 Great, many thanks to everyone for joining today. I think we'll kick things off. So first of all, thank you very much for joining this webinar, which will be a closeout for the data-driven electrification program that the IA has been doing in collaboration with Power Africa and USAID, as well as our partners at MIT. So while today, we will be mostly focusing on a new tool that is, if I do say so, quite cool to try and figure out the best ways to use GIS to map which regions of Africa remain unelectrified and what the likely demand of those buildings will be, I also want to give you a little bit of context in the broader program and what else we have done in the context of this. So this program has been a multi-year program that is really focused on how do we get better data around electrification. And I think that there's already a great foundation that we've been building on with really great survey work that has helped lay the foundations there. But we know that there are limits to that process, that those surveys are expensive and take five years between different surveys to be able to sort of re-up that data, and that there's an increasing focus on how to move faster and new technologies like off-grid systems that are really changing the equation on how we do this. And so this piece of work that we put together was really thinking about what are the different levers we have to give countries and investors and the development community a better sense of how things are changing in real time and being able to build these data sets with more granularity down to the household and building level to really understand and gauge progress. So I always like to say data work is frequently the work that is not necessarily the most glamorous. People don't love to fund it. People want to be thinking about projects and investment, but I do really view it as taking the vitamins or eating the vegetables that we need to to make sure everything else is functioning. And oftentimes some people say, well, data analysis is important, but it's far afield from making investments happen. This is a notion I think we really want to push back on and also in the context of this initiative, we are seeing that really the provision of data and the right types of data are absolutely essential to making or breaking some of these projects in terms of taking them over to bankability. And what I mean by that in particular is when financial deals are done, the first thing that different investors do is they go out and they look for benchmarks. When different people are preparing financial plans for projects or expansions, they're trying to make sure that they're using credible data, incredible assumptions that have been backed by rigorous science and authorities that have vetted this data to make sure that they are being able to make a strong case that this is a credible project, incredible financial projections. So in many ways, having these benchmarks I think is absolutely essential to being able to make the business cases that the investment community feels competent in moving forward. So while this is sort of a stretch where there's many different pieces of the supply chain to get to there, I do want to draw that connection because I think it's really important for when people think about how this data is relevant to the real world and how these data foundations really help support the movement of more money to this critical and use of electrification in Africa. So what that is is sort of a big picture thing. I just want to say concretely what we've done in this. So one, the big focus first and foremost was to upgrade and provide a new guidance on how do we use supply side data sources in Africa to track progress on electricity access. So this is really saying each year how to make sure our utilities by region are providing information and data, net sense. This is how many connections we provided in what regions and what types of connections and ideally geodocating those as well. But also how do we work with mini-grid developers and off-grid systems and using their sales data and their permitting data to make sure we're getting a stronger base there. We did a ton of work with a lot of different African statistics divisions standing up this new supply side methodology and survey approach. And really we're already seeing reaping the benefits of that where people are getting more granular data, more detailed and timely data on this. And for the first time, the IEA was really able to actually parse out and provide an estimate at the continent-wide level on new connections each year by type of connection. And what we saw was that the revolution in solar and batteries and new business models around off-grid and solar systems meant that last year or actually in 2022 that half of the new connections were from off-grid systems in Africa. And this has been a big pivot where grid has really been dominating things historically, but we're seeing with the debt crisis and the slowdown in sort of utility revenues, we're seeing that model being facing strains and without sort of additional concessional support and that we might be shifting to a mode where we need to rely more on these off-grid approaches. So I think that's one area where we've seen this data work actually be able to produce really important outcomes that are really resonating at the top level when people are thinking about their strategies. And the second piece on this as well, is really trying to make sure we're working with different African stakeholders and using the GIS tools that are already out there. And today we've just launched a repository of all the GIS data sets and tools and ongoing projects in different African countries related to electrification that can be used. And so I really encourage practitioners to explore and use these and explore this data set and figure out what tools and data sets may be available in your country and can be used to greater effect. The last thing as well as we just had a, we ran a pretty comprehensive bootcamp training on some of these GIS foundations with our partners at Climate Compatible Growth, along with the World Bank, WRI and SE4ALL. And this was really a great way to make sure we got everyone in the room understanding what are the tools available and how do I, as a local practitioner in government, say what are the data sets I support? How do I stand those up? And how do I make sure that we are putting the best tools available to work for us in our electrification planning? And I think that brings us to the new tool today, which my colleague Darlin and our colleague Stephen from MIT will go through. But really this tool I think provides a first of its kind ability to one, at the building level really understand what buildings likely have electricity and which we don't. The number of times we've gone through this mapping in different countries where we've seen a community that we know with 80% certainty they have electricity today and 400 meters down the road a smaller outcropping that has no electricity access. These are very low hanging fruit ways in which companies can say grid companies, utilities, off-grid solutions can say these are easy customers to electrify. And I instead of having to go village to village to make this assessment, I can now sit at my computer at my desk, put together a customer acquisition strategy and plan and this will greatly reduce the cost of me finding these new customers. Additionally with this, we use they use the latest machine learning and AI algorithms to look at millions of different photos of these different buildings and look at their surroundings. And these clock on to small things like the pavement of roads, the density of buildings and all of these different signs that may signal their access to commercial markets, their ability to access additional employment that could help raise incomes. And this also comes a very important way to be able to target which buildings are gonna have the highest demand and also the higher ability to pay for higher amounts of electricity. So this really collectively, I think strengthens the business case for many companies to be able to use this tool, identify priority communities and then also think about with new business models with the right incentives and government backing how do we find ways to broach this challenge in further field communities as well. So I will leave that there, but just again thank you to our partners in Power Africa for laying this foundation. I think you'll see today as we go through this that this has not just applications to utilities and Oxford providers but it's gonna be useful for funders, philanthropies, local statisticians, people who are running these planning operations in governments. This data set I think is gonna be massively helpful for really taking electrification planning to the next level and making sure that, and the last thing I think I'll say here is the IEA stands ready along with our partners to work with other countries to extend this data set to your countries as well because we really do think this can be an instrumental move to really jumpstart the electrification process here. So with that, thank you again for joining and I'll hand it over to my colleague Darla. Perfect, thank you very much Dan. I think you should be able to see now the presentation chair. Good afternoon to everybody. My name is Darlan. I'm also in the IEA and of course I'm part of Dan's team and together with MIT and with the support of Power Africa, which is also together represented here. We've been developing this and implementing this project called Data Revenitification in Africa. I want to walk you through the rationale of this initiative and the agenda for today will be developed as follows. So we will start with overview of the status of access to electricity in Africa. We will then present a new data tool which is available right now on our website which is called GIS Catalog for Energy Planning in Africa. And then we will jump together with our colleague from MIT on the new model that we are also releasing today. And then of course there will be space for conclusions and Q&A that I invite you to post in the Q&A function of the Zoom webinar so that we can collect them and address them by the end of the webinar. So the first point as I said on the status as you can see, as I think this is non-use for all of you, we stand at a moment in which the population of Africa has been increasing quite a lot, I would say, in the past years and is currently the region which has the highest growth rate in terms of population year. You can see how was the trend in the past 23 years so from the beginning of the 21st century. And you can see how Africa's population as a total has almost doubled. This has created a situation in which access expansion has struggled to pace with this growing population. As you can see here with this aggregation in terms of access to electricity, you can see how there were huge improvement in terms of access to electricity because the population having access to electricity has been growing a lot, but you can see how the total number of people not having access to electricity across the continent as you didn't change so much and as actually increased by 100 million people in the last 23 years. There has been different trends across these two decades. First, we've seen from 2023, sorry, from 2013, the population not having access to electricity decrease even if not at the required rate, I would say, but it has been on a declining trend. What has happened is that from 2020, this trend has inverted for the first time since 2013 and of course as you might guess, this is a result of the impact of the pandemic and for the first time, the population in Africa not having access to electricity has been increasing. Last year, so in 2023, we've seen for the first time after the pandemic, this trend having an improvement even if modest, which again, despite being an improvement, is remain significantly below the progress required to achieve SDG 7. So as I said earlier, in terms of numbers, in the past 23 years, we've had 450 million people, so almost half a billion people that were able to gain access to electricity. So this is a huge improvement and is a testament of how good stakeholders in the continent have been working towards SDG 7, but we've also had, and as said earlier, one of the reasons is of course, this huge population increase, 100 million people more that don't have access to electricity. If we go down and we break down these numbers, depending on the type of region, we can see how people that today don't have access to electricity are four, like four out of five people that don't have access to electricity right now live in rural areas. This, if we go deeper and we also see the different trends in the different type of regions, we can see how in urban regions, electrification rates, as I was mentioning earlier, have surged significantly, but conversely, as you can see on the right hand side, the rural communities that most of the time, like absent only modest enhancements in electricity access, this of course has entered economic challenges and is related to the complexity of extending the grid to low density areas that are most of the time also partially populated and also have limited demand and purchase power. Additionally, the off-grid sector, which of course includes solutions like mini grids and solar arm systems, which would serve as viable alternatives in these kind of scenarios, have also faced their own set of challenges. And of course, I mean, this might not be new to the audience, but there are many reasons for this. The main are of course, unclear regulatory frameworks, and one that, as IEA and as Stan was anticipating earlier, the lack of clear information and reliable information for areas for which of course it's very difficult for countries and statistics office to have very detailed timely and also granular information on which the private sector in this case, when we talk about off-grid solutions can really develop strong business cases. As anticipated by Dan, again, off-grid systems are becoming increasingly crucial. Here you can see in the past years, at least the order of magnitude of the number of people that were gaining access to electricity per year in the continent. And we can see how this value is more or less between 20 and 30 million per year. But if we go, and this is really thanks to the support from our Africa that we're able to have such granularity in our last estimates, we can see how off-grid systems and in particular solar systems are becoming, as I said, increasingly crucial. And in 2022, alone solar systems, so both the smaller ones and the larger ones contributed to more than half of the access increases in sub-Saharan Africa to conclude this part. And as mentioned, this was like the improvements in terms of the granularity of the data that we are able to produce and also publish for the general public were done thanks to the contribution and the support from our Africa. The IEA has been the first, actually, to produce global database of electricity access information since the beginning of the century and is also one of the co-custodians of SDG7. So here, this is a portal that is available openly from our website in which you have the desegregation and the data and trends for both access to electricity and access to clean cooking. And as also mentioned earlier, we've been producing through this program this guidebook for improved electricity access statistics, which is a step-by-step guide to develop access to electricity indicators using supply-side data. Through this work, we've been interacting with and working actively with the statistics office in all the countries in the continent in order for them to gain the IEA expertise in terms of statistics and being able to report on a yearly basis with the highest granularity, at least for the time being possible, in terms of the access to electricity data. So while performing this, of course, this is the access part. We've highlighted how crucial it is to work on planning, especially on the electricity access side to attain SDG7 by the end of this decade, given that right now we are not on track to do this. So for this, we've also, we are also releasing today this new product that is the GIS catalog for energy planning in Africa. But for people that might not be familiar with what GIS is, which stands for Geographic Information System. A geographic information system is a technology that captures, stores, analyze and also present data with a very specific granularity, which is their position on the Earth's surface. So here, just also to flag this presentation, we'll be shared with all the attendees later on. So you will also have the possibility to reread it. But the key component of GIS are data input, data management, data analysis, and visualization. So this is the same for all kinds of data. This is nothing new. The new and the innovation that's in the peculiarity of GIS data and is that this data, often big data also have a geographical component and geographical attributes. So the IEA, as I said and through this program and in general through its work has conducted an extensive evaluation to understand the utilization of GIS in shaping electricity access strategies across the continent. These assessment involved interactions in that sometimes, interactions with the approach spectrum of stakeholders, including ministries, utilities, ratification agencies, and also partners in the private sector. And here you have a couple of the results of these interactions and also this aggregates some of the results from our interviews and surveys and also information from our partners from the World Bank through their Global Policy Support Platform, which is the rise of regulatory indicators for sustainable energy database. These examples, I think it's pretty obvious how GIS is already used a lot across the continent in planning and expanding electricity access and it demonstrates the critical role of geospatial analysis in helping the continent meeting its electrification goals already today and hopefully in the future. So during this assessment we've seen, of course, each of these countries have their own history of using GIS for planning with different scopes. But what we've seen is a lack of coordination, sometimes awareness of the available datasets and models and also sometimes lack of coordination in terms of which are the responsibilities in terms of like between the different stakeholders within a country. And for this, we've been basically collecting and doing like a comprehensive review of what are the available datasets in the sector right now for a GIS-based planning, what are the existing models and tools that can be used to perform GIS-based access planning and also a list of the institutions per country that work currently or are starting right now to integrate GIS within their operation for, of course, energy planning. So all of this is now contained in the IEA GIS catalog for energy planning that we are also releasing today, which is in our idea, I would say a one-stop shop and like a resource for GIS-driven planning for professionals across the continent or that work with the continent to have a comprehensive overview of what is available in the space right now in the sector right now to do GIS-based planning. So again, as I said, we will be sharing right after this webinar or the links and the material. This is a screenshot from our website, but as you can see this catalog right now has three sections, one for datasets, one for models and one per country. And you can see how you can filter all the datasets available by type of data. You have the possibility to, as I said, to see the models too and decide that can filter them by type of model you're looking for. And of course, also by country, we have a quick overview of what's the status of the GIS integration within their operation. And you also have an overview of what are the relevance stakeholders in each of these countries. And the possibility, of course, to also filter again those datasets and models that are present in the first two sections, specifically on the country you're interested in to see if there are datasets that are country-specific or models that are country-specific for the country you're interested in. And also, in case you... There are also datasets and models that are country-wide that, of course, can be applied to country-specific contexts and analysis. So this is a quick overview. Again, you will get all the links to access those resources. But as you've seen also in the model sections, there are many... Again, I went through all the participants of this webinar. And I know many of you are professionals in the GIS phase. So I'm pretty sure most of these won't be used. But I wanted to go through a bit some of the kind of applications that GIS is currently with which GIS is currently used across the continent to do planning. And of course, one of these is least cost electrification. So for people that might not know, what we do is overlay different type of GIS layers and basically with different resolutions. But let's say the idea is with as much as possible granularity being able to decide for different settlements, different contexts. So with this very context-specific approach to assess what's the best, so the least cost electrification technology among on-grid, off-grid, and sorry, among on-grid, mini-grid and central systems. And this is of course based on different types of layers of considerations like the distance from the existing transmission and distribution grids. There are resources available in terms of renewables and of course the energy demand in the location. This is an example from our Africa Energy Outdoor 22 in which we perform such an analysis at the continent level. And you can see how this resulted in this chart in which we can see very quickly that to achieve SG7 by 2023, these are the, this is the desegregation in terms of type of technologies that we think are the most suitable for each of these contexts, for each context in the continent that currently don't have access to electricity. And from this we can find out that at least 70% of people can gain access to electricity for the first time through renewable energy. One thing that specifically we decided to address with this project, and again, of course, this is one of the components that you will see in the Africa GS catalog. And but again, if many of you, as I know, are working on GS-based energy planning, they will know that demand, the demand estimation in an area really influences the result of at least cost electrification model. Why? Because if we have different demands, of course, the technical economic best like optimal solution from a technical and economic point of view will differ depending on what the local demand. So here you have an example, for instance, of three results in terms of least cost electrification, depending on different scenarios in terms of demand, low, medium, and high. Even if we don't go through the details here, you can see how in terms of like visually how the results of GS-based least cost electrification tools differ depending on how much demand we are anticipating to provide to each of the settlements and actually the population that live in this region. And imagine when we do it at the continent level. So the reason, and this is one point, so this is the importance of the demand part. The second thing is that during this assessment of available datasets models, et cetera, what we notice is that there is a lack of available datasets out there and models actually that allow detailed demand estimation. And that's why we decided to partner with MIT to address the specific gap in the sector. So I will leave now the floor to Steven, our colleague from MIT to walk us through what is the building level electricity access and demand estimation model. Up to you, Steven. Thanks, Darwin. It says I can't share because you're sharing. Yes, you should be able to do it now. Fantastic. Thank you all for your time today. I'm here to introduce our electricity access and demand layers and also our platform, Open Energy Maps. What we're doing is developing an open source state-of-the-art machine learning system that can scale to provide what we really think of as the most comprehensive and granular view of electricity access and demand possible. This is really aimed at aiding governments, utilities, planners, researchers, and so on. And we believe that we're really addressing a pretty significant set of data gaps. This data could be used, for instance, in planning. And as Darwin mentioned, how to design systems and choose different electrification supply technologies. Today, we're gonna start with three countries, Uganda, Ghana, and Senegal. But as we'll describe, actually all the remote sensing features that we use are actually available across the African continent and elsewhere across the globe, actually, so we can extend to other regions. So this is our website, openenergymaps.org. It's officially live now. And you can click Explore. And I come right to our kind of choose a maps page. You'll need to create an account and log in for your first time visiting the platform. And we can kind of move into the maps and see a description of what our platform is currently showing, but also some documentation that you can click on and read up on really what's going on under the hood here. And we wanna actually, I was thinking to start with our outputs very quickly. So we have different layers on the left panel. We have an electricity access layer and a demand layer. And let's just look at the access layer for now. So you can see we have a lot of dots covering Uganda currently. And as we zoom in, this will resolve and we end up seeing individual buildings and building footprints. What you're seeing right now is just the output of our models. We have all of our different building shapes covering the whole country. What we'll describe later on where that comes from. But what you're seeing here are different colors. And this is really sort of a heat map where the yellow color corresponds to higher likelihood of electricity access and more orange and red and purple colors correspond to lower likelihoods of access. And if we click on a given building, you can see actually a distribution that's being estimated for whether or not that building has access. This particular building has something like we think a 44% likelihood of having access. You can click that off and move to the demand layer. And actually we see the same buildings. But what we're seeing here is a different color scale. It's also a heat map. But for all the different buildings in our geography, we have estimated electricity demand in terms of kilowatt hours per month. And this rainbow color scale where blue implies lower consumption, lower demand and yellows, oranges and reds imply more and more demand. And as you can see, we can pan across our region of interest and see really the diversity of demand that's being estimated across this sort of spatial scale. So I'm gonna start by talking about these buildings. These have been really a game changer for us in how we've been able to do our research in the last couple of years. So these are just the buildings themselves without any layers added on. And they come from Google and Microsoft and they actually have computer vision models that scan all of Google Maps and Bing Maps imagery. And they actually outline the building, all the different buildings. And what we're doing is we're actually taking these two independent data sets and merging them together such that we have really what we think of as a very complete view of what should comprise most buildings in our countries of interest. And they're important because, well, we're estimating demand and access at the building level but also they provide what's called the primary key where we can actually relate other features of interest to these buildings. And you'll see each building has its own unique identifier that is that primary key with its own lat long description. So actually our two first features of interest that we compute are building density and building rooftop area. And the rooftop area comes directly from the data set itself. It's some simple geometry to compute that area. The building density is a little bit more complicated but really what it is is we draw a radius around a building and we count all of the nearby buildings inside of this local area. And this gives you a sense for the urbanization of a given building. And we actually have these calculations and all the features replicated for every building across all of our countries of interest. And what we do is so in addition to these density and rooftop area features, we add additional features and one key feature as has been introduced is high resolution satellite imagery. And this is now very globally available. What we do is we actually take a clip of a large mosaic imagery and we take really a tile that's directly overhead of our building of interest. And this will tell you a lot of things about, well, what's the local area look like? What does the building itself look like? It's rooftop, it's local roads and so on and so forth. And there's a lot of really rich information in that imagery itself. We actually also use another type of satellite imagery. This is nighttime lights imagery and our platform lets us zoom out and see Kampala and these different irradiance readings that are taken at night. And you can click around and see the different values that have been obtained. And this gives you a sense for street lights. And you might think the more brighter the area is, likely the more economic activity is happening locally and likely the higher the demand is. Our next feature is actually a layer called internet speeds. And what this is is an aggregation of different speed tests that people do when they're testing their internet connection and how fast it is. And one of the companies that does this Ucla actually aggregates this information and makes it publicly available at these grid cell levels of aggregation. And you can click on one of these grid cells and you can see things like average download speed, upload speed, so on and so forth. And the number of devices that have tests. And this is also a correlate for electricity demand and access and maybe access is pretty straightforward. If you have fixed internet connections, it's very likely you have electricity, it's required. But actually we have these layers for mobile as well. And also the idea is that you might think the higher the internet speed is and the more people who are testing for internet speed, the more affordability there is in that local region for higher demand as well. And lastly, we don't have it visualized here but we have this comprehensive land use dataset that we're using. So we're gonna go back to these buildings and it is that we really have a very rich description of each of these buildings. And what we do is we then map all of these buildings to whatever utility data, metered utility data that we can collect. And we have a lot of people to thank for that access to that data through the IEA and other contacts. And what we do and what our model does is try to learn relationships between all the different features that we talked about, the internet speeds, the nighttime lights, the satellite imagery and map these relationships to ultimately observed metered consumption data. And this is really where we think and we use the power of machine learning to learn a model that's capable of finding these relationships and that's also capable of being extended elsewhere to regions that actually we don't have that metered data where maybe there's no connections at all or maybe there's electricity connections but there's no metered data present or even in some cases where maybe someone in the world has that metered data but we don't as planners, we can still estimate that consumption and that access. And in the end what we get are the different layers that we talked about, the access layer here and this is in Kampala very electrified so that's not as interesting but we have the demand layer as well. And we have for all of these different buildings now what we think of as really key inputs to planning methodologies. So with this platform we make all of our different data sets available for download and the current view is hosted already and what we actually have is grid cells for all of our regions of interest and you can click on one of these grid cells, find a download link and this is a fairly dense area it might take a little while but you can see a GeoJSON file is being downloaded that's a pretty universal file type that can be used in many different GIS softwares and programming languages, et cetera. Now there's a couple of limitations to this work this is a really hard problem from a machine learning perspective to solve. Really there's an inherent lack of information it's hard to know with very high accuracy what's going on inside of a given building or household given just what we see from outside and what we're seeing from remote sensing features really you can't see what's happening inside of the households themselves and there's actually a lot of variability even in buildings that look identical to one another that might have the same exact remote characteristics. So what we do is we actually treat this problem probabilistically we say well we're not gonna say we know exactly the answer but we know maybe what the distribution for a given building may be given all the other observations of metered consumption that we've seen previously and that's what we're providing here we're providing this distribution and the way that this is read is that actually there's some likelihood that an entire range of consumption values may occur at this building or a range of demand values may characterize the building but maybe we can provide some one mean estimate and use that if we were to just give one value. Another challenge is that it's actually really difficult to generalize so we have a lot of our ground truth metered consumption data in East Africa and it's challenging to think well how well can you do in other areas if you don't actually have the ability to stick and have ground truth in other areas of the continent and lastly we have challenges due to class imbalance where you can imagine there's actually many, many more buildings that may look like this these lower demand buildings maybe they're rural or more suburban and we have relatively fewer very large consuming buildings and this just makes training machine learning models very challenging and is a source for error as well. So to conclude our machine learning model is able to learn relationships between we think of as a very rich characterization for most buildings on the continent and in the world and what we're actually looking for is to grow this platform we've spent a couple of years building this but we think it's also just the start of likely much more to come and we're looking for partnerships and resources and specifically high resolution data and that can help us really improve our results and otherwise thank you for tuning in I'd like to give thanks to the IEA Power Africa our collaborators at UMass and EGuide and of course our utility and government partners without whom we couldn't have built any of these systems. Thanks very much back to you darling Thank you Stephen let me share my screen again just to conclude So thank you very much Stephen I hope and I think actually it was very clear and just wanted to recap a bit the overall structure of the model and then we can finish with the conclusion so just to be like clear and summarizing you might know again nothing new for most of you I think that we might have three types of situations in Africa in a country we have buildings that don't have access to electricity we have buildings that have access to electricity but no meter or maybe they have a meter but it's not geolocated so we are not actually able to characterize those buildings in terms of consumption and then we have electricity access like building with access and meter data and fully geologated what we've done basically has been since we were able to collect with the partner utilities from the three pilot countries and partner countries that were in this project that were Ghana, Senegal and Uganda we are able to collect those information in terms of months so we know actually what's the demand of some of these buildings as mentioned by Stephen we have the building footprints for most of the buildings in those countries and what we've done has been basically to train this algorithm in characterizing each of these buildings based on this layer that Stephen has been describing to solve this problem which is how much is electricity consumption both for areas that don't have access to electricity so in this case we're talking about the needs for those buildings even if they don't have access to electricity right now in the future in case we had to electrify them so this helps the planning phase as you might guess but also for areas in which we have electricity but we don't have meter data so that we are able also to support utilities in understanding and characterizing even better their customer base and on a very granular in terms of geographical so going to the conclusions there are many potential applications as you might guess by this just wanted to go through some of these of course for decision makers I don't know something in my life decision makers funders in like doing decision making what we want to do is being able to enhance local planning capabilities so providing granular data to empower local statistic divisions and planning teams we want to support project development and project developers of course so that they can have more detailed data for business planning and then last but not least but this is where I will stop here in terms of policy and planning to the integration of course this is an input data most of the time for many of the tools that are available and the great tools that are available right now in the space so what we want is provide a new data set that could be integrated through these tools to have better understanding of where access to electricity like where access to electricity is and how much is the demand of those areas that lack access to electricity today and what's the consumption of those area that currently have access to electricity I would be a bit quick here just because in the interest of time and want to leave space for a Q&A but as many of you are GS professionals you might guess what are kind of GS analysis that you could do with this having for instance settlements polygon and aggregating those data from the building level to the cluster and settlement level so being able for the community level to say how many people have or don't have access to electricity and also characterize buildings and households and productive uses by type of level of electricity consumption what we've done in the case of the three countries that were partnered in this program was to apply this methodology for the entire country and here you can see the results and one of the most striking outcomes that we came out with is that areas that already have access to electricity are actually consuming less than they could most of the time the reasons are two one is affordability and the second one is reliability so the second message that we want to to convene also from this analysis is that it's really important yes to focus on providing access to electricity to unoccupied areas but it's also really important for utilities to focus on areas that currently already have access to electricity but might have either as I said affordability issues sorry affordability issues and reliability issues to really reinforce the grid and make sure that people that could already have the possibility to use this electricity also from for productive use could actually do it and could have 24-7 availability and of course afford that service so to conclude the target stakeholders for us here you have them here but it's all of you so government agencies, utilities, NGOs, researchers and local governments and of course also investment and financial institutions be able to prepare better and make better decision and better investments in terms of next steps we have two things the first thing and is the immediate one is the integration as I said of this model with the existing jazz based models and tools and collaboration with partners so we welcome the possibility to partner with many of you to actually apply this model and these results and integrate those into your models your analysis as Steven said all the data and the model actually will be openly available for download then you can already use them but feel free to contact us also in case you would be interested in further insights on those and the second thing and it's a bit more of the long term is the geographical scaling and the model enhancement so on one side as said, stated by Steven there is a lot of data in terms of input for the model to be trained and being able to differentiate between different areas and characterize them and having really lot of context specific results but this means that we need data from activities that are a comprehensive sample of all the type of geographies and specificities that we have within a country and of course across different countries and of course we will be also working together with our colleagues from MIT from with Steven in particular with the model enhancement because as you might guess and I'm pretty sure you will guess there are many features that we could include these starting from for example the type of buildings that somehow is embedded right now but could be also an input to the model itself so this is it, thank you very much for your attention and I think we go through some of the questions we have in our Q&A box right now and then we will conclude with some remarks also from our sponsor and partner from Power Africa thank you very much so feel free to put your questions in the Q&A box that we go through them right now and maybe answer some of them and also defer to Steven some of these but as I said after this webinar we will share all the materials with the link etc and in case there are some questions we won't be able to answer we will share them with the like share them and ask where we do so I see many of these questions actually were posted during the presentation and were basically answered one thing I see here is in terms of question which is related to the possibility to include new tools and models and datasets in the GS data catalog yes so feel free to send us an email to suggest if you have any dataset and model that you think is relevant here we will be more than happy to evaluate that and to include it in the GS catalog we will do it right now like right after the publication with a short-term update to include what we might have missed but then our goal is to have probably a six-month update of this data of this catalog to make sure that new initiatives, new models and new datasets are included in it yeah I think so looking at all the questions I think actually most of them have been answered by either myself or Steven and also in the interest of time I would prefer to leave the floor to Samsung to close with some remarks from Power Africa and then as I said we will collect all the questions answered to them and share them with the audience and also abroad their audience thank you very much it was a pleasure and Samsung over to you thank you thank you very much Darlene and thanks to everyone that admit time to attend this session on this work we've done together with the IEA as Power Africa our goal is to increase generation capacity in Sub-Saharan Africa as well as bring more connections and to increase generation you really need to understand what the demand profile looks like what's the existing infrastructure look like and how do you cry out proud in the necessary stakeholders to to build out and invest in this asset and this data-driven electrification program in Africa that we've supported with the IEA has really demonstrated the capabilities of the new tools that are available and as Darlene said there's nothing hidden under the sun and so far the sun can see it there's some level of details that could be just partially derived a popular code from the energy foregold says there's no industrialization that is not high that is low on energy consumption so and you could see some of this work has shown that a lot of the people that are getting connected are still on the low tiers of consumptions and how do we move them up as Power Africa we cannot do it alone we can only do it increase generation increase connection and also increase demand productively together with our partners and we are proud to have supported this work and thanks to the IEA and the stakeholders of MIT for doing this detailed level of work it's been a couple of years we've been working on this and I know we still have sort of many questions to answer but I think at this point we feel confident that this could go out in public and more importantly also to capacitate people in countries people that are making this decision making this data available and bringing the necessary capacity to do this modeling building of this data and really making insightful decision at the local level on where the right electrification should go the method technologies and the kind of access that that is desirable and building this over a couple of years with that I will stop not only myself but also have Pamela Chaukar together with me on this program and the whole Power Africa team thank you very much to the IEA and our stakeholders and that will be it from me over to you Dalit Thank you very much Samson so we are over with our time so I will use this last minute to thank first of all Power Africa for the support all the stakeholders with which we will be working through this program in particular from the three partner countries we had which were Ghana, Senegal and Uganda thank you for to MIT also of course for the support especially in the development of this building level electricity access and demand estimation model we will be sharing all the materials the presentation the links to all these pieces of work that we've been mentioning together with the documentation on how to use the different tools and thank you very much it was a pleasure and all great days thank you