 Yeah, thank you. We're going to actually present Cognicity, but first we're going to introduce that or present it through Pedda Ben-Chana, which is the program that it's being used for now, and kind of walk you through that. And I'm going to give you a little bit of background about what Pedda Ben-Chana is and how it works and what it is that this crazy collection of people do together. Before I turn it over to Nachi Matani, who is one of our project designers, to talk a bit more about how Cognicity works specifically. So Pedda Ben-Chana is a publicly accessible and publicly generated online mapping tool for managing disasters in Indonesia. We have mostly been working on flooding, because it's a major problem in Indonesia and Jakarta specifically. But we can't handle other disasters, they're just self-sorting, you know. So this collection, we have people in our team that are, that kind of run from philosophers to architects to geospatial geniuses to computer science, urban planning, kind of the whole of it. And when we all got to Jakarta a few years ago, we were all sort of looking at this region in terms of the fact that it holds more than half of the world's population within this circle. And that 14 of these major megacities are sitting in River Delta's and 18 have experienced major flooding in the past decade. And that affects nearly 300 million people. And what can we sort of do about that? That might be a novel approach to a problem that a lot of people are sort of working on, and using computer platforms to do so. So in 2013 we sort of all found ourselves in this flood situation in Jakarta, which believe me was much worse on the ground than it looks in this photo. And also knowing that these floods at this scale have been growing more consistent and much worse since 1996, even though they haven't even learned our history. But when you're on the ground there, you understand that while the government doesn't have like an amazing capacity to deal with this problem, though they're getting a bit better, the people on the ground do. They've been living in this for a very long time, and they have ways that they communicate with each other, that they know how to respond, they know which areas that need to be evacuated and how quickly that all of that needs to happen. And as we're sort of tromping through the city in water up to our waist, we're like, what do we do with this scene, which we're sort of seeing everywhere? And in Jakarta and Indonesia alone there are 79 million people using social media, these are numbers from 2016, and over 66 million of them are using social media on mobile devices. Well, we were just in a conference last week where we learned that there are around 600 million people actually living in Southeast Asia now, and over 700 million smartphones in use currently. So there are over 100 million smartphones more than there are people in this region, and that is an amazing statistic, and it opens a lot of opportunity to do some really interesting stuff. So we took this kind of information and thought, well, what can we do? And so in that first year, we just sort of turned on a kind of ambient listening within Twitter, which as I am not like a computer programmer with, I can't tell you how to do it, but it's like a thing that exists in Twitter. You can just do it, I'm sure all of you can. So in 2014, 2013 monsoon season, there were over 13 million tweets that we were able to geolocate about flooding within this bounding box around Jakarta, because people had geolocation on. And we're like, okay, that's enough people having a conversation about flood that we can do maybe something about this and kind of bring the conversation that's happening on the ground to the government, to other people that can make bigger decisions and kind of bring local sort of crowdsourced ground knowledge up to the top. And maybe we can design something with a cross-section of residents and people in the city that can do something a bit more powerful. But do it in real time in a way that is responding to what's happening in the moment, not trying to collect a historic sort of database of what's happened or be predictive about what is going to happen. So we developed a platform called Cognicity. Cognicity is a series of tools. This is all available on GitHub. That is how Padavenchana works. And it can be used for anything. And I'm going to turn this over to Nasheen to give you a little bit more information about the specifics of this platform and how it is a bit different from other sort of crowdsourced mapping programs. But our pilot project for Cognicity was called Padavenchana. And I mentioned this because you're about to hear this term about a million times in some videos. And Padavenchana is the same project as Padavenchana, but as our scope and scope of services kind of and locations sort of increased, we changed the name from Padavenchana to Padavenchana. So for Padavenchana and Padajakarna, Cognicity is an open source software that integrates the management of both social media as well as API source data to make hazard information open and accessible, not only to the residents but also for the government agencies and non-government organizations. And last year we were awarded with the Open Data Showcase Award from the Open Data Institute. And with their support we produced a short video that explains a little bit about how Cognicity works. And we just wanted to play you a few chapters of this, the full version is on YouTube. And the complex hydraulic infrastructure. Among the many hazards faced by residents in the city, flooding during the monsoon season is particularly severe. We lead Jakarta's residents also generate 2.4% of the world's tweets on the social media platform Twitter. With some of the highest concentrations of social and digital media usage in the world, residents actively use their social media networks to gather and share information about disaster events, especially floods. But rapidly urbanizing cities like Jakarta are often thought to be data scarce. By enabling residents to share flood data openly and in real time, a wealth of information can be leveraged to promote resilience to extreme weather events. Cognicity is free and open source software developed by the Open Source Geospatial Lab at the Smart Infrastructure Facility at the University of Wollongong. It gathers, sorts and displays crowdsourced data, thereby making accurate information available in real time to residents, community networks and government agencies. Powered by Cognicity Software Peter Jakarta.org is the first map of its kind to produce a megacity scale visualization of flooding, using both crowdsource reporting and government agency validations in real time. The map displays the rapidly changing conditions that affect infrastructure systems and their users, providing immediate information for residents, first responders and municipal agencies. By enabling reliable, non-trivial communication between users and government agencies, the platform promotes civic co-management as a form of megacity climate change adaptation. As our website says, information is the most important resource in a disaster, which is why it's critical that this is all open source and open data, because decision making has to be a democratic process. One of the critical successes of the project is that it bridges this information gap between the disaster management agency and residents. And so in order to gather, sort and visualize data from a variety of localized networks onto these lightweight platforms, the software is organized into three software modules, and this next chapter briefly explains the architecture. In display accurate time critical information from a variety of data sources, Cognicity is organized into three software modules. The reports module is Cognicity's interface with external data collection platforms. It collects information about flooding, river hikes and infrastructure performance from a variety of sources. The database module stores the collected information for a specified duration of time, in order to ensure the relevance of real-time flood information. The server module serves this data to a client interface. The Peter Jakarta.org webpage, for you either on a desktop computer or on a smartphone, providing users with access to data in real-time. By gathering, sorting and visualizing data, Cognicity transforms the noise of social media into critical information for residents, communities and government agencies. The free web-based platform Peter Jakarta.org has been used by hundreds of thousands of residents during monsoon flooding. It is also used by the Jakarta Emergency Management Agency to monitor flood events, to improve response times and to share emergency information with residents. But beyond sharing information within a city, filling in this information gap between residents and the government agencies, it's also critical that we share this data with other tools and organizations that are working towards similar goals in order to make all this information available for anyone who needs it. And so this next chapter briefly explains our output API stream. As new tools, applications and software are adopted by municipal governments and NGOs for the identification and management of urban risk as a result of climate change, the need for greater integration of various data they collect becomes acute. An output API stream gives secondary applications open access to Cognicity's data, enabling them to optimize their performance with an array of open data. And so when we say software as infrastructure for climate adaptation, one of the critical advantages that soft infrastructure has over hard infrastructure is its flexibility. And Cognicity servers are designed to scale to optimize the flexibility and ensure that we're always delivering a lightweight platform even during peak times. During and after disaster events, evidence shows that residents self organize using digital infrastructures. In order to handle the peak of data traffic during these emergency situations, when access to accurate time critical information is most important, the web server running Cognicity is programmed to scale virtually to meet peak demand. In this configuration, Cognicity is able to handle high volumes of traffic during disaster situations without requiring expensive but idle service during off peak times. Running a dynamic set of virtual servers on top of a well maintained hardware platform allows Cognicity to provide resilient, flexible services that respond to extreme weather events. Designed within Amazon web services, Cognicity leverages the scale and capacity of cloud services, thereby minimizing Cognicity's impact on the climate change events that the system helps to monitor. With that name change from Peta Jakarta to Peta Bentana last year, we scaled from serving around 28 million people in Jakarta to 56 million people across these four cities alone. And that was just kind of dealing with flooding as the problem. And I think one thing that what Nashin just showed you where we've seen all of this really hard work going to Cognicity like really high off is a couple of weeks ago we had a pretty major flooding event in Jakarta. So we had over four million map loads in less than 24 hours and the system was able to kind of take that in really seamlessly and let it back out. And another issue in Jakarta is the, not just in Jakarta but Indonesia in general, is that the internet doesn't always work so well or it's really really slow to load. And so we've also kind of worked very hard to kind of design a system that can load them out very quickly even in like poor service conditions. So we're going to try to show you the way that we've also kind of designed this system to be sort of media and platform agnostic. Because as we have also scaled to these different cities we have learned that not every city uses the same platforms to communicate across. So for instance, while we have a high usage of Twitter in Jakarta, we have very few people using Twitter in Surabaya. So we have to learn how to have communications with people who want to report in ways that it's both about building a community. So we're not just sucking up ambient information and putting it on a map. But we're actually having conversation with people and trying to bring them into a sort of community building system to sort of co-manage the cities in a way that is productive. If you're using one of the ways that we've worked to solve this problem is to make a set of cards that can be used via Twitter, they can be used via any kind of messenger service. So if someone wants to message us about a flood, if you just use Twitter you can just hashtag Benji or hashtag flood and we'll get that message and then we'll send you a response to these cards. But you can also, if you're using Telegram, and this is one of the tricky things, is like Facebook and WhatsApp. WhatsApp has all these members, all these people on Facebook, but it's owned by Facebook, that's what WhatsApp is, and they don't like to play ball with us. So we can use them, we can send these messages out, but within 24 hours or so they catch us and they change their API and we can no longer do it. So Telegram on the other hand is fantastic and they're super fast, though it's not going to be that fast with this situation. So if you're using Telegram you can search for the Benchana bot and send us a message. We'll send you this back, and so you can drag this map around and pick your location, we'll just skip that for the end of the time. You tell it where you are, so you can take a photo and you can just do this live. Just take a photo and it'll upload it seamlessly or you can grab one from your photo library. If you have ever information that you need, and I should tell it that this is a test, so that the government doesn't... At the end of that it'll send you a link to show you your report on the map and you can see your contribution and review. So that's kind of how the thing, with Peta Benchana, Peta Benchana specifically covers Indonesia, but we are now looking at other hazards including earthquakes, tsunamis, typhoon, fires, like I know Singapore even have a lot of problem with our haze that we like to send you guys, so we're kind of starting an investigation into that, which brings up a lot of really interesting questions about privacy. Because it's important, we already have a system that provides complete privacy to people sending in reports, but with flooding it's not that big a deal, everybody's problem fires a little bit of a different matter, particularly in this haze issue. And then we're also working on a project with MSF, Doctors Without Borders, to kind of redesign the way that they can monitor, and assess, and engage with disaster response management across the globe ultimately, but for the moment it's just Southeast Asia. And one of the other things that we're kind of here for this weekend is to promote a project called LabNet.Asia, which is really to kind of bring all of these amazing people from across the entire region that are working on issues that are both directly urban but also are not but could possibly be, to further the use of cognizity and try to apply all of these amazing thinkers and interests to humanitarian use, and just kind of hacking our urban environments to better suit all people. Thank you very much. Just a cool question from my side, how long has this been in use and production, and how you've seen and what you're in? Next, you mentioned all the different kinds of disasters that are, how has the utility of this system been scaled down? So Jakarta, so we've been in use in Jakarta now since 2014, and what's interesting about Jakarta is that, I mean they always had a sort of disaster, like a flood mapping and disaster kind of response management system. It just wasn't very good, but it's been growing and there's lots of interest from lots of NGOs and different groups from all over the globe trying to help Indonesia deal with this problem, but we are now like, we are their flood mapping system. So they, like, we bring in information from the other things, but this is like, this is the, to respond to situations across the city. And they are actually the ones that have really requested, like they're really pushing for us to do more and more disasters, the fires included. So it has directly and almost, like, it's almost the only thing that's really impacted policy change in Jakarta with regards to sort of flood governance and hydrological system governance, which is kind of like everything that you could sort of hope for, is that you can design something where the people have a voice in changing the direction of government policy that better serves the city. With what? With other NGOs. Yes. So first, yes, all of the reports are only visible on the map for two hours. So you, which means that you can't, like, this is because to prevent the government or anybody from going into the map and like seeing information that is 12 hours old and trying to act on that information when it's no longer valid. So it's always meant to be real time. You can, all of our information is also, it's, we have a completely open API, so you can go in and download all of our information including historical stuff and do anything you want to with it. Yeah, of course, that's all fun. The validation, verification. So one, because we don't just map everything that mentions Flood, we have a conversation with that person, with that report before it goes on the map, that we have had almost no issues with people submitting false information. Sometimes it's not super helpful information, of course, but that happens. But also the government, BPBD in Jakarta, they are constantly monitoring the information as well. So it does go through our AI system, all of our box kind of managing that. But then they look at it and if it's correct, they verify it and you see, like, the little report goes from white and blue to blue and white. So you can see that it's been validated by official systems. So then it kind of lets you know that something is, you know it's right or you know, maybe that's not quite it. That's our way of doing it. I was living in Jakarta, something a little bit of a company told me that we need comparison because you're also working with hot OSM, I believe, provide your map. Do you have a large capital cities in the areas and that's where the area's existing maps are? Yeah, sure. We're not only looking at mega cities, but hot, you guys are homies, we can't go anywhere without them really. So if they're not willing to go into a place and try to map it or help us do that somehow, then we kind of, we have to wait until that information can come up to serve certain areas. In Indonesia, that hasn't been so much a problem. It's pretty well mapped. There are some cities that aren't. Yeah, I mean to a minimum level of mapping is required, but say Ternada, for example, we've been asked to look at Ternada and Ternada doesn't have the same quality of mapping as Jakarta. But it's okay as long as there's like a basic level of understanding of where a person is in the city, it's fine. We can go ahead and start working in there and hot can work at the same time and we can get the information up to date as it's available. And that works for us. Because it's more about people on the ground knowing where they are and as long as they can place themselves and place what's happening then great and we'll get better information as it's available. I think it's a great project. I like it. It's awesome. But I think to get a point, how do you make use of the data? Are you using some kind of algorithm like predictive analytics? So you get this kind of data. First of all, can you predict where we'll be the next floating? Because you have such a lot of data. Second, what you do with the images, so you have to text an image. How do you process this kind of thing? This is like homogeneous data, so you need a couple of capabilities. So I haven't seen anything in your slide how you actually crunch this kind of huge amount of data. And this is useful because you can evaluate outlines. You can validate your data. It's the same. So as soon as you have two millions of images and you have 100 of images, like spam or something like that, you can filter it out, cluster it, you can eliminate all these kinds of outlines. So I haven't seen anything relating to how you crunch the data, how you take use of it in an automatic way by an algorithm, by a deep learning, by AI. Can you tell us more? Yeah, well, it's because we don't right now. We're doing so much data and it's so useful. Why not push it in, you know, in a sort of deep learning thing and you just get one predict, you know? Yeah, so Jakarta has taught us a few things. One is that you cannot predict flooding in Jakarta. You never know. I mean, I think anybody who's been in Jakarta for very long, it's like, you never know, you really never know where the water is going to be and how deep it's going to be and how quickly. And that is only getting more and more true because of the normalization project that is completely changing the flow of water through the city and kind of in a haphazard way. So we intentionally do not try to do predictive analytics with our information. I, particularly from my background in machines, I know our predilection is to do that, but intentionally we sort of hold back from trying to. There are other organizations that do pull our data and try to do that with it. As far as the working to sort of crunch all of the reports and the photos and all of the kind of stuff, that is something that we're sort of working on for other projects, maybe not as much with flooding. But I think that will be particularly important with other disasters when, I mean, say, a major earthquake hits, like you get millions of reports in a very short amount of time. And that, I think, becomes more important. And I don't know, we don't have the answer to that yet, but we will. You know, you'll be interested to connect also to IoT. Now we see people as smart cities. Now they have the lamps, you know, all the lamps that connect to IoT, they can give you humidity, or your data, you can pull it. But I also think this is the key thing about Cognizity, about this project in general and about Cognizity, and its further use is that it's all open. So you can come in to, you can become a member on the GitHub and just do everything. Thanks. Thank you so much. Thank you.