Precision agriculture: data and AI for the field of the future

Brand Studio for Microsoft

Precision agriculture stopped being thought of as something futuristic and distant to become purely present.

This agricultural discipline is increasingly used and allows the best possible decisions to be made, in the right place and at the right time, preventing generalities.

In this direction,

Andrew Nelson takes multispectral images to document drainage and the amount of downed weeds.

Photo: Dan DeLong for Microsoft.

Aimed at researchers, data scientists, and farmers, this toolkit can be used to turn agricultural data into actions that help anticipate weather events, increase yields, reduce costs, and solve connectivity issues.

In addition, there are other Project FarmVibes tools that will be open sourced in the future, this includes FarmVibes.Edge, which intelligently understands vast amounts of data from drone scouting flights.

the field of the future

Andrew Nelson of Farmington, USA, is a fifth-generation farmer and partnered with Microsoft Research to add data to his 7,500 acres of wheat, barley and lentil crops.

All of them

This information that he receives daily allows him to see the variations in temperature, soil moisture and nutrient levels, plant health and among others.

Andrew Nelson, reviewing multispectral drone imagery.

Photo: Dan DeLong for Microsoft.

Nelson, in turn, presents that data in Project FarmVibes,

 He is currently using

FarmVibes.AI.'s algorithms, running on Microsoft Azure, predict the ideal amounts of fertilizer and herbicide to use and where to apply them;

forecast temperatures and wind speeds in your fields, inform when and where to plant and spray;

determine the optimal depth to sow based on soil moisture;

as well as learning how different practices can affect the carbon in your soil.

The US farmer said: “Project FarmVibes allows us to build the farm of the future.

We are showing the impact that technology and Artificial Intelligence can have on agriculture.

This allows us to save a lot of time, money and also helps us control any problem we have in our field”.

Andrew Nelson performing maintenance on his White Spaces TV broadband tower.

Photo: Dan DeLong for Microsoft.

Until recently, Nelson's course was like many others around the world.

He had internet at his house, but the Wi-Fi signal was dead when he walked through the door.

Its 7,500 hectares were a dead signal zone.

To solve this problem, you are using one of the Project FarmVibes tools, called

This provides broadband access via TV White Spaces, the unused spectrum between channels.

Currently, Nelson has a solar-powered antenna that acts as a Wi-Fi router, but has also allowed him to cover most of his field.

That connectivity has allowed him to use:

• Async Fusion:

   Combines drone and satellite imagery with data from ground sensors to deliver insights.

  Nelson, for his part, uses this tool to create nutrient heat maps from multispectral drone imagery and soil sensor data.

  These serve to vary the speed at which you sow and the application of fertilizer, helping to increase yield and avoid overfertilization.

  In turn, Async Fusion can also create soil moisture maps from sensor data.

  These indicate how deep to sow and in what order to do so.

  As an added benefit, they can help prevent tractors and sprayers from binding up.

  
  

• ​SpaceEye

  – Uses AI to remove clouds from satellite imagery.

  This helps Nelson fill in the gaps in areas he hasn't explored with a drone.

  He can then feed these images into AI models that identify weeds, helping him create maps to distribute herbicides only to areas that need them.

  And even at spray time, they allow you to vary the application rate, delivering more volume to densely weeded patches and a lighter load elsewhere .

  
  

• ​DeepMC

  –

  Uses sensor data and weather station forecasts to predict temperatures and wind speeds for your field's microclimate. 

  
  

• ​A

  what-if analysis tool

  that estimates how various agricultural practices would affect the amount of carbon in the soil.

  
  

However, the US farmer is currently testing other Project FarmVibes tools beyond FarmVibes.AI which will be open sourced in the future.

This includes FarmVibes.Edge, which intelligently understands vast amounts of data from drone scouting flights.

This efficiently builds images small enough to upload to the cloud via FarmVibes.Connect.

Undoubtedly, these technologies generated a great impact both in the fields of Nelson as well as in their economic yields.

Food in sight

Food is one of the biggest problems in the world, which is why by opening its most recent research tools, Microsoft wants to disseminate them to contribute its grain of sand on this issue.

“By 2050, we will need to roughly double global food production to feed the planet,” said Ranveer Chandra, Director of Industry Research and Director of Technology and Agrifood at Microsoft.

Ranveer Chandra, Director of Industry Research and Director of Technology and Agrifood at Microsoft.

However, as climate change accelerates, water levels drop and arable land disappears, doing so sustainably will be a huge challenge.

“We believe that one of the most promising approaches to address this problem is data-driven agriculture.

At Microsoft, we are working to empower farmers with data and artificial intelligence to increase their knowledge about agriculture and help them grow nutritious food sustainably,” said Chandra.

The

 available on GitHub.

This includes: Async Fusion, SpaceEye, DeepMC and the "what if" analysis tool.

Chandra points out that most farmers around the world are not like Nelson, who is as comfortable coding as he is in a combine.

Most will not download these tools on GitHub.

Instead, Microsoft wants to inspire industry and academic partners to translate this research into tools that can be used by all farmers, including small farms in the developing world.