in ,

How TV White Space Is Enabling The Farm Of The Future

FarmBeats, a program of Microsoft
Photos courtesy of Ranveer Chandra

You’re sitting down to watch the evening news and you start flipping through the channels but you have to flip through a bunch of channels of that annoying white noise. Now thanks to FarmBeats, a program from Microsoft, they are looking to use Microsoft cloud and AI technologies to enable farmers to easily and cheaply use precision agriculture on their farm.


Can you take us through the genesis of FarmBeats?

FarmBeats started as a research project at Microsoft. The reason I started it was because of my background. Even though I do not have a Ph.D. in Agriculture (I am a Ph.D. in Computer Science), I spent a lot of time growing up on a farm of my grandparents in Bihar – the poorest state in India. Those farms did not have any electricity or toilets and through my research at Microsoft, I have always wanted to give back. Hence, I started this project in early 2015 at Microsoft Research. We also participated in the Microsoft hackathon and this won the Industry Award at that hackathon. More recently, in November 2018, my team, along with myself, moved to the Azure product team to transition the research into an actual product that could benefit the most people – farmers and consumers alike.

Tell us about the mission of FarmBeats.

The goal of FarmBeats is to enable data-driven agriculture. We believe that data and insights, coupled with a farmer’s deep knowledge about his/her farm, can help the farmer be more productive, reduce costs and practice sustainable agriculture. However, getting all the data from a farm is very difficult. Most farms do not have Internet connectivity or power outlets and it is not feasible to put lots of sensors in a farm since it interferes with the farmer’s day-to-day operations. With FarmBeats, we are able to capture large amounts of data from a farm and aggregate many data sources – such as from satellites and weather stations, which farmers can then translate to actionable insights.

It’s amazing how many farms don’t have reliable access to the internet. TV White Space is so exciting and seems like such a simple and cost-effective way to make data-rich farms possible. Tell us about the technology used to make this happen.

The TV White Spaces can enable a scenario where you can access your Wi-Fi connection from several miles away. The way it can work is by putting Wi-Fi signals in empty TV channels, referred to as TV White Spaces. These channels are usually just white noise when you browse over-the-air TV – the ones you watch using antennas. The reason these channels are so beneficial for sending and receiving data is that they are in the lower frequencies and, compared to Wi-Fi at the same power level, the signals propagate 4x further in UHF TV Channels and 12x further in VHF TV Channels.

The reason they are so appealing for agriculture is that the TV towers are in cities – where most people are – while in the middle of the farm most of the channels are just white. The more white noise channels there are, the more unused capacity there is. And this can provide over 500 Mbps of unused capacity in the middle of a farm – across which we can not only send data from sensors but from drones, cameras, tractors and many other devices.

What sort of data will be collected by FarmBeats? How will that help the farmers to increase their yields?

FarmBeats helps partners gather data from sensors, drones, tractors, satellites, weather stations and other sources. It will aggregate all these sources of data using Artificial Intelligence (AI) techniques to predict data points in places that do not have sensors. Finally, it will enable developers to easily create new AI algorithms that can bring the benefits of data science to agriculture. With all these tools, farmers will be able to get a view of their farm that was previously incredibly difficult, such as a soil moisture map of the farm. With these views, partners can provide insights to farmers that can enable data-driven agriculture in a more affordable way.

FarmBeats 101

Precision agriculture is the new buzzword in farming. The problem with this is that it is often expensive to enact and you need reliable internet connectivity, which a wide portion of the global farm population does not have. FarmBeats is Microsoft working with farmers to enable data-driven farming. Some of the research that they’re working on include…

  • Techniques to merge drone imagery with ground sensor data. 
  • TV white spaces based long-range sensor networks on the farms. 
  • Automation of drone flights in the farms. 
  • Improving drone battery life. 
  • An IoT gateway device for agriculture. 
  • Cloud support and machine learning services that are useful for agriculture.

How far out are we from seeing this technology adopted by the masses?

FarmBeats is in private preview and, at Microsoft, we are working on building a platform that makes it easy to collect large amounts of data from the farm and to aggregate different data sources (such as from satellites and sensors) to create additional insights. However, as you know, Microsoft is not an agricultural company. Our goal is not to sell a product to the farmers, but to build a platform by which agricultural companies can enhance their offerings to the growers.

How affordable do you hope to make this technology?

Our research has shown that the TV White Spaces can bring down the cost of sending data, compared to alternative technologies. Similarly, the AI algorithms can provide insights using far fewer sensors than are otherwise needed. Our ultimate goal is to significantly bring down the cost of data collection and analysis, enabling farmers of all sizes – from smallholders to large scale farmers – to reap the benefits from data-driven agriculture.

What do you think the farm of the future will look like?

The farm of the future will have a farmer who knows a lot about his/her farm and is equipped with tech to make the farming more productive, efficient and sustainable. Technology will provide that farmer with data from all parts of the farm, AI to predict how the different parts of the farm will perform in the future and ground and aerial robotics that help the farmer act in a timely manner. As a result of this new tech-infused way of farming, we will start to see many younger farmers in agriculture – helping us bring back the younger generation to farming!

Ranveer Chandra Chief Scientist, FarmBeats, Microsoft
Ranveer Chandra
Chief Scientist, FarmBeats, Microsoft
Photos courtesy of Ranveer Chandra

What do you think?

US Annual Total Farmland Returns

The Democratization Of Farmland

Dr. Xin Sun

Shining A Light On Precision Ag