 I hope you enjoyed the thunderstorm yesterday. I think it was a nice difference to the days before. But even if you don't like rain, it's crucial for our survival. Without rain, the farmers can't grow the basis of our nutrition. And that is not only true for vegetarians. Without corn, there is no bacon. As you can see, we are dependent on things high above us. And soon you will see that the sky isn't the limit. Please welcome Lisa Gutembeut. Hello. Thanks everyone for coming. I guess everyone survived the storm. Today I'm going to talk about satellite imagery and agriculture, questioning privacy, data protection and autonomy in the field. And I'm going to talk about two areas that are interconnected. The first is... Sorry that I have to interrupt, but I have some other information. This talk will be translated into German. So if you have a decked phone, you can call 8012. And you will hear the colleagues from the booth who are doing a great job. And yeah, Lisa is from... ...they have a tent right next to us. So if you have any questions or are interested in projects, you can go right over there after the talk and ask and talk with them. So now I guess I have all the things from my list done. So the stage is yours and good luck. Okay, thanks. So again, I'm going to be talking today about satellite imagery and agriculture, questioning privacy, data protection and autonomy in the field. And there's two separate areas that are interconnected that I want to talk about. And the first is the data security issues of farmers using software and farm technology, which have the possibility of collecting, using and selling the information. And then also the tension between privacy and transparency with the given availability of satellite imagery in agriculture. I also wanted to give this talk at CCC because I think we can use it as an example. I think sometimes when we think about these massive issues like data protection and the internet, it can be too overwhelming to kind of grasp and solve. And if we look at kind of micro examples like how farmers have dealt with data protection issues, then it can help us to kind of solve bigger issues. So and there's two really specific things that are different about how farmers dealt with data protection issues. And that is, first of all, that they were really concerned about the data security from the get-go. There was no, I have nothing to hide. It was more that I don't want anybody to do anything with my data without me knowing about it. And if someone's making money off of it, I want to cut. And they have also gathered together really well and created an agreement which has been undersigned by a lot of agricultural technology providers called privacy and security principles for farm data. So just quickly, an overview of what I'm going to be talking about in order. First, I'm going to talk about how satellite imagery is used in agriculture actually and how that's led to the digitization of agriculture and the widespread use of farm management software and how the availability of satellite imagery for observation of farm activities has been so widespread in the past that it's really changed the field and how there's an opportunity for transparency, but also the privacy rights of farmers haven't really been analyzed or addressed thus far. So first, how satellite imagery is used in agriculture. Here's an image of, when we think about satellite imagery, we think of some equivalent of Google Earth, I think. And that's only a small portion of what satellite imagery can be applied to. That's only one kind of set of spectral bands that shows you visual light or visible light. And what's more revolutionary, I guess we can say, in agriculture is near infrared light, which reflects off the chlorophyll in plant cells and therefore shows the photosynthesis levels. So when it's applied to a whole field, you can see the variation which is in the center photo. And they've also created indices to kind of refine it even more. So you have NDVI, which is Normalized Difference Vegetation Index, which shows more, refined more exactly or more easily readable where there are problems on the field. And I'll explain a little bit further. So back to basics, this is the electromagnetic spectrum. That tiny rainbow strip on it is what's visible light. Light bounces off objects and back into our eyes. And our eyes are kind of perfect receivers for these bands of light. And then there's decreasing wavelengths which are like x-rays, which we found ways to capture this information and use it. And the same is for infrared, which is slightly increasing wavelength. And specifically near infrared, we use essentially filters to capture this information. So just to show, this is an image that I took last year for a project in the Prinzessninggarten. And the image on the left, I think, is an infrared image. And I used a public lab toolkit to essentially hack a camera where I took the manufactured filter out and put an infrared filter in so that I could take pictures in infrared. And so you can see the pink areas on the photograph are plant life and everything that's not plant life is not pink. So essentially near infrared just bounces off the chlorophyll in plants. And also I should say this is a close-up image and it's more useful at a distance so you can see more variation. And the image on your right is also part of the public lab. They have a tool called Infragram, which enables you to upload infrared pictures and gives you more information through it. And again, this is just a depiction of how it works. This is RGB, like visible light bands, and they reflect off the surface of a plant so we see that it's green. And then the infrared, which in this picture is purple, goes a little bit deeper and reflects off the chlorophyll cells. So again, there's the visible light and infrared light and when this information is put together, it's used very often with normalized difference vegetation index. And you can see in this picture that there's some red areas which aren't doing as well. They don't have enough nutrients. There's something wrong there. And what this has enabled farmers to do is actually spray field inputs exactly where they're needed on a field. So previously, they would just spray kind of blanketly over a field where fertilizer and pesticide was needed and there were areas of the field that actually were growing just fine and that wasn't necessary, so they were essentially just spraying way too much overall. And this has been great for farmers because it lowers economic costs because on the whole, they have to spend less money on fertilizers and other inputs and it's also meant less environmental damage because specifically with nitrogen, there's less nitrogen leaching and there's less kind of pollutants in the waterways and drinking water. Another way that satellite imagery has been applied in agriculture is in yield forecasts. So knowing how much you will have at the end of the season to harvest and satellite imagery is kind of traditionally or previously been known as a documentation tool, something that tells the past. So it's great as being this unbiased overview tool that you can uncover many different things because of the spectral bands, what the spectral bands enable, but in agriculture it actually does tell the future essentially because if you read the photos and the levels at a certain time in the season then you can make pretty close to perfect predictions of how much yield you're going to have. And this has actually been used, this technique has been used by the governments for decades. In 1977 during the Cold War, the US ran the LACI program which predicted wheat yields in different countries. So they did it in the US and they did it in Canada and they also did it in Russia. And I think one can imagine if you know what the wheat yields in a country are going to be and you know maybe it's not going to be enough and they're going to have to import foods from other countries in order to feed their people then it's kind of geopolitical and intelligence importance. You have a bit more bargaining power then. And the FAO, the Food and Agriculture Organization has also used the Amos program to predict when there will be a food shortage so they can prepare and there's actually an image of that, just a screenshot of what it looks like. It just shows that they focus on wheat, maize, rice and soybean and it's open to the public, it's released monthly just to give ideas of how things are doing and farmers use it now to predict their earnings and plan for labor and resource needs. Actually in Germany with sugar beets it's a big thing because it's a very labor intensive crop to harvest. So both of these things, these two examples have led to really the digitization of agriculture. Previously I think farmers were one of the most disconnected communities and now they're using farm management software, they're online, they're using big data and just to illustrate this a little bit I collected a few quotes so Tom Wilsack, the United States Secretary of Agriculture says data is quickly becoming one of the most important commodities in agriculture and this also illustrates that the technology firms that have been offering the software recognize data as a commodity as well, they know that the information that they're gathering is very useful and the economist says it could be the biggest change to agriculture in rich countries since genetically modified crops and it's proving nearly as controversial since it raises profound questions about who owns the information on which the surface is based. It also plunges stick in the mud farmers into an unfamiliar world of big data and privacy battles. And six months ago several members of the American Farm Bureau said data wasn't even a topic on the agenda. Today it's in PowerPoint slides, hallway conversations and question and answer sessions. So what are their concerns? What are farmers' concerns? What's going to happen with their data? So some of them are that the commercial secrets could be sold or leaked to rival farmers that the data could be used to buy underperforming farms and run them in competition with other farmers. If you're a big company and you know how well a farm could be performing but they're not doing so well, you can make them an offer and buy them out essentially. And you can use data on harvest to trade on the commodity markets and influence the stock market. And just basically many farmers are independent in principle. They've wanted to kind of live itself sufficiently and not be bothered by all of these issues. So they've organized around this. And what's been done this far is that privacy and security principles for farm data have been written up and actually signed by many of the largest agricultural technology providers. And they've also created the Open Ag Data Alliance who seeks to create a secure data ecosystem that enables data security, privacy and interoperability for the entire agriculture industry. So just quickly, this is a screenshot. It's actually only a two-page document. It's very concise but comprehensive about what farmers can expect and what agricultural technology firms can expect to adhere to. And it's not legally binding, but it's a really good first step and maybe something to consider in other sectors where there's more specific needs rather than providing broad overarching rules and agreements. But the thing is that even if farmers aren't using software, their activities can be closely observed by satellite imagery that's available. So satellite imagery, it's actually been a great thing that satellite imagery has been opened up to the public, like Landsat data is available and governments and international organizations have kind of been using this information for years. So this is just an example. This is the Brandenburg viewer and it shows infrared imagery around Brandenburg. This is a development seed tool. It's really interesting and just came out, I think, earlier this year, which makes it easier to obtain up-to-date Landsat imagery. And they also offer different spectral bands. So these are true color images that you see on the screen, but you can also get infrared, you can manipulate it. A lot of different things can be revealed. So one of the ways that this could be beneficial and bring kind of transparency to the agricultural field is by exposing environmental issues, just as an example. So this is a picture by Mishka Hanna of kind of cattle lots. And he wasn't even looking for this. He was looking for something entirely different and just stumbled across these cattle lots with these blood lagoons, as he calls them, these really like environmental disasters kind of. And he has a whole series of them and I recommend looking at them. They're very interesting. So yeah, satellite imagery can be used as this way of exposing wrongdoings in the agriculture industry. And the European Union Common Agriculture Policy has also used this. This is just an example of illegal dumping. But yeah, actually, Common Agriculture Policy uses satellites to check on farms that they're doing what they've said they've done in order to obtain the subsidies. And that's regular practice and has actually saved a lot of money from the previous method of field visits. And this is actually a picture. It's kind of hard to interpret, but what it is in Australia, it depicts over time the lands that was taken over or the forests that were cut down to be used for agricultural purposes. And here we have to kind of question to what degree do we demand transparency from the agriculture industry but also small farms and to what degree do they have a right to privacy. And so in thinking about this, I use this scale or just put some things together and made this scale of, as individuals, privacy is a human right. We demand privacy. But on the other side, governments, we want a lot of transparency. We want to know how money is being spent, how decisions are being made. And in the middle somewhere, there's private companies where they have some degree of privacy, but they are held accountable to things like consumer protection and not doing environmental damage and providing good working conditions. And then there's agriculture, which is somewhat like that, but it's also that it's highly subsidized in Europe at least. This is the information for Europe, which is that 43% of EU spending in 2013 went to common agricultural policy, both in direct payments to farmers and for rural infrastructure. And this was 57.5 billion euros. So it's obviously a public interest. It's a taxpayer interest. So you have this expectation that there's more accountability expected there. Yeah. But at the same time, there are individual producers and there's big agriculture firms and there's small producers that... I don't think that we want to go to this place where in our organizations and private companies that we have a video camera in our offices, which is essentially what satellite imagery uniquely provides for agriculture fields. And just to kind of depict this scale a little bit more, it's Neil Abraham of CIS, in DSS Privacy Protection to be inversely proportional to power. So actually, that's essentially all I have to say, and I just wanted to introduce the topic. And to summarize, I wanted to say that the agriculture community has taken great strides in a short amount of time to data security of software and equipment they can use. And I think we can use this as an example as a community. But at the same time, agriculture is unique and that most activities are highly observable by anyone. And there are still open questions regarding the tension between privacy and transparency of farmers with the abundant availability of satellite imagery. And this is an open question and I hope that I've provided enough information that that can be a dialogue about this because in my research I haven't really come across anything that's thought about this or started addressing it. So thanks. Thank you very much for your talk. As we always suspected, cameras and satellites are a dangerous combination. But as we have seen, there are also good things that can be done with this technology. If you have any questions, you can go just to one of the microphones that are on the left and right side, I hope. And oh, there we have a race. I would say the one on the right side has the chance to pose his first question and then we will go on to the rest. We have a little bit of time left, so I guess everyone can ask. Hello, thanks for the talk. What exactly are you working on and how much people are working on it and where can we get in touch? That's actually a difficult question because I work for tactical technology collective in Berlin, so this is kind of the privacy and data protection issues that I work on. But I also am just completing a master's degree in agriculture economics at Humboldt University. And my thesis was on satellite imagery in agriculture, how it's used and kind of the environmental benefits that we can gain from applying this and also the privacy issues. So yeah, and for that work, I did interviews with farm management software companies and this is actually kind of the idea for this talk because I was discussing with them about data protection issues and they were so aware that especially German farmers were very concerned about the data protection policies being very strong and they didn't want their data sold to any third parties. So, you know, they reinforced this and referred me to their Datenschutz Erklärung but then when I asked the question, but most of that information you can get anyway by the satellite imagery that's available, just see this change in their face, like, oh, like I didn't think about that. So yeah, yeah. So how can you get in touch? That's my email address. But I'm not sure how I'm going to take the project forward, actually. I'm just kind of registering everything and getting my studies done. So, yeah. Thanks. So we can switch sides. Your question, please. Thank you for your very informative talk. Very interesting new field to me. I'm wondering whether the government's interesting to have data and perhaps too much data to exercise control or too much control. Yeah. I mean, there's a whole dialogue about satellite imagery in general, not even specifically to agriculture but there's a big discussion about the resolution and how they just passed, I think earlier this year, that it's down to 30 centimeter resolution in the US and that's only what's publicly available for sale. Like, we don't know how refined the government satellites are and exactly what they can observe. So it's an issue. I'm sorry, that wasn't my question. My question is that lawmakers and policymakers in Berlin and other capitals, they have agriculture policy and they want to have information for that purpose and they may want to have too much information which hurts privacy. Is anything happening in the field? It might be required by law, for instance, just to collect it to make these data available. Yeah. I mean, the only laws that, if I understand the question correctly, the only laws that are really restricting what satellite imagery can be accessed and actually Mishka Hennad is a really good art project. Is that not the question either? But it's like a military or government sensitive thing but other than that, all satellite imagery is available and there's, to my knowledge, no discussion happening about what can and cannot be observed besides that. So the next question, please. Thank you for the very interesting talk. One thing that this reminded me a lot of was of colonial projects to map agriculture in North Africa, for example. And I was wondering if you were looking at also the ways that not only the ways that satellite surveillance can give governments and corporations an idea of what's on the ground but also how it may provide them with the mistaken assumption that they know more than they actually know and craft faulty policy based on this knowledge. Yeah, I mean that's an interesting question and I don't think I have a real answer. I mean of course there's this thing of trusting data too much and making assumptions based on what's available but actually I have to say with satellite imagery there's not a lot of human intervention. It's really just unbiased kind of images that you can take different spectral bands but I think the information that's being gathered with satellite imagery is like to this hyper level of accuracy. So whether they're interpreting that differently is, yeah, I don't know, but the data that's being collected itself that's the kind of revolutionary thing about satellite imagery is that it's extremely accurate and unbiased. It's just taking pictures. Hello, thank you for your talk. You talked about near-infrared imagery. Have you ever heard about synthetic aperture radar imagery which is done by radar waves in another spectrum area? Do you know if this is also used or used together with infrared? Yes, actually today I only talked about infrared but there's long wave infrared, thermal infrared which actually telling the heat of different plants also does a lot and also this radar thing as well is applied in agriculture and actually one of the firms that I interviewed was looking into it but they hadn't really implemented it yet and the benefit to radar is that you don't have to worry about cloud cover because it's a really big issue in getting satellite imagery in general. So if you have radar you don't have to worry about cloud cover because the data is just collected anyway. So yes, it is applied in agriculture for sure. One last question from the right side. Okay, basically it's about your personal opinion or attitude or where do you locate yourself on this axis between privacy and transparency? For example, in one concrete case the European Union Space Agency has this Copernicus program launching these Sentinel satellites. A few weeks ago there was this Sentinel-2 satellite which will offer wonderful data in different spectral ranges to basically everybody in Europe or in the world terabytes of data which will be public. So basically what do you think about this? Is it good that basically everybody with connection to internet can download everything about the planet? And is there any... in regards to these data shoots, for example, could there be collisions like one party trying to persuade the public that this shouldn't be public or that their privacy issues, privacy rights are in danger? So my personal opinion, I think that there's been more good done with... I think that it's brought so much transparency and the fact that citizens actually have access to this data is really important because governments have had it for decades and they've just been kind of watching over. And now that it's in the hands of citizens, that's great. But it's really hard to tell. I had a few discussions this week about it because there's also different-sized farms and big agribusiness, you might want to hold more accountable or not more accountable, but they have more power and therefore maybe more should be demanded of them than small-holder farmers. But it's just so varied. You can look at a plot of land and gather soil information, plant information, land use change, weather, like carbon emissions. There's so many different aspects that you can look at a square hectare of land for that it's really hard to say, okay, well, this is a small-holder farm so it's my land and I don't want anybody else to gather any other earth-relevant information to it. It kind of gives this other dimension to it that goes beyond property rights, I guess, if that makes sense. So my answer is I don't know. I'm stuck in the middle because I see both sides and I see the really great things that have been exposed with satellite imagery, but I also see the potential for that to be really big privacy issues. So the next talk is coming. It will be about FPGAs. I hope you will enjoy this talk as well. Once again, a big thank you to you and to all of you.