 Good morning, good afternoon, good evening everybody, thank you for joining our webinar, of course we are very excited that we have participants from more than 13 countries from all over the world, and of course for start I would like to introduce us to you, my name is Marcin Bielecki, I'm product manager in Cloud Ferro and of course I'm responsible for the SEM4CAP solution and other EO based services, with me is also Sylwiana Shioska, project manager in Cloud Ferro, and warm welcome to Kozminkara, technical manager from CS Group Romania, developer of the SEM4CAP solution. Maybe a few words at the beginning of the upcoming agenda, so first of all I would like to tell you a little more about why should you use Cloud solution and why you should choose a Cloud solution for the SEM4CAP. After that Sylwiana will show you a live demo of the SEM4CAP and of course after the demo Kozminkara will tell us a little more about the upcoming version 3.0. So of course if you have any questions regarding the webinar, regarding the SEM4CAP, feel free to put them in the chat room after Kozminkara presentation we will try to answer all of your questions. A few words of introduction about who we are. Cloud Ferro is a company founded in 2015, managed by a group of experts with over 20 years of experience in the IT industry. And of course we have been trusted by leading European companies and scientific institutions from various big data processing sectors including the European Space Agency, Elmetsat, the European Centre for Medium Range Weather Forecast, also the Mercator Ocean International, German Aerospace Centre and of course many more. Our main product we specialize in is of course cloud computing and a very large storage at the additional services and I think if we will check our storage from all of the projects that we delivered and of course we are still developing them. We have around 100 petabytes of raw data, of course. What makes us different from the market is of course our specialization towards processing large data sets of Earth observation in Hungary, of course. As you can see we grow very fast, in the last five years we have expanded our team by 10 times. We of course from 2015 delivers and continue to develop a few big projects related to the Earth observation satellite sector, the climate data store for the ECMWF, of course the CREOBS our main project. The Wikio Dias, of course the Dias related to the German, they could be going further to the to the Centre for Cap. As the survey shows us, some of you are not familiar with the Centre for Cap so perhaps a few words on the objectives under which it was created, of course. The development of the software was aimed at of course increasing the efficiency, traceability and the cost reduction based on the use of EO products with connection to the validation of operation practices and agriculture monitoring. This action was also intended to redirect paying agencies efforts to leverage cloud environments and to activate the use of Sentinel products, of course. As you can see on the map of course, we operate with many countries around the Europe and the most of the users are from paying agencies, from government departments and from our experience the software based companies and the satellite data related organizations are also the part that delivers Centre for Cap on CREOBS directly to the paying agencies. What is also interesting that the Centre for Cap is also used outside the European Union, not only for Europe but also we have users from different parts of our globe. For instance, we have testing users from South America, from Asia and from Eastern Europe that is outside of course of European Union. Before we start a live demo, I would like to tell you about the Centre for Cap on CREOBS, how it's managed of course. So, from the start of the Centre for Cap from the release, we create a Centre for Cap image from a virtual machine that is already configured of course with everything, you don't need to do anything more. You don't need to put any additional configuration, you just use the image that is already in the cloud environment, which of course Sylvia will show after my part. And the second most important thing is the data, because Centre for Cap requires Sentinel-1, Sentinel-2 and of course the Landsat-8 data. So the data is here, you don't need to use any public apps to download the data into the virtual machine. Of course, for simplicity of use the platform, everything is here. So, after the installation of the virtual machine, you have all needed components such as Maya of course, PostgreSQL, ApartServer, the Centre for Cap is landed, let's say, in the CentOS operating system. And as I mentioned before, it's directly connected to the EO data repository with Copernicus data. Also, the virtual machine includes the SRTM data, which is required in the Centre for Cap. And if you would like to connect directly to the virtual machine and you would like to change anything in the backend, you have two ways of connection. You can use the SSH protocol or you can use the VNC protocol, which allows you to connect via, for instance, the X2Go client with the graphic user interface. And of course, in a couple of minutes after the creation of virtual machine, you are ready to go. You can create your new site. You can start processing in the Centre for Cap platform. From user perspective, of course, there is no need to do any additional configuration in the data sources. You don't need to put any additional credentials to the hub or something like that, of course. From our experience, of course, clients, our clients, our customers, users will gain a few benefits. So we have always the latest version of the Centre for Cap. So there is no need to worry about any upgrade. You can always use the latest version from our cloud dashboard repository of virtual images. You have always available Airfobservation data experts and safe support if you have any additional questionnaires regarding to Centre for Cap. The main important thing is the data. The Copernicus data is always used and always up to date, of course. One most important thing from the users and from our experience is that everybody wants to have a very simplified billing model. So for the Centre for Cap clients, we have prepared a solution that you only need to use the virtual machine, of course, and the storage. And of course, we are calculating the pricing based on how much storage will be lost. We are charging per one gigabyte of storage, so there is always a simple solution for users. Okay, so what's next? From our experience and contact with users, it appears that many of you need additional training in the operation of Centre for Cap. So probably in two or three weeks, we will launch, let's say, a new bundles of services for the Centre for Cap that, for instance, in the contract for about six months or maybe 12 months of using the Centre for Cap, every user will gain a package with hours. So users can change the hours for training, for internal webinars, and of course the consultancy with the expert and data scientists. And from our perspective, it will be very helpful if the user are at the beginning of the journey of the Centre for Cap. But also, if there is a need on the part of the user to generate specific results, we can also perform such an operation, of course. So we have come to the end of the business and product part of the webinar, and from now on I will give the floor to the Sylwia Nasiłowska, who will present the practical part. Thank you, Marcin. So this part will be about practical aspects of Centre for Cap. Firstly, I will show you how to establish a dedicated virtual machine, and next I will show you the overview of the graphical interface of Centre for Cap. So in order to do this, I will share my screen. Yeah, and first step is to go to our CruelDSU platform, and here you have everything you need to work with Centre for Cap. Firstly, of course, you have to register. And if you are a register, you can log in, and the access to the dashboard you can have by tools, cloud dashboard, or the same is accessible with your cloud environment. And what is also worth to mention that we have very rich knowledge base, and you can search here for the help. For example, for the Centre for Cap, there is plenty of articles and tutorials how to work with Centre for Cap and how to establish the virtual machine. But right now I will show you this, exactly how to do this, and I'm going to my cloud environment. I'm connecting, and here you can see the dashboard, and in the section instances, you can launch the new one. You can name it as you want. I call it Centre for Cap Webinar 7 July. Next, you have to choose the image with Centre for Cap 1.3. And if you are an advanced user and you know that you will have a lot of data, huge region, and plenty of images, you can here also establish the additional volume size. But for the beginners, it is not necessary, so here we are clicking No. The next step is choosing the flavor. You see that the orange triangle is showing you which kind of flavor, which kind of parameters of the virtual machines are not enough for Centre for Cap. The Centre for Cap demands quite a huge processing power, so we encourage you to use this EO2, 2X2 large flavor. So here I'm choosing it. Next step is choosing the networks and security groups. How to establish it, you can find on our Centre for Cap QDS webpage. And then the last steps, very important, is to create Keeper. Because if you want to have access to your cloud environment, you need two keys. One is private and second is public. Public key can be visible and you can access it all the time in this dashboard. But private key, it is available here only once. And this is this moment. You can create a new pair of keys. Create Keeper. Okay, I don't know what is happening. Okay, sometimes it lists a few seconds to create this Keeper. Sorry, I have no idea what is happening here. But okay, you can use some old one and launch instance. And here in the list of instances, you can see the emerging new virtual machine. In few seconds it will be ready. And what you have to do in next step, you have to associate floating IP in order to access to this virtual machine just with your browser. And you can hear, associate floating IP. Choose the new one. Allocate IP. Associate. And right now I can show you already created environment of the SEM4CAP. It is very easy. Just place the external IP to your browser. I have created this virtual machine a few days ago, so I have one example to show you. And I can log in here and you can see here the SEM4CAP interface. Here you can do all things related to your work with agricultural products for using earth observation data. So first step, it can be creating new site. So here you are establishing the region where you will be working. And in order to do this, you have to create the new site. Just naming it with the big letter at the beginning. And then upload the shape file, but this shape file should be in UTM projection and it also has to be zipped. So I have an example here. And I can save new site and my site is created. And in order to arrange the time ranges of your analysis, you can add the seasons. You can of course add as plenty of seasons for your analysis if you want. Just click here the green plus button naming the period and choosing the days. Here I'm choosing just one week, but I think that you can choose longer period for your analysis and enabled here in on. I will not do this right now. But I can show you the results of my previous test site. It was done for Portuguese and we have already here a few products generated. And those are for the north side of Portugal. And here in this section of products, you can see what was done for this area. You can of course download each single image if you want just to see what is inside. But this is not necessary because this sent for cups software is working in every step you need and you don't need to download anything. There's also a few another sections which are telling you about the pace of the processing. You can manage here the production of the thematic layers like like crop types or grassland moving or another agricultural practices. You can observe here how many images are created here. You can add another users and you can see from where data are taken into this software. So this is how sent for cups of environment looks. And for more, you can watch my colleague webinar from December 2020. And he's he's showing you exactly how to use the send for cup environment. And he's showing examples of the product and example of usage of this. So right now I would like to come back to the presentation. And maybe I will show you what kind of products you can acquire from the sent for cup. And those layers which can be acquired from sent for cup are very useful for everybody who is working with agriculture and who wants to use satellite data for this agriculture. First of all, you can acquire very good classification of crop type. You can also get the information about grasslands. For example, when these grasslands were cut by farmers, you can also get can get the plenty of indicators with vegetation indicators like NDVI, Liferia index or F upper. And all those indicators are telling you about yield or biomass. So those are all this information are very useful for agriculture management. And here you can see a few examples of the results of sent for cup. And you of course can have the crop classification. And at the end, the system is giving you a lot of shape files and statistics about the results. So you have not only the polygons with the attributes, but also you have plenty of statistics of this classification process. So for example, confusion matrix you will have and all necessary accuracy for you to work better with agriculture from satellite data. And at the end, I would like to summarize that and indicate that cloud ferro has over two years experience working with sent for cup for common agriculture policy. And we have been working with several various institution paying agencies, companies, research institutions in order to find solution for agriculture. And here you can see the map of regions where sent for cup were implemented. So you can see that it was lamented on the north of Europe and south of Europe. So that means and that proves that sent for cup can be applied in any place, whatever it is independent on climate or some landscape characteristics. So using our And also this is very important that the data is close to the processing. So this solution is very, very helpful for their purposes. And what is also worth to mention that from next year. So using air observation data will be obliged for European country. So this solution we are talking about today is ready and convenient for everybody who will want to face this issue next year. And this is the reason why in cloud ferro we are developing developing constantly and we will be providing also new version of the software. And that is why we invited today Mr Cosmin Cara, who will tell you more about future development of the software. So Cosmin. Yes, thank you. Thank you for this introduction. Good afternoon to everyone. First of all, I would like a confirmation that you are able to hear me. Yes. Okay, thank you. My name is Cosmin Cara. I'm the team leader of the same for cup development team of CS group Romania. I'm glad to be here to share with you a glimpse of on the on the evolutions that will be part of the next 3.0 simple cup release. This version is scheduled to be released in September, mainly due to summer holidays that are beginning for everyone. The system version will benefit of rework significantly work of the web user interface previously. We relied on some server side web framework which induce some dependencies that we managed to remove and replace with the true HTML and JavaScript nice web interface. The actualization of the raster products was enhanced as you could have seen in the demonstration from Sylvie in the current version it is possible only to view a quick look of the products produced by the system in the upcoming version. It will be possible to visually inspect the products at their true native resolutions and 20 meters with the possibility of zooming in and out moving images nice something similar with what you can see in Sentinel playground for for example. Additionally, more information or metadata about products is available on the selection of a product in the new interface with it will also be possible to overlay this raster visualization with the LPIS or parcel data. You can obtain and you will select individual parcels to obtain marker information either in the form of text or properties, or in the form of a graph of time series of the respective markers. All the configuration settings that are currently available only by means, let's say advanced configuration settings are currently available only by means of SQL scripts or a dedicated Linux application will be exposed via the same web interface very nicely. Of course, basic user or a regular user doesn't need to change the default values but for those who are interested in playing and tweaking with the system. This can be easily done without the need to open a Linux terminal and enter some batch script commands and so on. Besides the web interface, there are some changes in the core of the system. We have secured the rest API that is exposed by the Stanford Cup system by protecting it with the HTTPS protocol and authentication token usage. This is mainly intended for those that for business to business or machine to machine communication if you would need to interact with the Stanford Cup system from an external system. The marker database is extended with additional markers, for example with the mean and standard deviation for reflecting bands configurable you can choose one or more such bands. This is the number of valid pixels used for computing this mean and standard deviation for per person. We have done some fixes or corrections related to recent changes that took place in the data source part of the system mainly related to Sy Hub for Sentinel data and USGS for Landsat data. We are still expecting the future to have some fixes for USGS because their API is evolving and from time to time it breaks. We'll have dedicated Docker containers for the atmospheric correction processors for Maya and Centukor, which will allow a better management of the version so that if an update of the used atmospheric correction processor is available. There is no need to reinstall and reconfigure the system is just a matter of updating the Docker. It will be possible for the L4C M1 up to M5 markers to be produced, even without parcel input tables. Right now in the conversion it is necessary to upload this table in a special format in order to obtain this markers. A big improvement has been realized on the classification processor in the classification step. And now the computing time is almost 10 times smaller than in the previous version. Finally, some maintenance work on the version of the operating system libraries. The SLRM packages, SLRM is used for a distributed execution, were updated and switched up to mainstream. I was mentioning that even if it's not in the new line of evolution of 3.0 version. Starting with a 2.0 version, the current version, it is possible to switch the digital elevation mode of previously only SRTM was available in this version 2.0. It is also possible to use AsterDEM version 3 for a higher latitude areas of interest. I mentioned that because I saw a question in the chat box about the Copernicus DEM, but I can answer to that in the Q&A session. So that would be it and stay tuned to the upcoming release in September for the new version. Thank you. Thank you very much Cosmin for the teaser of what awaits us in the version 3.0. And I think that we can go forward to our question and answer section. Of course Sylvia or Cosmin, if you see any questions that are not covered, please feel it. Yes, sure. In the audience there are also two colleagues of mine that I see that started to answer the chat questions. I'm trying to browse them to see if something skipped. First question would be, if we start now to test current version of SEM4CAP and in September we move to 3.0. Can the new version be upgraded automatically in the dashboard? Yes, we will provide with this release as we did with the previous versions, update scripts. And the update will go smooth from the current version to the new one. Another question from Tanasis, are you planning skipping the pre-process with Gidal and move to DataCube? I think that we need to wait a little bit about the whole DataCube projects to confirm that. Cosmin, could you refer to the question from Roland Bradshaw? Yes, there has been increasing interest recently in carbon soil storage. Can this application be used to quantify carbon soil storage? Your answer now at the moment is that the system was designed to be accessible for additional functionality or information. The new module can be implemented, but since the ISA contract for the SEM4CAP software system ended on May, we can discuss on a case-by-case basis this kind of evolution. Okay, maybe in addition to this question, one comment on that from, of course, our side. Because a few days ago in the CreoDS, in the Finder tool, we have implemented a new processor that enhays the resolution of Sentinel data from 10 meters to 5 meters. So of course, I can give you a sneak peek that we will try to implement this solution into the SEM4CAP for, of course, better accuracy of the results. But in context of that, we will inform you in our social communication tools. Are there any new processors planned for the upcoming versions besides the existing processors of L4A, L4B and L4C? Actually, we don't know how it will go beyond this 3.0 release. So we have to discuss also with the ISA, or maybe ISA has to discuss with the European Commission to see how we can continue evolving the SEM4CAP system outside of the contract that we had. But of course, that would be a nice idea to extend the system with additional processors for that. Maybe it would be interesting to create, let's say, like some polling topic on the SEM4CAP forum, if you are familiar with that website. And try to survey what users would like to additionally have as processors in SEM4CAP system. Okay, about the resolution enhancer. The resolution enhancer is ready, but it's ready only for, let's say, sentinel products. Yes, you can for now use the enhanced products in the SEM4CAP solution. We will start working on it. So we will inform all of the community if we release anything. But we encourage you to check our sentinel to resolution enhancer in the CreoDS and, of course, to evaluate if the products after the processing to the 2.5 meter could work in your other projects, of course, related to any agriculture activities. And I think Marcin also a question for you. Are you based on AI for the resolution enhancement? Yes, of course. The solution was based on the AI, so sure. Yes, Thomas, the machine learning was used. So I would incline to say yes to this question. Can the application create historical timelines of vegetation for specific sites? I would incline to say yes because you have the possibility to retrieve the vegetation markers time series and you may keep them if you want to have a history for the respective landscape. I'd like to know a bit more about the biomass calculation provided by the SEM4CAP. Kozmin? Yes, I think my colleague Kozmin can reply to that in the chat. The ATBT or the algorithms are public and are available on the web. To be honest, I'm not sure if they are on the SEM4CAP website or if they are available on the Centro Agri site, which is a system similar. In fact, the answer of SEM4CAP software. To answer to Kemal, questions so far as I understood, there is a small possibility of further developments for SEM4CAP in 2022. Even if there is an uncertain situation for the continuation of the development of SEM4CAP, we are committed to still improve it however we can. So if we have the chance to do new developments in the frame of other projects, we'll do them and the open source software system will benefit of them. Yeah, and of course, maybe I can cover this question from the side of the Cloud Ferro, and of course I can confirm that we will try to develop our solution to the SEM4CAP because of course the SEM4CAP is open source software and we can try to change this a little bit, let's say. And trying to implement before mentioned Sentinel resolution. Anyhow, if there are additional questions, maybe not at this moment I will put the address of the SEM4CAP forum in the chat box and anyone can take a look there or ask additional questions. I'm sure that the small community that started to build around the SEM4CAP system is able to respond, if not ourselves. Okay, so of course if you have any additional questions, of course feel free to ask as Cosmin said directly to the SEM4CAP community forum and of course directly to us through the CreoDS or through the Cloud Ferro website. And we are of course always for you to answer all of the business and of course product related questions. Also, I would like to invite you to Cloud Ferro website to our next webinar that will be out 15 July and it will cover the topic of introduction to the CreoDS. I would like to thank you everyone for coming to the webinar in such large numbers. And according to the webinar we will send the presentation and the recording via email of course. I would like to also invite you to visit our website, our socials in the Facebook, in LinkedIn and of course the Twitter. And if you would like to see previous webinars, feel free to visit our YouTube channel. So thank you all and hope that we will see in our next webinars. Thank you very much. Thank you and stay safe. Thank you and bye bye. Bye bye.