If rural connectivity challenges are going to be solved, private sector investment and innovation will be necessary. Dr. Ranveer Chandra, the Chief Scientist of Microsoft Azure Global and Partner Researcher at Microsoft Research, joins the FarmBits Podcast to discuss Microsoft's investment in enabling data connectivity and interoperability in agriculture. Ranveer initiated the FarmBeats project at Microsoft Research which has recently led to the launch of Azure FarmBeats, which is a cloud platform enabling data-driven insights in agricultural applications. The FarmBeats project has also resulted in innovative approaches for establishing broadband connectivity on farms, such as using vacant TV white spaces to facilitate data transfer in rural areas. In this episode, Ranveer covers everything from the development of this TV white space technology, to emerging technologies in the developing world, to how value can be extracted from data once it is consolidated and made available. While Microsoft is not a traditional player in the agricultural industry, the data pipelines and enabling technology they provide have significant potential to enhance agriculture alongside the expertise of experienced agricultural professionals.
"For growers, it would be more about thinking, 'Hey, I know a lot about the farm, which are the parts if I had this additional data that I could do better?'" - Ranveer Chandra, Ph.D.
Opinions expressed on FarmBits are solely those of the guest(s) or host(s) and not the University of Nebraska-Lincoln.
On this episode
FarmBeats Website: https://www.microsoft.com/en-us/research/project/farmbeats-iot-agriculture/
Azure FarmBeats: https://www.microsoft.com/en-in/campaign/azure-farmbeats/
GatesNotes Wi-Fi Chip Blog: https://www.gatesnotes.com/Development/FarmBeats
White Space Deployment in Africa: https://www.microsoft.com/en-us/research/wp-content/uploads/2016/02/main-5.pdf
Economist Article: https://www.economist.com/science-and-technology/2016/09/17/tv-dinners
Ranveer Chandra's Info:
FarmBits Team Contact Info:
Samantha's Twitter: https://twitter.com/SamanthaTeten
Jackson's Twitter: https://twitter.com/jstansell87
Jackson: Welcome to the FarmBits podcast, a product of Nebraska Extension Digital Agriculture. I'm Jackson Stansell. (and I'm Samantha Teten)
Samantha: and we come to you each week to discuss the trends, the realities and the value of digital agriculture.
Jackson: Through interviews and panels with experts, producers and innovators from all sectors of digital technology we hope that you step away from each episode with new practical knowledge of digital agriculture technology. Welcome to the third episode in our connectivity series here on the FarmBits podcast.
Samantha: We've heard from Dr. John Fulton and Dr. Keenan McRoberts so far in this series regarding the importance of connectivity and how public and private partnerships can drive improvements in connectivity.
Jackson: For this episode, we are excited to welcome Dr. Ranveer Chandra, the Chief Scientist of Microsoft Azure global and partner researcher at Microsoft Research.
Samantha: Ranveer initiated and is still heavily involved in the FarmBeats project at Microsoft Research, which has recently led to the launch of Azure FarmBeats, a cloud platform in enabling data-driven insights for agricultural technology companies and other experts.
Jackson: In this episode, we will hear from Ranveer on many topics associated with FarmBeats from how the FarmBeats projects brought innovation to real connectivity to how Azure FarmBeats is enabling better decision making once that data actually gets to the cloud.
Samantha: This interview kept us on the edge of our seats so here's our interview with Ranveer Chandra.
Jackson: So, you know Microsoft is not a traditional entity that's involved in agriculture and so why was the company on board with you kind of starting this project and getting involved in ag?
Ranveer: Yeah so, at Microsoft Research we are expected to take risks expected to do things that are out of the way that is not necessarily something that Microsoft does right now and that is why when I started palm beach of course it wasn't a big project. I was given four interns, I was told okay see what you can do and that's how the project started and the key thing here was that we believe that the future of agriculture is data-driven. It's going to be driven by data, by technology and even though Microsoft might not be the ones building the tools directly used by the farmers we could be building the basis for a lot of tools in ag tech, in food tech and that's what we are doing at Microsoft. At Microsoft right now the approach we are taking to industry be it with agriculture be it in any other industry is we are building the digital stack for agriculture and the customers for us are our partners essentially. They are the ones who are building solutions on top and taking those solutions to the growers in this case with FarmBeats.
Samantha: Awesome so, can you expand on that data-driven decision part a little bit and how is the Farmbeats project working to achieve that what are you guys doing exactly?
Ravneer: Yeah, I know that's a great question Samantha. You know, when I started Farmbeats the first thing I did is I went and interviewed lots of farmers. I went and I talked to them. I volunteered on a couple farms. I stayed for one week in in upstate New York in this farm, and I essentially attended these meetings at 5am. I saw how hard farmers were what I realized when I met each of these farmers is that the farmers know a lot about their farm. They have been farming there in some cases for years in some cases for decades, if not for generations. Some, like one farmer he could feel the soil and see what's going on. Another farmer he would just taste soil and see what's going on they know a lot about the farm, yet a lot of decisions these farmers make is based on guesswork like they know a lot about their farm yet when to water how much to water when to seed when to harvest all of these decisions are based on guesswork, which is where I believe one of the things that we could enable is with data-driven agriculture we could augment a farmer's knowledge about their farm with data and with data driven insights. That is we can get the data from the farm show, them what the farm looks like right now. We can then use artificial intelligence to start making predictions of not what is it now but what will it be an hour in the future a day, in the future up to five days in advance, so that a grower can take smart intelligent decisions and that's the kind of future that we believe Farmbeats can enable data-driven agriculture.
Jackson: Yeah, so how much of a challenge has the lack of rural connectivity you know here in the U.S. and obviously in other places as well you know how much of a challenge has it been in terms of getting that data to where it can actually be used for artificial intelligence and some of these data-driven applications?Ranveer: Yeah, no that's another great question. That's the first thing we need to do to drive the eye as you rightly pointed out is to get the data. Get the data about what the farm looks like right now and then you'll be able to make any predictions, but getting this data is extremely difficult. There is yeah you can get satellite data but satellite data is coarse-grained it is not seen below the canopy, it's not seen below the ground, it's good data it can't see through clouds lots of issues with satellite data right now for which you need on the ground information. But, if there is no internet how do you get data from the middle of the farm and that is one of the first things we realized that this needs to be addressed with Farmbeats and with Farmbeats we started looking at this problem of how pervasive is this lack of connectivity. It's huge like you know farmer's house might have connectivity a lot of the discussions that were happening was about bringing broadband to the households even that is a very important problem, but if you have to enable this data-driven agriculture things like precision agriculture and so on you need connectivity not only in the household but in the farm as well, and that is one of the first issues we have to address as part of our leads and then we have to look at other issues as well for example there's no power- there's no electrical outlets where you can plug in your devices. That is you how do you can't be putting lots and lots of sensors these sensors are expensive so how do you gather information without putting lots of sensors in the farm? So, these are some of the other issues that we had to address but yeah broadband was lack of broadband in the farm it was one of the first key realizations when we started to envision this concept of data and agriculture
Samantha: How are you actually going about addressing those?
We've heard of a couple of creative ways but can you kind of fill us in on that?
Ranveer: Yeah, one of the creative ways I'll talk about is something that I started working on in 2005 called tv white spaces, so this is one of the technologies that I've written some of the first papers on is on a way to use unused tv channels to send and receive data that is what the tv white spaces enables is. Imagine if you have say a wi-fi router that you can access a few miles away you know right now when you turn on your router as soon as you leave your house or your building there's no internet there's no one. With this we could go much much longer and the way we did that was we put these wi-fi signals in empty tv channels this is tv you watch using antennas over the air tv you know when you browse through these channels on certain channels you get a transmission the other channels all you see is white noise nothing coming there. The technology we had there, it was a way to take this wi-fi signal and to put it in these noisy tv channels in a way that doesn't interfere with your tv reception in an adjacent rack, so you could be watching channel 7 at home on channel 8 people sending wi-fi signals and the reason this is so cool is that compared to wi-fi at the same power level in UHF tv channels your signals go four times farther in in UHF. They go 12 times farther and that's in free space purely based on physics, if you put it through crops, canopies, forests the signals just keep going through. It goes much more because the reason is that just for people who are more tech savvy here is you know the lower you go in the frequency space your wavelength becomes bigger it can go through obstructions much better. So, this is something that we've done, we've deployed back in the day in 2010 the FCC chairman had come to Microsoft campus to see the demo we had put together this was made legal in the U.S. in 2010. Since then, we've gone out connected high schools, hospitals dispensaries in various parts of the world using this technology. In fact, with the air band initiative, Microsoft announced in 2017 we had made a pledge to connect 3 million Americans to broadband using this technology. In fact, if you look at what's happened most recently, in fact it was late October, November the FCC passed new regulations to make the tv white spaces even more favorable for rural areas for farmers. So, the new regulations allow for higher antenna mounting so you can mount the antennas higher they allow higher transmit power levels you can even geofence an area you can say this is my farm area and over here I can start transmitting at higher power. So, you can really envision the future of this is going to be like a wi-fi like scenario you want your farm connected put in this antenna put in this device and you can then start getting data from all your devices in that region and not just getting small sensor data. These are broadband signals you could be getting large amounts of data large amounts of bandwidth because the key insight that we had for agriculture is that tv towers are where people are people are the ones who watch tv so in cities most of the tv towers are in cities if you go outside the city most of the channels are just white noise like if you're not keeping the farm most of the channels are unoccupied the more empty channels there are the more available capacity there is so if you're talking even if say 20 tv channels are available we are talking of each channel is 6 megahertz right so 20 times 6 120 megahertz of available spectrum. We are talking about a few hundred megabits per second if not more using one spectrum so this is like unused spectrum and if the regulations are there we need that could be one way to bring broadband to the farm but that said one of the things I would like to highlight is that this is just one method there's no clear just one solution that we'll have to use a combination of solutions to address this problem. Like for example, we are also working with other companies for example with SpaceX, we announced a partnership with starling using low earth orbit satellites. You can use those to bring connectivity as well now of course the way you would use it is your bring connectivity from satellite to a point. You can then use a terrestrial solution for example like tv white spaces or it could be 5G, it could be private LTE. You could set up any of those in combination with so all of these technologies need to work together along with another technology called edge compute that is rather than sending all of the data to the farm, you could be running things on an edge computer and together you might be able to address the problem of broadband connectivity. But, going back to the point you both of you raised Samantha, Jackson that is we need to address this broadband problem if we are to get to the future of agriculture which is going to be data driven by AI. If, that is a future we need to bring broadband to the farm.
Samantha: Yep, those are really cool solutions that you guys are coming up to and we'll be interested to see how they get implemented. Another challenge can be the cost of technology, so there's a lot of interesting solutions out there and these costs can kind of pile up for some producers. So, how is Farm Beats potentially helping though provide solutions to connect some of these technologies or to maybe make it more of a feasible option right?
Ranveer: So, with Farm Beats what we are doing is building this data layer in the cloud what Farm Beats is right now is it's a data platform in the cloud or for ag tech companies what we're doing is we're allowing you to bring in data from different data streams and then run your artificial intelligence pipelines in that. That's the state right now, but this there are few things that we are continuing to do investigate as far as research is still in research on trying to bring down the cost of these solutions even more. So, tv white spaces is one way to bring down the cost the other way we are looking to bring down the cost is by reducing the number of sensors that are needed in the farm like for example suppose the question was what is the soil moisture level six inches below the soil throughout the park, what is the soil nutrient level if you have to solve this problem say just the soil moisture problem both of you are in the agriculture department you would understand this problem much better that if you have to build such an accurate map you would need lots and lots of sensors you would need a sensor say every 10-15 meters if you have to build an accurate map. But, putting lots and lots of sensors is expensive to deploy to manage it will come in the way of the farmer as the farmer does a day-to-day job, so the question we also said well if you have to build a map you can't put lots of sensors how can you bring down the cost of building such maps? The way we did that is by using artificial intelligence. Our key idea was well if you could use put only a few sensors in the farm and then you use aerial imagery and then use artificial intelligence to then bridge these two modes of sensing to build maps, the idea was very simple if you have a pond you intelligently decide where you put a few sensors and then you use technology such as drones or satellites to get the aerial image what you do with the aerial images then whenever you have a sensor you look at what does the aerial image look like. You use that to train a machine learning model and then you use that machine learning model to predict these values in other parts of the farm where you have the image. But, you don't have sensors and we go to paper on this we showed that using such a technology you can be three times more accurate than existing schemes that don't use aerial imagery. So, this was one way in which we can bring down further bring down the cost of solutions, but we could do more and we are continuing to invest in other ways to bring down costs like initially for example we were using drones and as you know drones can be expensive they still cost a few thousand dollars if not more, and they only lost say 30 minutes on a single charge. The question we asked was we went to Africa for example we wanted to do these trials in Africa and India with smallholder farmers, once we went there in one country we didn't get permission to fly drones we were told that he go get the permission from the ministry of defense at which point that they're never going to grant a few researchers on the permission to fly a drone. So, then our question was how can you get aerial imagery from a farm at low cost so the way we solved that problem is we went really low tech our idea was to use helium-filled balloons these are about four to five feet in diameter what and these balloons are tethered to the ground with a rope and they fly up to 150 to 200 feet what we built was a weatherproof mount where a grower can place their smartphone with a camera facing down and a battery pack attached to it and this thing can stay out from four to seven days taking images of the same part of the farm and in fact what is that. So, there is a farmer close to Microsoft campus he used this to monitor floods in his farm. He could see because right now what used to happen was he would come every morning to his farm, if he saw any amount of flood in the pond he would throw away his entire crop because based on regulations he cannot sell any crop that is touched by the flood. So, with this we actually had visual proof to see which crops are touched by the plant and only throws away those crops in fact we could do a bit more than that like in places like India and Africa where labor is inexpensive you get someone just walk around with the balloon over the entire farm and then we use computer vision and artificial intelligence algorithms to stabilize the imagery and then we construct these aerial views of farms, so you can see how because this is another way to bring down cost where if you have instead of using expensive drones you use these low cost balloons. There's some more cutting edge stuff that we are doing where the kind of vision we are enabling is can you use your smartphone to start measuring soil moisture and soil and seed that is by using wi-fi on the phone. This is a very new research which we did where you know these sensors that you buy right now, they're expensive they're still $300 if not more, but many of these small holder farmbers they won't spend a few hundred dollars to buy a sensor in the farm because the question they ask is what is the ROI, what is the return of investment, if I put a sensor it will only tell me what's happening here not what's happening a few meters away. So, then the question we the idea we had was that well farmers won't would spend a few hundred dollars for a sensor but many of them have a smartphone even if it is an inexpensive hundred dollar smartphone. If they have a smartphone, it has a wi-fi chipset in it, if it has a wi-fi chipset in it the key idea we had was that the time of flight of a wi-fi signal depends on the permittivity of the material and so for example if the soil is moist the signal will take longer to traverse the same distance except this distance is in this time is in nanoseconds and wi-fi doesn't have the bandwidth to measure this in nanoseconds and the next idea we had was that most of these wi-fi chipsets have multiple antennas in them. So, instead of measuring the absolute time of light if you measure the relative time of light then you could actually use that to measure soil moisture and soil electrical conductivity, so we wrote a paper on this we built a system we showed it to Bill Gates when he visited the farm that we have close to Microsoft campus and in fact the blog that he wrote on Farm Beats the title of the blog that gets his blog was can the wi-fi in your phone feed the world. So, we are super excited about this is like cutting edge tech just the first research paper has been written a lot more research needs to happen before it really becomes mainstream, but that I believe is the future where we can then equip every farmer with tools so that they know what's happening in the farm both above the ground and but these are all different things we are coming at it from a computer science side of things from an electrical engineering side of things seeing how can we bring technology to help the farmers to help build ag tech solutions that because this is all data we provide. Both of you I'm sure you could do lots of things in farmland's life.
Jackson: Yeah, that's the hope, and it's so interesting how you're thinking outside the box and making use of those technologies that are already in place that farmers already have but just how can we maximize the efficiency of them that's what it sounds like to me?
Ranveer: Yeah, no it is that is the case that is a lot of these tools in computer science are being used for other industry part of what we are also doing is thinking outside the box in the context of agriculture and trying to bring technology, which are being adopted in other industries to agriculture.
Jackson: Yeah, that's really cool and you know since you've kind of brought up the kind of the low-cost thing and also all the algorithms like the machine learning algorithms and artificial intelligence that are kind of that you're using now to almost for lack of a better word interpolate data from sensors I kind of want to get into the Farm Beats Azure platform, and so who kind of are your intended users and what are you seeking to achieve with that Farm Beats Azure platform?
Ranveer: Yeah, so a lot of what I was describing with tv white spaces with wi-fi based sensing balloons all of this was in research right we were all thinking in terms of two parts of Farm Beats- one was data acquisition how do you get a lot of this data to the cloud and the second piece is once you start getting all of these different data streams from sensors from drones from satellites from tractors how do you merge all of these to drive new insights like the interpolation algorithm was one we've done more in that space around microclimate prediction. So, then I transitioned over from research to products and we shipped the first version of Azure Farm Beats and Azure Farm Beats is this platform in the cloud for other ag tech companies to build their solutions that is we are able to bring data from diverse data streams from the farm and then allow farmers to then allow these ag tech companies to run their AI machine learning algorithms on top to take it to the growers, so that’s the platform that we built essentially it's about being for any piece of land what one beats would allow through APIs is you should be able to provide a polygon which is a shaped file and for that polygon which should be appeared, which could be a fun we then bring in lots of data streams from satellites, from weather stations, from farm equipment, from sensors, other data streams and then as you can tell these data streams have different time resolutions have different spatial resolutions with drones you get centimeter resolution with satellites you'll get meter level resolution. So, how do you merge all of these how do you run AI on top of this, so this is a platform that we're building with Farm Beats. But, then the way it goes to the market is through partners is through ISPS. We are not building a grower based solution, so for us a farmer might be using these arctic tools with palm beach behind the background but they might never know that they're using these solutions behind the scenes and towards this we've done several partnerships both in terms of partners whose data we can bring to the platform as well as in terms of part in terms of partners who build the solutions on top. For example, we recently announced a partnership with DTN, which is a big weather company because we are bringing DTN's weather to the Farm Beats platform, so once you define a shapefile or a polygon you can start getting those the data the weather data from DTN to the platform and then regarding the actual farmer-facing offering, we recently did a partnership with Land’O Lakes where Land’O Lakes is building their digital agriculture platform on top of one beats and we are bringing the latest in tech while Land’O Lakes is bringing all of the agricultural expertise to build those agricultural solutions on top of Farm Beats.
Samantha: Yeah, that's really cool that is we actually heard you speak at a conference a couple years ago and just it stuck out to us that collaboration piece we just remember you talking to these people and saying like we want to work and learn the agriculture side from you guys and that was just so cool and so with this collaboration piece there can be some other challenges how do you guys distinguish like the ownership of models and algorithms created from these other companies or whatever and putting them together with Farm Beats if that makes sense?
Ranveer: Yeah, no that's a good question Samantha. It's about how do we drive this partner ecosystem how do we build this collaboration on top? So, in this case a lot of these AI expertise that comes on top like for example building a corn-based model or a model for wheat we are not the experts in corn and we will never get them and these are the things which we want to empower our partners together we want to empower our partners with the latest tools, with the latest in artificial intelligence, latest in internet of things edge compute so that they can then start building the solutions on top the models. By itself it belongs to them, we are providing the platform and they build the models, but we want them to take their models to the agriculture to other companies who we work with. For example, one in the Land O’Lakes scenario which is going back to them one of the partners we work with there is tavant technologies, they are the ones who are building a lot of some of the technologies there which we are then taking to Land O’Lakes so the IP then belongs to whoever's actually built the model while we are more in the process of bringing the technology, empowering them to build these models which otherwise would have been very hard for them without these technology tools. So, we think of ourselves if you think of it as a tools provider as someone who's building these tools for anyone in the ag tech space to do more right now they're at a certain level. With Farm Beats with the tools that we're building we want to take them to the next really big step up from where they are right now.
Samantha: That's really cool, that's incredible. So, what do you foresee being Azure Farm Beat’s biggest opportunity for growth in the coming years?
Ranveer: Yeah, so see the vision here is to make every farm farmer use data driven technologies right the goal that we have the what we want to get to is a world where every so it's about food it's about not just feeding the world but nourishing the world and nourishing the world sustainably without harming the planet. If you think of that as the future, what we are doing now with Farm Beats is just scratching the surface Farm Beats is just around the farm but once you start bridging it across the entire food supply chain is when you can start unlocking new efficiencies. If you look at so if you look at the entire food supply chain all the way from input companies to equipment companies to the farm to the logistics companies the food distributors the processors the detailers it's an entire chain. Each entity in the food supply chain could use data to make themselves more efficient, like Farm Beats for the farm you could do a similar thing for a manufacturing plant. When these entities start sharing data with one another we believe you could start unlocking even new efficiencies in the entire food supply chain and that's the vision that we want to get we want to get to a place where you should be able to exchange this data you want to be able to use this data to drive new efficiencies to make sure that you're growing with less, you're directing food, you're reducing food waste, you're taking it to the people who need them and you're giving everyone the food that they need. The ultimate vision here is that consumers would treat food as medicine. They would treat, they would eat food that is best for them and everyone has some different food could work differently for each person and you the world will get there it's about how do we accelerate people, accelerate the entire food industry to go in that direction and the farm by itself plays a very important role. That's the genesis, that's the central point yeah based on which the entire food chain would be done. Yeah people are talking about alternative proteins and things like that lab based meats but if you even if you look at that the genesis of that they're looking at some of the food the inputs to that are things like chickens and moon beans and millets, so these are things which again have to be grown somewhere and they need to be grown the right way, so that you can make the right food. So, yeah so that's the vision we are after of getting everyone the right food that they need and in order to get there, Farm Beats is just scratching the surface. There's huge amounts of opportunity in the entire supply chain.
Jackson: Absolutely and I mean all of that has to be enabled, and I’m going to kind of backtrack a little bit or I guess really not backtrack but tie things together right, so all that has to be enabled by connectivity and so if you're thinking about this this rural connectivity aspect do you think that addressing this this particular issue is going to be more of a private and public emphasis on just kind of bringing existing technologies into cooperation and just you know implementing them in rural areas or do you think it's that we still need to develop more technologies that actually fit these rural areas appropriately fought for broad for connectivity?
Ranveer: Yeah, I think for connectivity we need to do more than what it says. I think we don't have all the solutions if someone says we do, I don't think we do. We still need to do more just to prove my point around the world about 40 percent of the world doesn't have internet access and the reason they are not on the internet is not because they don't have internet coverage like 90% of the world has some sort of internet connectivity to connect to the internet it's just not affordable enough. So, I think being able to build these affordable broadband technologies is key especially if you're, when you're talking of the farm a farm there's no human there usually in the middle of a farm you need to get internet of things data how do you get this data at scale so that you can drive all of these decisions is I think a problem that needs to be addressed. They are very promising technologies like tv white spaces I would call this promising technology leo satellites is another promising technology. But, a lot more needs to be addressed because the other thing I wanted to point out about connectivity in a farm is the definition of broadband. Right now, the way the FCC defines broadband for example is 25 megabits per second download and three megabits a second uplink connection. From a farm, most of the transmissions will be opening and not down you might want to even flip it for farms when you're talking of broadband it should be three meg down 25 or even more but existing technologies don't get you there like for example with leo satellites again your capacity is in the downlink direction not in the update direction the uplink capacity is still constrained, so how do you get to a future where you can provide pervasive high bandwidth connectivity, uplink connectivity to the cloud is an unsolved problem. Part of the solution is edge compute how well you don't need to set everything up, you could be doing things on the edge, but that's also evolving. Like, we have a solution for Microsoft and we need to adapt it even more to see what can be done on the edge. For some other things, you really need to send things to the cloud so those are things that need to still be addressed. I think though if you were to ask me what is the state of the art, I would say we have a lot of promising tech but we still need to address, we still need to come up with innovations to address the problem of broadband and the other piece is part of it is a technology problem. The other part is a policy problem where you need the right incentives, you need to have the right policies in place. The FCC has been doing the right things, but I think a lot more needs to be done and worldwide I think FCC is still the leader. As far as policy makers are concerned, I think worldwide other regulators need to be moving fast as well to make the right decisions to enable broadband in the farm.
Samantha: Sure, from some of your responses it sounds like you know developing nations have kind of provided a lot of inspiration for some of these technologies that really push some of these things forward how do you go about developing a technology that fits you know developing nations and where the United States in our agriculture infrastructure is and how is it going to impact these different countries and their different agriculture systems?
Ranveer: Yeah, that was that's a good question right Samantha, and I struggled with it as well when I started doing this project as to who do we target and if you were to ask me right now I think we have built technology that works reasonably well for farms that are tens of acres or more. It's something that if you talk of data-driven agriculture solutions, they work at that level they still don't work for the smallholder farmers at mostly because of the cost and that's something that needs to be addressed because like around half a half a billion farmers majority of them are small holder farmers. So, how do you get technology to them so that's why part of the technology that has shipped are things that work at tens of acres or more while we are still thinking in terms of how do we how do we get to the last minute how do we get to the farmers who are small that's still something where we are continuing to do research like the things they talked about, if we can get this wi-fi sensing to work that can really help solve a problem. In fact, I'm writing along with Stuart Collins at the Gates foundation we've just written a white paper on how do you enable digital agriculture for small world of commerce, this is in the review process but that's something where we put down our thoughts on what are those technology breakthroughs that are needed in things like connectivity, in artificial intelligence, in speech in human interaction that is you know some things we assume how to use a phone well it's we can't just assume that for many people around the world the literacy levels are not the technology saviness is not there so that's in that why people we try to cover a lot of these problems where we are still continuing to actively invest in addressing that problem because that really needs to be addressed if we are to make farming worldwide to be data driven.
Jackson: Yeah, that's amazing, I'm looking forward to seeing that white paper and we'll be sure to share that with our listeners as well on our streams because I think that's something valuable that I think it's easy for us to kind of put out of our minds when we just think about our lives here but you know for anybody who's traveled internationally I think they understand you know what the different challenges are especially with those smallholder farms because it looks nothing like what you know a giant row crop farm here in the Midwest looks like.
Ranveer: The reason I started Farm Beats was because of this exposure to the farms in India that is all the projects that I've done at Microsoft at Microsoft Research. The ones I've started have this element of trying to impact the people worldwide not just the people in cities but people in remote areas and that was one of the triggers for Farm Beats as well that is I want to build tech to get to improve the lives of farmers of everyone worldwide. So, I guess one question I really would like to ask you after you know hearing all of your responses is not only what do you hope to see in the agriculture industry in the next five to ten years but what do you really think we're going to be able to get to within the next five or 10 years what do you think is a practical goal to have I think we will see we will start seeing more adoption of these data-driven agriculture techniques these digital techniques I think what I'm seeing based on the trends we'll see of course much more adoption in the medium to large farms especially in the developed world farms in the us they all go in that direction to start using more technology adopting more data we'll start seeing more and more of that we and part of the part of the trigger would be of course driven by productivity. But, it will also be driven by sustainability that is for using the right practices in the farm in order to get carbon credits for example like with the carbon bank proposal that's coming up at the USDA and so on you would need to provide information about what's happening in the farm you really can't go and manually be getting all of that information you can start using data, once you start adopting these techniques like for example if you're doing cover cropping and you're using noted or reduced you need to know what the impact of that is going to be on your productivity how can you still be productive even more productive while using some of these regenerative agricultural practices part of that would be driven by data. So, what I'm seeing is an increased adoption in digital agriculture techniques because of these that the trend towards being more sustainable the acknowledgement and the need to put this sequestered carbon in soil to make the soils richer, but that's something which I think is going to happen is going to accelerate worldwide. I hope a lot more happens, I think that's where I’m a bit more concerned. I'm seeing that and it needs to happen otherwise this problem when we talk about this divide between what's happening in the developed world and the emerging markets is going to keep on increasing and that's where the governments need to take the right decisions to make to help these farmers adopt the right tools to be more productive because you know the other thing that we see we talk about climate change. Well, when we talk of climate change we talk about how agriculture could be a solution the thing is climate change could impact farmers as well and it will impact the smallholder farmers even more because these are the people who are not equipped with the right tools to deal with variation in weather. For example, a few degrees here and there unexpected rainfall these guys are going to suffer, so unless we address our problem we need to move towards more climate smart agriculture being able to provide these tools these digital tools so that they can adapt they can take the right, so I think that's why it's imperative for us to even start thinking in terms of smallholder farmers and how we can bring technology to them so that we can help them be more productive as well.
Samantha: Yeah, absolutely yeah the sustainability piece and the carbon and like we keep hearing more and more about that which is great and I'm sure we could have a whole another episode on just that part as well, but to wrap everything up what piece of advice do you have for individuals involved in the agricultural industry for you know developing the skills necessary to be successful or how to get connected in this digital agriculture landscape?
Ranveer: Yeah, so people in the ag space now of course they're like you on your grad students are doing cutting-edge work you're already in this learning a lot of digital tools and working with the right people but anyone else who's in this space, I would encourage them to pick up some of the core data skills which includes artificial intelligence, machine learning just the basic skills of how do you how do you use any of these tools in your daily decisions. I think that would be helpful just the fact and for growers it would be about how do you they don't need to learn machine learning but it'd be more around if you have data to augment your knowledge in the farm how would you use this data and there are a few examples and there are few farmers who we worked with who've been really amazing. There's this one farmer he takes out he farms over 9000 acres in in eastern Washington spread across 45 miles and whenever he takes out his tractor he looks at the predicted weather forecast using AI. The thing that we're doing for micro levels to see whether it's the right time to be spraying and over the next few hours because once he takes the tractor out he's going to be out for a few hours. If it starts getting windy, it's going to be a disaster so how do you use data to start taking your decisions he uses that to even he knows his farm he knows some of the topographies in his farm what the land looks like and then he uses this data that we're providing with how his crops are growing to even decide how he should be seeding the next year. So, this is where I think for growers it would be more about thinking hey I know a lot about the farm which are the parts if I had this additional data can I do better right and that's something which the doors would do and the other ag tech professionals I think it's both about data and learning the skills where needed because the more of these tech skills that you need the more you'll be able to do the more productive you'll be able to do and the more sustainable agriculture we'll be able to practice.
Samantha: That was a great discussion with an incredible innovator Ranveer Chandra regarding Microsoft Farm Beats and Ranveer talked about a lot of resources and articles that they're working on in publishing, so if you would like to read into any more of that please check out our show notes for those links.
Jackson: Yeah, those are going to be great resources for people out there I think you know when I think of this interview the word that comes to mind for me is enabling because that's really what I think Farm Beats not only are they doing but that's what they're really seeking to do is they're seeking to enable ag tech businesses that are trying to build on their platform they're seeking to enable farmers in the Midwest and they're also seeking to enable small shareholder farmers with all their innovation right. So, I think it's pretty cool and I just think enabling is a good way to describe what they're doing.
Samantha: I completely agree they're doing a great job of identifying challenges globally but then not just thinking about what can we invent to fix this they're thinking about what can we make that's accessible and that's incredible you think about your cell phone your smartphone advance example that was awesome to hear about yeah all really cool technologies.
Jackson: And so, we hope that you all who are listening enjoyed that episode as much as we did and we also hope that you enjoyed hearing Ranveer's vision for the future of agriculture. So, next week we're going to jump from this global company Microsoft that we all know about to more of a local startup Realm Five that's actually right down the street from the University of Nebraska and learn how they're addressing real connectivity challenges for data-driven solutions we look forward to seeing you then.
Transcripts are generated using a combination of speech recognition software and human transcribers, and may contain errors. Please check the audio before quoting in print and write firstname.lastname@example.org to report any errors.