food supplies months before any other currently used methods, which in turn can also help predict upcoming food crises.
Descartes Labs collects and analyses about five terabytes of images daily from NASA, ESA and various commercial satellites. It then cross-references this information with other relevant data such as weather forecasts and prices of agricultural products.
This data is then entered into the company’s machine learning software, tracking and calculating future food supplies with surprising accuracy.
By processing these images and data via their advanced machine learning algorithm, Descartes can collect remarkably in-depth information such as being able to distinguish individual crop fields and determining the specific field’s crop by analysing how the sun’s light is reflecting off its surface. After the type of crop has been established, the machine learning program can then monitor the field’s production levels.
“Corn are these little factories that absorb or reflect certain kinds of light based on where they’re at in their growing life cycle,” said Descartes Labs CEO, Mark Johnson, in an interview with Motherboard. “Basically what we’re looking at is how much energy is being absorbed to create corn. We don’t have agronomists on our team — we have a bunch of physicists.”
Preliminary results from Descartes Labs show that the machine learning technique is accurate already
At this point in time this technique is only being employed in the US, with data from all other countries only reading “soon” for now on the company’s website. That being said, Descartes’ machine learning algorithm has been able to produce extremely detailed information on a weekly basis for each of the 3114 US counties. That is very impressive when considering that the US Department of Agriculture (USDA) is only able to release monthly reports on a state by state basis.
According to Descartes Labs, with the company’s first test of the algorithm on 6 August 2015, it came within 1.9 percent of the USDA’s report on corn yield. More notably, this prediction came a remarkable five months before the USDA report released in January 2016.
“Our theory is observe every field in the country every day, rather than what the USDA does which is take surveys from thousands of farmers every month,” CEO Johnson explains. “If you’re only reporting every month, a lot changes in that time, like weather forecasts. Knowing those things as they change is really important.”
Descartes’ bank of geospatial data already boasts 3 petabytes of detailed information. As this database of imagery increases in size, so does the accuracy of the data.
Featured image: Yumi Kimura via Flickr