 So, my name is Clara Duzal, I'm a data analyst at SDSN and I'm part of the Fable Consortium, which aims to understand how countries can transition towards sustainable land use and food systems. In particular, we ask how countries can collectively meet the SDGs and the objectives of the Paris Agreement. Thank you so much for joining us today and I should let you know that this webinar is being recorded and that it will be made available on the SDSN website later. We will have an opportunity for questions and answers with the panelists. So please do you the Q&A option from Zoom to submit any question that you have as they come to mind. So today's webinar, as you might know, is about digitalization and sustainable agriculture. So why we wanted to do a webinar on that topic? Well, because the world stands at a critical moment to deliver on the 2030 agenda for sustainable development and innovation is seen as an important means to achieve food security and the sustainable development goals. Innovation is important to realize the productive potential of family farmers, especially in small and medium sized family farms that occupy a large share of farmland and produce much of the food in low and middle income countries. So on top of the existing challenges that the food system was and still is facing, the COVID-19 pandemic has put unprecedented stress on agricultural food and led bear many weaknesses in our institutions and systems. This has generated much attention on digitization as a solution to several challenges and many have emphasized that digitization can reduce the need for middlemen and brokers and also to facilitate a greater share of profits going back to farmers. Many also emphasize its potential to build in resilience to shocks as well as reduce inefficiency through greater equality of access to robust information. So we are going to have the opportunity to explore some of these questions today by presenting case studies in digitization from different national contexts and operating at different scales. So I'm very pleased that we have with us today Xinyi Lim, she's Executive Director for Sustainability and Agricultural Impact of Pinduoduo in China. Pinduoduo is China's largest agricultural platform, connecting 16 million farmers to over 800 million active buyers. We'll also have Daymika Linsen and he is CTO for ISDA Africa and today we'll be presenting virtual agronomists. A virtual agronomist is an artificial intelligence designed to help smallholders apply the right amount of fertilizer and predict realistic prop yields. We will also have Emeka Muatine Mere, who is CEO of Kitovoo Technology in Nigeria. Kitovoo provides smallholder farmers with the data to make small decisions about what to grow that would sell and how to grow them optimally. And finally, from a more global perspective, we will have Jaouku, who is a Senior Research Fellow at IFPRI, the International Food Policy Research Institute, which is a food policy think tank that conducts research to develop policy solutions for reducing poverty and ending hunger and malnutrition. His team studies the potential of digital innovation that can help low and middle income countries achieve SDGs. His presentation today will focus, we'll discuss new challenges that policymakers will need to understand and address for effectively managing climate risk using digital technologies. So I'm going to hand over to you, Ksenia Lim, who will present Pinduodoo. Thank you. So thank you very much for inviting me to be part of this sharing session. I'm Ksenia Lim. I'm the Executive Director of Sustainability and Agricultural Impact at Pinduodoo. Hopefully you can all see my slides. I've just got a couple here that will give you a little bit of an idea of who we are and what we do. So we were founded in 2015 and we have since built up a very large user base of over 860 million buyers on one side. And then on the other side, we've got about 16 million farmers who are connected through our platform and they're able to sell their produce directly to consumers across all of China. In 2020, we facilitated about $42 billion worth of agricultural produce sales. So that includes things like fresh produce, fish, meat, rice, oil, tea. So pretty much running the whole gamut and really helping the farmers in China gain more market access. And this has been something that we've been focused on since day one. So when we were founded in 2015, we started out doing agricultural product sales before we then branched out into other categories of goods as our user base got larger. So because we've been working on agriculture related digitization since the first day of our company's founding, we are also very aware of the opportunities for digitization to bring more efficiency into the agricultural sector. So to that end, last year in August, we announced our 10 billion agriculture initiative, which aims to address critical needs in the agricultural sector and rural communities. So that's us pledging up to 10 billion RMB of our profits, starting from the second quarter last year to go towards investing in projects. So it could be research projects, it could be projects relating to farm training or infrastructure, for instance. So in a nutshell, what we do is basically digitizing the agricultural product sales. So on the slide here, you'll see on the left screenshot of our user interface. So being a platform that started in 2015, we're a mobile only platform that emphasizes more savings, more fund, that's our slogan. So it's very intuitive and interactive kind of user interface that you see here. It's a feed that is presented to you. So not so much oriented by search, although there is a search bar at the top as you can see. But it is basically dynamic and adjusting according to how you interact with the app, what sort of interest you're showing. So here we've toggled to the fruit category. So you can see there's a variety of listings that are being recommended to the user. And so as the user shops that interact, they might be able to discover some agricultural products that they previously maybe weren't aware of or were not specifically looking out for. So this in particular is also very helpful because for agricultural products, in particular, we find that users can also be influenced by their friends around them. So potentially the orders can be aggregated. So what do I mean by this? If you look on the right side of the slide, you'll see the detailed page of a listing that we've clicked into. So this is dried mangoes. So on the platform, we also have a function called the team purchase. So on the bottom of the slide, you'll see that there's a pink box that has 21.6 inside. So that's the individual purchase price. And then the red box has 8.3, which is the team purchase price. So significantly cheaper if you buy as part of a team. Now how do you form a team? Because the platform is already the largest in China by number of users, actually just forming a team involves just one more person. So it's actually very easy. So what I could do is if I want to buy the dried mangoes, I could click on it and share the link with a friend. So say I want to invite Clara and I sent her a link through my social networks, in this case WeChat. So she could just click on the link and join me in the purchase. And we would both pay 8.3, which is a team purchase price and get the box of dried mangoes shipped to our homes individually. So it's very convenient for the user. They're incentivized to help share the listings. And it's also very powerful for the farmer because now instead of me buying the mangoes and maybe Clara would buy the apples from another seller, she actually got persuaded by me sharing the link because I'm a friend. She trusts my recommendation and she thinks, actually the mangoes are pretty good as well. So I don't mind having mangoes instead of apples. So that then helps to aggregate more orders onto a narrow range of suppliers. So they actually get a lot more visibility into the orders that they have in a given window of time. So this then translates to more upstream savings, whereby they can actually coordinate on the logistics and also then pass on some of those savings as the cost proposal comes down because of volume, pass it on to the consumer. So this just illustrates what I was talking about in terms of streamlining the distribution chain. So over here on the top half of the slide, you can see what is a typical kind of offline distribution value chain. So the farmer who's selling garlic in this example maybe gets quoted a price of two RMB per kilogram. So a lot of the farmers in China, the smallholder farmers, they don't really have the ability to set pricing. So what then happens is that there's a lot of different intermediaries involved, the regional distributor and the wholesaler etc. Before it finally makes its way to a retail outlet where the consumer maybe in a big city picks it up for 16 RMB or kilogram. So all of that markup was actually lost in the middle where there was the manpower cost, there was the warehousing cost, distribution cost etc., which in today's digital age doesn't actually need to be incurred. So on the bottom half of the slide, what you see is a more streamlined distribution chain that is also more transparent to both the farmer and the consumer. So on one hand, starting on the right, you've got the consumer who now maybe is able to buy the garlic at 4 RMB per kilogram. So it's a no-brainer for them to pay a quarter of what they used to pay, but at the same time, the farmer is able to still make more, 30% higher income because now after taking out the additional kind of packaging or shipping cost that they previously didn't have to be footing, they're now able to actually still make a higher income. So that's also a win for them and the consumer gets fresher produced in a shorter window of time. So there's also less food loss and waste that's incurred along the way. We estimate it's about a 40% reduction in the volume that would have been lost otherwise. And another way that we're trying to transform kind of the midstream efficiency of agricultural sales online is through this service called Duo Duo Grocery, which we launched just over a year ago. So what Duo Duo Grocery does is trying to match localized supply and demand together so that the consumers are able to get the products in less than 24 hours. So in the model that I outlined earlier, there could be some specialty fruit or produce that's only grown in one part of China. So if I live in another part and I just really want to get that specialty fruit, I would have to wait maybe about two to three days for that box of fruit to reach me. But if it's something that is actually a lot more common like rice or eggs for instance or parsley, things that are grown in a lot of places and also available locally, you could actually have that delivered and travel a shorter distance to you. So for these sorts of products that many people are counting on for their daily necessities for cooking, they're actually able to place the order on our platform by 11 p.m on the first day and pick it up themselves after 4 p.m. the very next day at a local pickup point of their choice. So what this does is that it cuts out the last mile delivery costs, which then becomes a cost saving that we can spread out with the consumer. The consumer has a better user experience. They're walking maybe 200 meters to the nearby barbershop or the convenience store and they're picking it up and they're also getting pretty fresh produce because it's traveling a much shorter distance. So this then gives users a lot more flexibility in terms of having another way of buying fresh produce and getting it in a very short turnaround time. And so ultimately what we hope to do is leverage the insights that we're able to have as the leading agricultural platform to help the midstream where we work with the third party logistics providers become more efficient. So we are then able to reduce further food loss and waste if we can actually use some of those insights in terms of predicting a certain amount of demand is going to show up in a certain place by a certain time, what is the volumes that need to be moved, what is the best way that they can actually do the route planning so that they minimize downtime, how can they improve the efficiency or the utilization rate of the trucks, etc. So what that does is that it further reduces the cost of shipping these agricultural products to the end consumer. Now the midstream is one part of it but the upstream is also very critical. So for all efforts to consistently yield a good effect, we need to actually work with the farmers. So we have a program called Duoduo Academy where we've been training farmers both through online and offline courses in terms of how to run a business effectively. How do you manage customer service? How do you take good photos? How do you do a live stream? How do you manage sales? So these are things that may not be apparent to a farmer but are critical for success online. So we're proud to say that we've actually trained about half a million farmers so far and of them we have about 126,000 new farmers who are basically younger people who have gone back to their hometowns to start their own businesses, be agricultural entrepreneurs and they in turn mobilize the rest of their community to be part of their businesses. So either they procure from them or the other people also join in in the packaging or in the logistics side of the businesses. So we also have partnerships with agricultural experts for instance through China Agricultural University, we're able to provide further classes online to the farmers for instance on specific topics so that they can manage their crops better and we've committed a training another 100,000 new farmers over the next five years. Duoduo Farm is another example of how we're working on the upstream to improve the efficiency of agricultural production. So what we've noticed is that sometimes there's technologies that are available and are known to the farmers but they're not being applied perhaps because of challenges around scale. So with the partnership of local governments as well as experts from agronomy institutes what we've done is to actually create a series of farmer cooperatives. So in this example here on the right side of the slide that's a picture from our citrus farm in Yunnan. So when I say our citrus farm we don't own the farm but we help to start the formation of the cooperative. So the farmers who live in the area are now organized in a cooperative they collectively farm a much larger plot of land which means they can now actually deploy some technologies like drip irrigation and drone spraying which then reduces the labor inputs that are needed which is very important for an aging farming population and also reduces the fertilizer costs. So they're using 15% less fertilizer as a result. So with a tap on their phone they can use drones to spray instead they can you know turn on the fertilization system etc. So all of that is automated it's labor saving and ultimately you know the farmers are able to make a better income as well because with the expert advice they're able to grow a variety that is also using a higher price because it's you know it bears fruit of season so there's greater demand for these kinds of produce. The smart agriculture competition is another way in which we're trying to yeah so we're trying to bring technologies closer to the farmers by demonstrating to them what some of these cutting-edge technologies like IoT and machine learning based kind of crop models are able to deliver in terms of better crop yields. So in 2020 we held the first competition where the AI teams that competed were able to use their crop models and also harness the power of IoT to grow three times more strawberries than conventional farmers competing in the adjacent greenhouse and this is competition which just ended also had a very successful result whereby AI teams were again able to harness their crop models to not only improve the nutrition profile of the tomatoes that they grew but also create a disease prediction platform and now you know these AI solutions are being brought to market and scaled more widely across the country. So ultimately what we hope is to serve as a convener to use our platform for good and to invest where there are critical needs to improve the overall efficiency and productivity of the overall agri-food value chain. Thank you for your time. Thank you very much, Zingyi. Next we'll have Jamie Kodinson presenting virtual agronomist. Thank you Clara and thanks Zingyi for the presentation it was fascinating for me as well to listen. I'm Jamie I work for ISDA we're a mission-driven company and we have focused on Africa in particular our mission is to raise the income of African small-scale farmers. And how do we do this? Our leading product is called virtual agronomist so it's been known in the research community for a long time now since the 90s that site-specific nutrient management isn't known so that's making an individualized recommendation for an individual farm instead of a blanket regional recommendation offers significant savings, improvements in productivity, less fertilizer, more profits. And so how do you do this? You know our dream is to roll this out this technology for every farmer. So we're a member focused on Africa and for African small-holders the question is how do you get this out when digital connectivity isn't necessarily there? What are the things in the way? Why can we not go and give individual advisory to every single smallholder? The first one really is they just aren't enough agronomists support systems are stretched so extension agents often are looking after thousands of farmers and so they just can't give that depth of advice. So what we aim to do is provide a digital solution that means that lower-trained field officers can provide that same value. So we're putting AI in the palm of their hands so that they can do a similar job to more highly trained extension agents or agronomists. Another problem is that advice is often regional or actually aimed at larger farms. And seeing you was mentioning this you know big data in AI means we can start to tackle this problem we can start to look at much smaller farms and provide individualized advice. And the last one I think is a big one really what prevents investment is uncertainty. And so what farmers need is the confidence that they can understand the system they can understand their farm and they can understand what they'll get out of an investment. And so what we really aim to do is to make this a tool for users to be able to understand their farm try different scenarios understand what are the risks what are the rewards of putting in that money. What does it look like I'm not going to give you a demo here but let's just spin through. It's very easy to imagine some AI system that's really complicated. But actually what's key really is understanding what's going on on the ground. What we need to ask is the kind of questions that a local agronomist would what's happened in the field before what are current yields at the moment what's what's the best yield that your neighbors getting you know what have you been growing before what do you plan to grow is there any pest problem here what's your plan you know what what kind of practice do you have. So we ask simple questions this app can be you know a field agent can be trained up to use this very quickly. And then the app is really providing that expert knowledge so it's been tailored for the location for the crop and also for the economics of the value chain. And by working through this the app is then providing a target yield. So many of these applications if you've seen them before this kind of approach it's often the user supplies what they're trying to reach but that obviously isn't realistic you know the an agronomist would come and tell you what's reasonable and that's our approach. What's the output I mean our aim is to provide an individual plan for every farm. So this you can see on the side is a is an individual nutrient application plan for that particular field. So it's very important to know what products are available locally what fertilizer products what their prices are what their compatibilities with those products are. And by combining these individually for a particular field a farmer can often make 30 to 40 saving and that's an economic saving but it's also more sustainable we're not throwing nutrients away by doing this so broad applications tend to be less efficient. And then of course by applying this year after year you have better information from what the farm did last year what the farmer tried last year what worked what didn't. And so by building this up we can work towards continuous improvement that's really what we see as being a sustainable way forwards for small holders is that they can build year on year keep building their profit keep understanding their field keep changing things learning and adapting changing conditions obviously a reality with climate change. So we see this as a holistic approach it's very easy to just look at agronomy in in the small but what we need to do to make recommendations if you are speaking to a real agronomist who's coming to visit your farm they take a holistic approach. So understanding the field history in the region that's very important but it's just as important to understand the economics what should you grow why if the prices are changing inputs and outputs then obviously what you should do will change. It's also understand it's key to understand the dynamic factors so whether pest and disease you need to understand those to factor in what's possible at that location and what the risks are. And really where we start is the soil so this is an important thing to understand each farm each field its capability is really defined by the soil and the ongoing soil health is critical. So how do you do that for small holders they often can't afford small soil tests. So we built a digital soil map we call is the soil this is based on over a hundred thousand training points from across Africa covers the whole continent except the Sahara and by using remote sensing and machine learning we can predict soil properties in particular as a a pH map we predicted at 30 meter resolution and that means we get a pretty good idea of the variation at field level. Obviously uncertainty needs to be taken into account here we can't literally test the soil for every location so this is a prediction. What does 30 meters mean this is a 30 meter grid and those are typical small holder fields so you get some idea of the variation. Now when you can do that you can offer a very low cost recommendation because you understand the soil without having to go through expensive extra hardware or lab tests. So soil maps mean you've got a base level of information and as farmers make more money they can afford more sophisticated tests they can understand their land better and as you build up that data that can feed back in to the digital soil maps which everybody can use so we really see you know this is the data aspect we can build this up as a community and that actually is why our soil map is open day for us it's open for everybody and we encourage people to donate their their soil scans to it. We take a decision science approach so this picture illustrates the common saying you know never cross a stream that's an average three feet deep because farming is inherently uncertain so what you have to do is advise farmers so they can understand the risks they can understand the investments and that is really key to us to understand what risks what investments farmers can make. There are lots of stereotypes about small holder farmers but actually our experiences they can really understand these deep things they just need to be presented right they can understand the pros and cons they are at heart entrepreneurial and that's what we that's what we embrace. Just aware of the time so I will just skip through this. Since this session is about digitization I should mention that really farmers and farming is a personal thing so there's this we don't underestimate the impact of the human factor the last mile so what we actually do to build this digital system to customize it is we work with the local output of input aggregators who are working with small holder farmers every day and we work with them to build this customized for their locality that's the only way you can get this really local knowledge in Africa to build that into the system otherwise you end up with very broad top-down approaches it has to be bottom-up and of course crops respond differently in different locations so models need to be customized for local conditions so we use the best scientific data available including these large-scale donor funded projects to be able to train these crop models so that virtual programmers can understand you know how things will work in that particular location that particular value chain and then I should end just by saying you know the approach then is this is inherently uncertain and so you have to monitor what's going on so we then get feedback from the individual farms and we're looking at what works what doesn't work how can you improve the system and I think that to us is the key to the sustainability pieces you keep getting better year on year I'll end there but if anyone has any questions feel free to put them in the chat and or send me an email if you'd like thanks very much thank you very much Jamie it was really really interesting thank you our next presenter will be Emeka and Wachinamele if you want to unmute yourself Emeka I see that you try to have your video on but it's not yeah but we can hear you perfect it's working now thank you very much so several years ago when man discovered farming it was quite very very simple all you needed to do was get your seed put it in the soil and you were sure of a great harvest at the end of the day since we are not complex there was adequate land and if you wanted to increase production you just expanded the band you had the weather was predictable you knew there were patterns that were used to the soil was fertile and you basically sold to markets that were near you so there were no complex logistic systems to worry about there were no distant markets with different standards different crop varieties in demand and all that and because you are selling within your environment or near your community you quickly sold out to people who are close to you we didn't have to bother about storage infrastructure and all that but all that has changed and it is affecting food production as we know it like you take you through the journey it's a journey of a smallholder farmer in Nigeria so a few years ago just after I finished my national youth service I wanted to find out the reasons why there were some of all those limitations in the agricultural sector because in the course of the national youth service I had done my festival farm so I didn't set out to get involved in agriculture my plan was finish school finish my salaries get a professional job in quotes preferably in the oil and gas sector and then you know build a career from them but in the course of farming I found out that youths were very very poor the outcomes were not provisional today amount of hard work I put in and then was still at the end of the farming cycle it was difficult finding fair buyers the middlemen offer prizes that basically didn't make sense at all so at the end of that period I began to try to find out why this was so what made farming the way it was in Nigerian by extension across sub-Saharan Africa in the course of my search it took me to several parts of the country and in one of those journey I met Hawa Hawa is a typical smallholder farmer cultivating under a hectare um she and she and them other smallholder farmers like her produce over 80% of all the food consumed in Nigeria today but the sad thing is that Hawa and people like her cannot readily afford healthy food stuff for themselves because of their low earnings and these low earnings are largely because of the disruption that has happened in the agricultural space there is climate change the reins are not predictable anymore aside that the farmers I have a situation where the funds are already spent and they cannot afford to do land rotation in Uganda because they are the availability of arable land is not as as um widespread as it used to be so the farm on the same land year after year rely on guesswork to make most of their decisions and because they do that the outcome is typically very very poor yield to put this in perspective an African farmer does one fifth of the output of farmers in the US I think that was not enough when the farmer has produced he loses 40 to 60 percent of everything he produces because he typically doesn't have access to a post-service infrastructure and then when he decides to sell he can lose up to 40 percent in incomes that he could have earned if he got better prices for new farms these are some of the problems that we wanted to solve several years ago when we started technology company so at Hitomu data is at the heart of what we do we use several platforms which allow us to solve the problems of farmers through a data driven approach first we have you max um you max is a satellite based agoramic advisory system he allows us to collect field soil um and uh crop data which we can then combine with satellite data run analysis to provide farmers with information on what type of fertilizer to use what quantity of fertilizer to use the variable application rates so that the farmer is giving the crop exactly what he needs so in a sense the farmer gets soil and crop specific fertilizer recommendations but that's not all we enable the farmer to get weekly crop health audits so the farmer knows how well the farm is growing and he can identify areas that are in need of urgent integration so he saves a lot of man hours from having to go around all his field doing scouting he also knows what the water stress is if he needs to look for sources of irrigation to help his crop but in addition we use weather meteorological data to make sure farmers understand the best good agrib practice to implement for time at the end of the crop cycle when they harvest this material the farmer has a choice he can either sell off immediately if he decides to sell off immediately we have a digital commodity supply service called e-procure which allows us to aggregate different crops that the farmers produce according to their varieties and match it to um demand from commodity buyers on the other hand if the farmers decide that they would like to store through a service called storage x we provide the farmers with um postpaid storage so the farmers get to store their commodities but that's not all when the farmers store their commodities we turn their commodities into an asset base by issuing them an electronic warehouse reset system which looks like something like this it's shareable with an api with financial institutions so what the farmer does is with the electronic warehouse reset system he can get up to 50 percent of the value of his crops that money is a lifesaver for the farmer because it allows him to um meet pressing financial obligations at the same time not be pressed to sell off his produce at the time that he can get the best prices from the market so well how does this work um because smallholder farmers in Nigeria and by extension south swan africa are um not uh tech savvy not literate we use an agent based system first of all we we work with um young people who we train in extension agronomy and um the use of our technology work with them to onboard the farmers in their different locations we then train them train the farmers on good agrib practice and collect the different data sets which we then analyze to begin to provide the services to them for the agronomic um um advisory services we typically bundle it with um actual products um product like fertilizer seeds and chemicals so that farmers are actually not only getting good agrib practice but they can be guaranteed that it's actually quality imputes that they are using they are not using imputes that are substandard in any way so by combining impute the right imputes with um good agrib practice and access to market we enable the farmer not only to increase his yields by over 30 percent but to quadruple summer's losses by 20 percent why achieving reduced impute usage through precision by another 20 percent as a result of this year on year the farmers who work with are able to increase their annual incomes by 40 percent this ties into our overall mission of trying to help develop a resilient food system for african agriculture where the smallholder farmers like hawa and people like her can earn enough to be able to afford nutritious foods um since we started with being able to um enable over 12,000 smallholder farmers to increase their crop yields by 30 percent they've also reduced poor service losses by 20 percent and the unique thing about these farmers is that all of them achieve 100 percent access to markets because we take a market-based approach we first of all understand what the market needs and we get the farmers to grow the exact varieties and specifications that are in demand which creates a captive market for them but in addition it also provides price guarantee for farmers and this as i mentioned before results in a 40 percent increase in annual incomes for them for us the long template here is to continue to aggregate data from different sources directly and from third parties which will help which will help enable us um continue to work to develop better and better products for smallholder farmers products that can enable um financial institutions understand the risk and credit uh credit and the uh risk of farmers so that farmers can assess better financing products data that can help us on um match farmers with um community buyers who buy at fair prices data that can help us provide tello satellite based insurance and even pension services for smallholder farmers the long term our goal is to help transform african agriculture by transforming smallholder farmers helping them produce more from less and in the end it will tie into the global needs to produce 60 percent more food by 2050 that will feed 9.7 billion people so this um what we do at hitovu and how we are helping you the resilient food system for african agriculture are we glad to take your questions thank you very much thank you very much emika i think we'll keep the the questions for after our last presenter who is jao koo from ifry jao koo if you want to show your screen okay um hello everyone i hope you are seeing my screen okay good uh so yeah again my name is jao koo a senior research fellow uh in the environment and production technology division at the international food policy research institute at ifry um so we conduct research to develop policy solutions that can lead to poverty and hunger and malnutrition and our team studies the role of technologies including digital solutions to help countries achieve sdgs so today i'd like to yeah present some stories um where either highlight what policy makers need to understand and how how to use technologies and what are the issues around the application of technologies okay so let's first talk about opportunities so okay good um so at the food systems level in the context of climate risk management digital technologies provide a lot of opportunities to make data acquisition and analysis faster and this is particularly relevant for the rapidly changing climate in the us for example the number of days between billion dollar climate disaster has been reportedly reduced to just 18 days in recent years so every 18 days there is another climate disaster compared to more than 80 days in the 1980s the digital technologies can speed up the data collection and analysis to react to climate risks more timely and climate impact are now more localized and difficult to predict and plan against predictive analytics can help plan the best protective measures at the local level and finally for policy makers that they do not always have the best timely and reliable information to act upon so digital technologies can improve public information systems and allowing stakeholders to access the information more quickly and monitor policy impact on the citizen however um yeah these new opportunities also highlight new challenges that need to be addressed such as the digital divide inadequate information and limited digital capabilities let me present a bit more details on what these challenges mean in practice the first is the digital divide the potential of digital technologies are quite clear yet each reach is not universal rural communities are often underserved and underrepresented in information systems and this can create biases in data and inaccurate information and even misguided decisions and just to quickly show you an example that we learned ourselves about this our hard way a few years ago we had a project that needed to understand the relationship between market access and the level of adopting agricultural technologies like improved seeds we did not have a lot of time so and the study country cell phone ownership was quite high so we thought we could use a quick cell phone sms-based survey the result was reasonable to some extent the father you are away from the market the level of adoption technology adoption decreased however the adoption level increased in very rural areas and can you guess why we tried really hard to understand this pattern uh but we couldn't and the reality was that um yeah the because of the high price of cell phone and uh and the low level of connectivity in rural area people who can afford to own a cell phone answer also in the survey that they are the ones uh better off and other technologies more so this was just biosystem data we uh for a while misinterpreted what's going on so knowing this has many implications for policy research and policy maker decisions any strategies for collecting data in agriculture should be particularly carefully evaluated and continuously revised to ensure the rural population is properly represented in the information system um okay the next the second one is inadequate information and there are many information systems already in place across economies we even started hearing that some countries have just too many information systems and they need to be coordinated better evidences show that however weak information systems with inadequate information waste budget excessive poverty and slow economic growth existing knowledge is not always useful this might be outdated or difficult to apply in practice so here is another sad example that we analyzed last year recently does anybody remember 2015-16 and linear that created drought and flood around the world when FAO issued a scary early warning uh in December 2015 uh in Zambia uh the government issued a precautionary export ban um on maize grains and however in reality what happened was that the extent of drought was not that severe and drought was only recorded uh to uh in the southern part of the country and other parts of the country received normal to above normal rainfall and produce actually is a plus of maize uh this could have been captured in real time in information policy makers and to revise the policy however due to the weak information system uh government just couldn't uh capture that uh anomaly uh on time so export ban remained in place and it ended up depressing price in the market um and even increase the poverty at the end so yeah this case this case again kind of um I confirm that silo data really do not support evidence-based policy responses that could have managed risk better okay the last example our last uh kind of topic is a limited capabilities uh even if you have really good infrastructure and state of the art technology is generating a lot of new data information you also need ability to understand what they mean and use them in decision making effectively more data in themselves are not enough to find all the answers uh technology investment should be complemented with investment skills as well uh here I'd like to show you another example uh that we failed to realize this important lesson early on ourselves uh back in 2020 uh we started a research pilot to test the value of state of the art seasonal probabilistic forecast data this data set uh provided detailed six month climbing forecast on rainfall and temperature anomalies anomalies around the world it was really exciting um so in this map generated using the same data set in April 2020 we saw the south asia especially in india would enter dry spells so we alerted our colleagues underground um to prepare for this but then what happened in May 2020 was the historic flood uh that recorded the large-scale property damage and hundreds of deaths uh the twist of this story is that in fact but the forecast was not wrong it was probabilistic and the map only showed the median forecast and we did have some scenarios a probabilistic forecast scenario that a predicted flood but we just didn't realize how to use the probabilistic information um and uh and we didn't know at the time uh what kind of different interpretation skills that we needed to generate insight um so what we what have we learned here uh there are many exciting digital technology pilots out there uh but they should be introduced together with continued research and even human-centered design to learn how to use and how to present them and also how not to do so otherwise we might keep on developing pilots that do not really scale even worse or misguide decisions so thanks for your attention and I will see you again shortly at the Q&A session thank you very much for this very interesting uh presentation um we have we've had a couple of questions in the Q&A uh box but I've seen that uh the presenters have answered most of them and maybe uh some of the for some of the questions that have already been answered some of you would like to uh go back on it and give more details if not I I have a question for Xinyi oh well you wanted to talk anyway so you can go ahead and I can ask my question after I noticed that there was a question in the chat box so I was just going through the ones in the Q&A so yes there was a question about the profit uh for our platform and how long it took for us to form and I saw that there was another one uh posed by uh Paulson as well about different profit sources for investors who develop an online platform so I thought I would just address both of them so on our platform um so just to make it clear we're an e-commerce marketplace so consumers on on the one hand they would install the app and they can discover all of these products and then the farmers they are the ones who are listing their products or their produce on the platform so what we charge is a very small transaction service fee so that has made me to cover the payment processing costs so the you know consumer is paying 20 RMB to the farmer or whatever it is and there's a small sort of payment transfer fee that has to be incurred anyway so we charge that on the sales and then the main sort of revenue stream on our platform where the majority of our profits come from is actually through advertising so on the platform the merchants can choose to advertise their products if they want to so it's all part of you know running a business so if you don't want to advertise it's okay but if you actually see that you know advertising gives you good returns helps you get more sales you can do that as well so actually when you compare kind of the the sales volume or the the dollar value of what we sell on the platform versus our actual revenue stream the revenue stream is very small right so because it's just dependent on that marketing so you know not every single transaction has a high commission fee or whatever it is and the idea behind that is really to lower the barriers to entry because we want to make it easy and convenient for farmers to onboard and sell their products so that's kind of the question on the business model there I think there was a similar business model question addressed to you Jamie for virtual agronomist yeah thanks yeah I mentioned briefly in the chat but our model is to is to essentially sell to the aggregators who are providing services to small holders so in our model that there's customization done for an individual value chain in region so you need to get together enough farmers that that makes sense you know that that can be affordable so we work through partnership with an aggregator and these aggregators already have typically a connection with their small holder farmers so they're often they often have some kind of extension I mean obviously by the way our situation is very different in Africa too it's as in using in China so the digital connectivity is not as strong there's still this last mile piece with extension agents and so they are offering often a bundle of services so input provision because the logistics is difficult as well as guaranteed offtake as well as credit to harvest and so they strengthen their service to small holders by adding this individualized advisory so that's that's basically how it works great thank you um I see that in the Q&A we have lots of questions coming from for a maker sorry so I'm just going to add them to you right now in maker because I see that you want to answer so um a question is has Kitabu been operating long enough to know whether this will be sufficient to cover your costs okay uh yes um we actually started operations in 2018 so year on year we've been increasing revenues uh in the long term we are going to be profitable one of the challenges we had or the drawbacks um in the fact that and I say this uh with a sense of you know um understanding of the problems uh for Africa digitization alone is not going to solve the problem it's not it's not a silver bullet we still have um critical challenges around the lack of farming and post harvest infrastructure so um and then you know the issues around farmers access to finance those two have the critical bottlenecks clause um so the approach we've taken um is to combine different uh product and services together bundling it together um so that um we can be able to provide something that is critically of value to farmers yeah so it's um it's we are increasing our traction every year and in the long term it's going to be profitable well that's great for you um there is another question that will be specific to gala for you emeka which is which digital technology is the best for linking rule farmers to markets in a country such as as ganna yes so um ganna like other sub-saharan african countries um there are two peculiarities that um stand out that's the fact that um in distant pockets where most of the production happens there is very low connectivity um and then of course there is also the challenge of the literacy levels of farmers so any solution um that is going to create linkages for farmers first of all in terms of the model it has to be agent based um it wouldn't necessarily interact with the farmers directly and then if it would if it would interact with the farmers directly um it needs to be sms or ussd based even then um it has to factor in the need for um aggregation aggregation and send with us because um you can't possibly source something that has effective um sense for supply from just one small holder because of the fragmented nature so you need to consider all that in trying to design a solution that works for either ganna or any other sub-saharan african countries thank you emeka thank you very much and maybe uh do some of the panelists have questions to ask to the other panelists since it's your area of expertise maybe you you would like to ask each other some questions if not i can ask a question and you can jump in if you if you think this is something that you you would have the expertise to to answer so um what can be done proactively to involve historically marginalized communities like women use people in extremely remote locations in the digitalization of agriculture okay let me let me jump in a bit um from the uh sub-saharan african context so um the reality here is that women don't don't or are not given a voice when it comes to decision making even when you want to do trainings or good agri-practice trainings um they are usually left out which also means that um they are typically left out of most interventions um so first of all i think um in designing digitalization programs um there has to be a level of intention uh and deliberation like this is what we want to do to empower farmers in our instance um because of the fragmented nature of farmers we usually make sure that we put farmers in clusters so if you put farmers in clusters it's easier to use all all sorts of mechanization among them so one of the rules to put in place for instance is that for for a cluster to be formed it has to be made up of 40 percent women um because um each cluster has the potential of getting some sort of impute financing that is they get the imputes bonded with the services on credit and pay at the end of harvest it's an incentive for them to be willing to change behavior otherwise they would not so there has to be that level of intentionality uh knowing what these challenges are knowing some of these drawbacks whether religious or customary um constraints that limits women participation in this um one has to do that into the design then then the other thing is um trying to also understand and then work in the shoes of the people you reach so so in some in some parts of my country for instance um the the active religion um sort of um has rules about how you engage a man versus how you engage a woman so putting that into context will allow you to navigate some of those conflicting areas for instance if you have areas like that you you make sure that your agents and the people who are your last my touch points with the farmers can be people of the same gender so that way it gives the locales it allows them the flexibility to trust you to want to be part of what you are doing um as against trying to um go against them and you know um escalating potentials of conflict those are some of the ways that we've navigated some of these issues and how I think that um in doing such solutions it's important to take all these sorts of things into context thank you thank you very much Eneka does anyone else want to jump in jump in on this question yes so yeah I mean this is really hard an excellent question and it's really hard to serve um so one of the cases that I heard it's not our work in Kenya I saw a case when they wanted to bring more women farmers to these digital innovations and use more digital tools and when they provide training and capacity building they actually invited men but they asked them to bring your wife bring your family members and bring them all together to this event only then you can be eligible to participate in training so I think that goes back to Eneka's point that she really intentional really carefully designed to bring that out uh bringing women out of their household and empower them into the system and yeah this will require a lot of innovative thinking and experimental research thank you Jim yeah just to chime in very quickly I think um at least the experience on our platform is that um because it is about increasing market access and at the same time I think the challenges that we have with the aging farming population is that there sometimes isn't really that much of a choice right oftentimes it is the men who have left to go to the cities to work in construction jobs or other jobs so it's the women folk it's the elderly who are still manning the fields so I think what we see is actually to the extent that now you know e-commerce is lowered the barriers um you know some of these uh people still on the countryside they're actually better to uh they're better able to engage right with the sales because you know sometimes actually for things like say live streaming right talking to customers handling customer service those are things where actually um you know there there are some benefits right for for say women to to handle those kinds of roles as opposed to just a pure like in a farming defined kind of role where you do have to have a certain amount of manual inputs so I think what we've been trying to do is to ensure that you know our courses are available both online as well as offline so even in areas where maybe it's a bit more remote or the people are less educated um you know it's it's below the poverty line we're still able to sort of send in our uh teams together with the experts to deliver the training that they need yeah I wanted to just add one thing quickly one thing we are very interested in trying it out in your near future decision science project so as Jamie pointed there are lots of data needs so there is a continuous kind of feedback loop of more data becomes more kind of insight and improving the quality of this recommendation and things and yeah there there has been a lot of decision science movement especially in research area because of the lack of enough never enough trading data and things like that so we are trying to understand if the decision science project participating in those projects also empowers women we are trying to kind of develop kind of army of women being trained to collect data like odk and different types of data collection tools and we want to track whether they really get more empowered by participating in those projects and get their own kind of um you know skills and they get propagated to other types of digital making processes so nothing there are probably again the innovative opportunities here and there we can learn and trying to understand the impact and also yeah to share across the world we have another question for you Jawoo on the Q&A about like did you ultimately find that the inputs and yields decline with the distance and time from the market ah so when we took that into account the low adoption of cell phone in the rural areas yes and that was kind of expected pattern we wanted to see how quickly that degrade from the market and yeah so that that was uh yeah so we kind of hypothesized that that's the trend and then we confirmed it but yeah just again we had the anomalies because of that uh imperfect our sampling frame and understanding how much of the cell phone coverage was low lower than we expected actually it goes back to Jamie's one of the slides in cartoon the average is three feet but actually there's a big one we had the exact same situation in that country the average cell phone number of cell phone for household household was more than one it's a 1.2 or something but in reality it's a lot condensed and concentrated urban area so rural area it was really not the case and we didn't understand before doing the effect of time thank you and I have a question for you Xinyi following so we you presented your 10 billion initiative have you already thought about where you think we should prioritize the investment to bring more efficiencies in the agricultural systems yeah thanks Clara so I think one area that we're very interested in is around the midstream how to make the midstream more efficient so you know I think compared to western sort of developed countries or like the US for instance the coal chain capacity per capita in China is still actually relatively low so it's only about a third so that means that actually a lot of the fresh produce it doesn't get transported in a coal chain truck so that also leads to a very high food loss and waste freight along the way so one part of the solution is actually just to improve the utilization of the trucks as well so obviously there is a need to invest in the infrastructure and to ramp it up so we're also in talks with the third party logistics providers on how they can do that but I think it's also important to show them that through actually a better use of the data through forecasting for instance with the existing assets that you have you can already just make better use of them right so you directly have an impact on reducing food loss and waste while also ensuring that the logistics provider still makes more money at the end of it all right so that's going to be a very important incentive to make sure that you know the people who are providing the services also see a tangible benefit right not just purely from kind of the environmental perspective or you should do it because you know it's it's better for the environment or it's better for cutting out loss but you should do it because you can actually make more money right you can be more efficient you can save on fuel etc so that's very tangible to them so that's one area that we're actually been talking to the third party logistics providers for quite some time and I think on the upstream regarding agricultural technology that's really where we expect to be investing more as well because we've seen a lot of great technologies all around the world but they need some you know adapting right for the small holder context in China so that's also where we've been spending more of our time and also looking at potential research collaborations oh and if I could just share as well I did share in the answer to one of the questions in the chat an example of the research collaborations would be we just announced a project with Wageningen University in the Netherlands where we're actually studying how we can use different types of lighting as well as treatment of the growing solution to alter the nutritional value of the crops so I understand you know in some places the emphasis is still more on the yield but I think looking at controlled environment agriculture is that also grows to become a bigger part of our food system right especially for food security purposes or to locate it closer to urban centers as the cost of these technologies come down that's also a lot more that can be achieved by using them not just through improving yields so that's also an angle that we're looking at because we do see that amongst the consumers there are also consumers with preferences for you know higher quality or more nutritional products but it's not always the case that these need to be imported products they don't need to travel from thousands of miles away or be brand name products you could potentially through the use of technologies through LED lights for instance you know produce more nutritional products for the consumers thank you and in the first question that we got in the Q&A in the chat lots of them were about sharing knowledge between different regions country or developed countries and and developing countries do you know if this is something that is currently really happening if there is current discussion for example this year you are based in China do you know if you have some countries that would be interested in learning from what you've done to implement it in in their countries or maybe Jamie same thing for what you've been doing in Africa do you have some regions that well not regions but governments that maybe have contacted you to try and have the same thing in their region um maybe I'll go first um yeah I mean there's a lot of um there are the large agricultural research organizations um the CG centers some of you will know um who do coordinate a lot of this kind of activity particularly between um there's a lot between African governments uh so we've been talking to the the government of veranda um about this they probably within Africa have the largest um the most advanced digitization program for their farmers it's part of their subsidy system um but actually probably the best example of the kind of thing we're doing uh globally I'm aware of is in the Philippines where there's a um a research project which is called Bryce uh crop manager uh has been adopted by the government so this is something that came out of research has now been adopted at very large scale um millions of farmers using it and they are generating the kind of longitudinal data that you'd expect for that there are there are examples in India too but it's a bit more fragmented amongst the states I believe rather than being at a national level so yeah my sense is there's a lot of data sharing is probably harder than idea sharing um particularly a topic we get into a lot is ownership of data um that you know farmers need to own their data it's difficult I think um historically though there's a lot of data being generated without really taking that into account and I think people are now becoming much more savvy about it um that's becoming a bigger issue it's also just very difficult I mean obviously researchers they produce data to to produce research they're not necessarily incentivized to share and so naturally you know that's an extra thing for them to do and that becomes difficult um so um yeah it's one of those things I think there's no easy answer is always a lot of work yeah I can quickly comment on that uh when there was a food system summit last year there were a lot of excitement and idea sharing also uh kind of planning for the future um kind of convening just like our Clara you mentioned for the all the different parts a different part of the food system actors who use digital services and digital innovation to collaborate and kind of working together in some kind of pre-competitive space so there is something cgi also in the we are participating something called global coalition for digital food systems um each kind of commitment for action and I will share the link on the chat I think I haven't seen actual like activities planned yet but there is certainly a lot of energy and enthusiasm and kind of momentum there are built on since the food system summit so I hope something will happen and so there will be a venue to continue developing the collaboration and your knowledge sharing and yeah innovation sharing yeah I think on that note actually uh so I think actually I was on another session with Jawu also I think last year so also sharing about the Pintoto model so I think there has definitely been a lot of interest particularly in the wake of COVID I think people were kind of trying to rethink you know how food distribution um you know maybe should work so a lot of interest spoken to people in ASEAN also in Latin America you know all with a lot of interest in terms of how they can replicate of course in the US as well so just yesterday I was sharing with a group of students in the US who are also asking oh why isn't there such a system in the US for farmers to sell their produce directly to consumers so I think um you know for us typically what we would emphasize is that a lot of it does depend on the local conditions so I think we're you know very lucky that when we start out in 2015 in China the logistics infrastructure had already been pretty well developed there was a lot of investments so you're actually able to have a pretty reliable and low-cost logistics service which may not be possible for other parts of the world today right so you do need some of that hard physical asset to be infrastructure to be built out and to achieve a certain level of reliability so that it then becomes economically viable to be shipping you know boxes of fruit right and ensuring that the box of fruit arrives in a given time at the consumer's place in a decent condition right so a lot of that unfortunately does require a lot of investments on the part of governments and maybe some corporates as well and then also the issue of digital connectivity that's also one as well so when we have started a lot of the consumers in China as well as the farmers they at least had you know a pretty basic smartphone right so low-cost smartphone where they can chat with their friends they can also exchange money or they can you know spend some money online and so they were looking to do that they had the tools to do it and so we kind of you know stepped into to fill some of that need for shopping for easy communication and sharing with friends etc in a fun and interactive way so I think there are a number of things that are important to have in place to facilitate such a system so you know we're very happy to share about our experience with other countries or other institutions that are interested but we also are very aware that you know there are a unique set of sort of conditions that do need to be in place for something like that to take off and succeed okay um let me jump in a bit um I think not just the sharing of data and knowledge is absolutely important especially if we realize that the problem we have ahead of us is is not one that can be solved by one actor and it's actually a global problem you know the problem of food insecurity is one that affects everyone one way or the other regardless of how how countries think they are far removed from challenges happening around the world is so connected now however you know um one challenge here especially from the sub-saharan african context is that um we have almost little to fall back on like you have almost no data at all so um in in in some other parts of the world so somehow you have maybe some sort of publicly available data that you can then build upon all happens in this context is you are almost doing everything from scratch the reality is that data is very very expensive collecting data is very very expensive now I can I'm saying this from the context of um you know uh startup players in the sub-saharan african space trying to bring a context as to why I understand why sometimes it looks like it's difficult to share share data and that's because um they've taken time to collect this data to develop this data they feel like this data should serve them be an advantage to them but I think that um one way um that innovators can go around this is that first and foremost we must come to the point where we begin to see some of this data is it's actually for public good right and work together to collaborate strategic partnerships makes it easy so that uh when you are doing something in an area where someone else has done that you don't have to completely reinvent the wheel but especially somewhere everyone keeps operating in silos it actually makes it much more difficult to scale because if I'm if I'm if I'm focused on having to do everything collect data in one area build upon that keep growing it would take me a long time that if I had some form of data sharing with someone um that I can build upon um it would be really great if innovators like myself in sub-saharan african um can learn how to collaborate even while they compete thank you amika I I have a follow up question for you and for for everyone so with the light we've put on the contribution of agriculture and the food system to the global gg emissions and um the fact that we the the international community could do something to try and and put more weights on the innovation to try and reduce emission and increase mitigation do you think that um it could be um a good way to to manage for example uh access to more data or or to to get more data because it would be justified by the fact that it would increase mitigation and it would get everyone closer to the 2030 sustainable agenda do you think this is something that could drive a movement to to get uh all the things that we really need to to to have um uh to reach mass adoption of innovation everywhere in the world to to use the tools like yours and everyone that has presented today yeah I can start and I I really believe yeah that's really kind of missing opportunity that that we can we can leverage and exploit here I did my graduate study in northern Ghana upper west region of Ghana it's really degraded uh sandy soil and low input system and what I did there was using remote sensing and different kind of crop modeling approach to estimate how much carbon sequestration carbon kind of mitigation can happen in that area even under that really low input system if farmers change their medium practice it and and I did that also in certain parts of Ghana and to be uh in comparative with each other and and certain parts of Ghana soil is more fertile it has more biomass in it the farmers are more well uh well aware of the soil condition and how to manage soil and things like that quite different but um but quite surprisingly we saw more potential possible uh in northern Ghana when soil is really degraded because of low degradation there is actually more opportunity to bring more soil carbon into the pool when they manage soil is better uh um you know in right way so similarly I think across Africa whenever there is a low input and very degraded soil condition it's really hard to manage in right amount of fertilizer and right amount of practices I think that could be easily not not easily that could one day be turned into opportunity for bringing more carbon into the soil and such that then might even turn into carbon credit they can use in other types of infrastructure investment yeah we talked about connectivity issue we ourselves are not not able to really absorb the problem but in the collectively that's a big another innovative revenue stream for countries to realize or if this could be really systematically managed so yeah again so I mean there's a huge potential to manage soil is better in low input farming systems and I think that would be really win-win solution that we all need to be yeah look after thank you very much Javu um if no one from the panel has anything to add I think I will have to give some closing remarks because we're getting closer to the hour um so thank you everyone for participating in today's webinar and we've had four presenters giving presentation on their different innovation and digital tools that they may be using across the globe and we had Javu giving that gave a more global presentation and the thing that I really take from this webinar is that we really need to to focus on giving access to learning materials and to allow marginalized population to really have access to all these innovation tools because they are at the places where they are producing most of our agriculture food for the small elders and other communities and yeah so I really want to thank you for this really interesting webinar I want to remind you that the the recording will be made available on the SDSN website and that we will also share the the presentations because most a lot of people I've asked for them and so I want to thank you again and I'll let you get back to everyone's work thank you