 Welcome back, the next talk is in English. Falls ihr den nächsten Vortrag lieber auf Deutsch hören, möchtet C3Lingo, hilft euch hier weiter, C3Lingo.org als Adresse. As usual you can raise questions to the upcoming talk using IRC, Twitter, MasterDone, with the hashtag 40 remote rein Ruhestage, that's R3S. And then those will find its way to the Q&A Sessions. And also if you have the time please consider to help out writing some subtitles for the talks. That would help a lot actually. All right, our next talk is Automatic Dependent Surveillance Broadcast, Automatic Identification System. Open Data is in the air. Our speakers are Skymaster, he's a Fan of Network Technologies and a Ham Radio Enthusiast. And the other speaker is JJX, a Network Administrator. He loves trains, planes, ships and especially high antennas on building roofs. With that one last remark, it is perfectly normal, if you only see the slides, enjoy the talk. Welcome to our talk, ADS-B and AIS. Open Data is in the air. Who are we? I'm Skymaker. And I'm JJX. We are part of the local Hacker Space Chaos Consulting, which is located in Isalon in the western part of Germany. Our space consists of about 25 members at the moment. And outside of COVID we meet up in our space in the city center. For the last seven years we are tinkering around with a lot of projects. We build mesh networks with Freifunk, contribute to OpenStreetMap, and have lots of Adrinos, ESPs and Raspis, where to colorful blinking LEDs. In the last two years we shifted our activities to the online world due to which some projects in our space came to an halt, but others gained much more traction than before. Like the one we are going to show you today. If you like, you can stop by our virtual space in the work-adventure world and say hello to us. Talking of Congress, we have attended the last real-life congresses in Leipzig and Hamburg. So maybe you have already seen us and our project. As the Congress is the most important event for our space, we are really stoked about giving our first talk this year. One of our projects is, as mentioned before, Freifunk. With Freifunk we provide open and free wireless network connectivity for 13 cities in our area. Besides the usual installations in the city center, we provide internet to multiple hundred refugees often with the support of the officials and sometimes without. We use our gear to build networks for local events and festivals. When we have time to spare, we like to experiment and push the envelope of what we can achieve with our gear knowledge. For example, we walk with a backpack full of gear onto hilltops and try to make long distance connections to the mesh. We call this hobby Freifunkwandan. On one occasion we were able to set up a 16.9 kilometer link with about 150 euros worth of gear. But what else can you do? As we are already on the roof, where we have network connectivity, power and mounting points, we started looking for other fun projects to do. One thing we experimented with are Freifunk connected webcams. From cheap security cameras from China to raspy cams and even real digital cameras we tried multiple options and delivered live feeds and timelapse videos via the mesh. Weather sensors are another topic we explored. For example the open source airhaw project from Stuttgart's OK Lab for the project on one of our annual summer camps. With the airhaw we measure temperature, humidity and particles with an ESP and send the data to OpenSenseMap, which is another OpenStreetMap based project. HamRadio is another activity that comes to mind when thinking about radio waves and antennas. Quite a few of our space members are actually licensed HamRadio operators, so it comes quite natural to incorporate HamRadio and SDR into our hobby. We are using open source software like WSJTX, OpenWebRX, GQRX and many more to listen to the different analog and digital modes. Building antennas and optimizing the setup is a big part of the fun. As you can see there is a lot one can do to keep you entertained for a while. But clearly we can't get enough and are always looking for the next interesting piece of technology to explore. ADSB is one of the things we found and took a closer look at. So, what is ADSB? In short, ADSB is a standard radio protocol which is used to track aircrafts. The automatic means there is no input needed from the flight queue, it just works. The dependent means it requires data from the flight navigation systems. So, it is dependent on that data, of course. ADSB transmits on 1090 MHz, like the old secondary radar, which is still used. That's the thing with the transponder if you heard about that. There are two different kinds of ADSB. ADSB in is mostly used in the US. There it works on 978 MHz, but it's roughly the same protocol and that provides better data and traffic information for the aircrafts. ADSB out is mandatory in the US since 2020 and Europe since 2016. So, every commercial aircraft flying through or from there should send ADSB reports. That sounds interesting, but how do you get the data? Of course, there's software to decode ADSB messages. It's called Dumpter 90 and luckily it's open source. It has about 900 forks on GitHub, but the most known forks are the ones of mutability and flight aware. Fun fact, mutability works for flight aware, a commercial flight data service for quite some time now, so he maintains the flight aware for. It supports the raw serial messages over UDP, but it also generates formatted JSON files. With these JSON files, it feeds a nice map. So, this is the web interface of Dumpter 90. You see the aircrafts above you, like you would expect. And aircrafts, which you received a position from, are highlighted green in the table on the right side. Above the table, you see the received data of the aircraft, like altitude and speed, but also when the last message was received. So, here's the next acronym. We've talked about aircrafts. Similarly to ADSB for planes, we can use AIS to track ships. AIS stands for Automatic Identification System. Its purpose is collision avoidance and for commercial ships, it's mandatory, even though you should have it on your pleasure craft so big ships can see you. AIS works in the Marine VHF radioband, roughly between 161 and 162 MHz. And its transmission is coordinated with time slots, there are 4,500 time slots per Minute. And the transmission coordination needs time synchronization, since we don't want all ships to transmit at the same time. The time synchronization happens over GPS, which is obviously needed already for the position calculation. In the AIS protocol, there are 22 Message Types in total. For us, only 4 of them are interesting. The first one is the position report. That's just the position report of a ship, obviously. But also there are base station reports, where base stations report their position. And the last two are pretty interesting, because ships can also transmit static and voyage-related data, like the name of the ship, the draft, the length, the width, or even the destination and the ship type, like a military ship or a container ship or a truck. And last but not least, there's an H2 navigation report, where visible boys or non-visible boys or even temporary danger areas can send their position or these can be sent also by a base station. Okay cool, so let me track ships. Of course, there's also software for that. It's called AIScatcher. It's fresh and new software from the Netherlands under heavy-day development. And it has significantly better reception than other software. It only outputs raw serial messages, so you have to decode these serial messages through GPSD, which is pretty old, but very good software on Linux. It has support for multiple SDRs, like the RTL-SDR, or the AirSpy, or the HackerF. And of course, it's open source. But before the software can do its job, we need some hardware to receive and decode the data. First things first, the antenna. A good antenna built for the desired frequency is key to a strong signal. We like to use ground planes, because they are easy to build with a minimal set of tools and cheap materials. An alternative are PCB antennas, which are readily available for AISB. Those antennas are small, robust and precisely built for the frequency without costing an arm and a leg. If about 20 Euro isn't an issue, PCBs are a preferred way to go. Either way, the building instructions for both antennas are available in our Git repository. Beware of cheaply made antennas with very poor performance, available from the usual online sellers. On the top left, you can see a ground plane antenna. On the bottom left, you can see a PCB for 1090 MHz. And on the right, you can see a PCB in an outdoor worthy enclosure. Next up, you need a receiver to convert the analog signal to digital. There are special ADSB and AIS receivers on the market, which are rather costly. Instead, we'll use the beloved RTL-SDR-Stick. These DVBT-Sticks are the cheapest way to buy a soft verdefined radio capable of tuning into a relatively wide range of frequencies. Speaking of which, not every stick is suitable for our use case. You want to look for a 820T or 820T2-Tuner, und avoid 50-Power-Tuners. Reason in the frequency hole and poor gain settings on the lateral models. To connect your antenna with the receiver, a piece of coax cable is needed. The cable should be as long as needed to mount the antenna in a good spot and as short as possible as every meter adds to the signal loss. The quality of the cable and connectors also has an impact on the signal loss. We usually go for RG58 with a pair of name brand SMA connectors on each end. As the SDR receiver is not designed for ADSB nor AIS reception, it has some flaws. Due to the wide frequency range, signals on neighboring frequencies, for example mobile phone networks, can overload the receiver. A solution is a band pass filter for the desired frequency, which will decrease signal strength on unwanted frequencies. The filter should be installed directly on the receiver input to be as effective as possible. The effect of a band pass depends on the location of the receiver, but in general is always a good idea to install one. While there are SDR dongles with built-in band pass filter, a dedicated filter is the better option. In some scenarios, with very crowded frequencies, the combination of both a built-in and additional filter yield the best results. The last piece of hardware needed is a PC to install the software on and process the data stream. While in theory nearly any old or new PC would work, we tend to use the ubiquitous Raspberry Pi. It has enough processing power for our task, while being affordable in SIPSON Power, which is a big plus running 24-7. Without active cooling, it is an absolutely silent setup, and without any mains power anywhere near it, it's quite safe to play with. Everything tied together into an auto-capable enclosure can look like this. Such a setup can be mounted outside near the antenna to reduce cable loss and can be fed entirely by power over either net via a single network cable. On the left you can see an AESB and AIS Outdoor setup. In the middle picture you can have a look inside such a setup containing a Raspberry Pi, USB Wi-Fi adapter, SDR-Dongle with filter and a boost converter for energy. We also use small mobile setups based on Pi-Zeros with a power bank for traveling. The two antennas of a mobile setup can be seen on the ripe mounted to a telescopic antenna pole. No show me the planes and ships. With all the data we needed a way to visualize it. The command line output however didn't cut it for us. We wanted strangely enough a graphical user interface. Wouldn't it be great to have a website to look at all planes and ships from all receivers combined. While there were commercial options for one or the other there seemed no solution for both. And we don't like the commercial aspect of most sites. The only logical solution was to start our own project and that is what we did 6 years ago. So this is the map. Obviously you can see the aircraft and ships we receive. On the right side you have the aircraft table and several buttons for different features of the map. We will show you some of them now. If you select a plane, this is what you get. In this case it's the approach queue in Frankfurt. You see the colorful tracks which change with the flown altitude of the aircraft. And you can also see the model of the airplanes on the map since we have icons for the 100 most common aircraft Models. You can also see the received data of the selected aircraft on the right side along with a photo of the aircraft model or even of the aircraft itself if we had one. Speaking of the plane icons, here you see some of them. Like already said we made icons for the mostly seen aircraft models like the Boeing 737 which you can see a second in the bottom line. But we also have icons for more rarely seen aircrafts like the Ilyushin 76 in the bottom right corner or above that the Airbus Beluga or the new Beluga XL in the middle. Here you see a police helicopter hovering at an altitude of about 600 meters. You may have noticed that the helicopter has a different color than its track. That's because we color special vehicles like military or search and rescue vehicles in red and green so they are easier to spot on the map. Also interesting airplanes and ships are highlighted in pink like the Airbus Belugas or the Antonov 225 so if they appear on the map they are definitely an eye catcher. The transfer in blue aircraft is tracked with multilateration. The position of aircrafts which don't send their position for whatever reason can be calculated using the signal delay between at least three stations. That's another benefit thanks to the work of mutability who not only developed dump 1090 but also the multilateration server. On the right side you see several more layer options for the map. JJ will tell you more about them in a few seconds. Some but not all ground vehicles on airports are also transmitting their position. Here you can see some tugs and also a truck of fire brigade in Düsseldorf. Not seen here is a transmission tower which we think for testing purposes only sends ADSB beacons without a position. Since we have three stations close to Düsseldorf Airport sometimes even that tower shows up on the map with the calculated position being accurate to around 20 meters. Just let you get an idea how accurate multilateration can be. As already said we are tracking ships and this is what it looks like. On the right we have the familiar sidebar with the ships info. Beside the ships name and call sign we show types, speed, direction and sometimes even the destination. On the map itself we see the vessels indicating the direction if possible and drawing a history track. The different special types of ships are color coordinated according to the aircraft Kalashim. There is another great open street map based project named OpenSeaMap. It shows ports, lighthouses and much more. As OpenSeaMap is very interesting to look at and it helps to understand the routes the ships are taking. We incorporated it as a switchable layer easily accessible from the sidebar. Talking of interesting ships here you can see the tall ship Gorchfock of the German Navy on the run from Kiel shortly after the expensive repair. Due to the high antenna position on one of the masts it can be received and tracked from a longer distance than any other ships as seen here. Another interesting thing to look at at this ship which is anchored. Moved by wind and waves it moves in a circular fashion around the anchoring point. But where does the boxy grid bike track come from? Well it is the lack of resolution of the latitude and longitude information transmitted via AIS. While it won't be an issue on the normal circium stances you always have to keep in mind that the accuracy of the ship's position is limited by this. The whole action on the map only works due to the numerous stations feeding us data. The stations are listed with some stats and the approximate positions on the map. When selecting a station you see some detailed info and a list of MLED sink peers. The more sink links the much meta multilateration will work. On the map you can see a visual representation of the sink links to the neighboring stations and the quality of each link, where green is optimal and red unusable. In an area with many MLED linked stations as shown here there's a good chance to track planes without position info quite accurate. The more data you have at hand the more important it gets to have good means of filtering the data to find the bits you're looking for. At the moment we have the following filters in place. The filter by fix option allows you to show MLED only planes for example. The by staged filter allows us to only show the vehicles received by one specific station, which is very useful if you want to check on your stations coverage or just want to see nearby vehicles. With the altitude filter you can limit the maximum altitude. As seen here we selected a maximum altitude of zero meters which is equal to on ground. With this setting you will only see planes on ground, ground vehicles and ships which do not have a flying altitude obviously. By the way what you can see here is our ground coverage in Frankfurt. The by type filter allows to show only interesting or military or fire rescue and law enforcement vehicles. The combination of all three is possible as shown here and the filter works for aerial vehicles and water bound ones alike. As we keep a history of all recent positions we are able to create a heat map. The heat map gives an indication of where we have good coverage and where we are lacking. It is easy to spot traffic hotspots for example here Frankfurt, Düsseldorf or Munich. The filters we have shown before are working for the heat map too. So we can create a heat map for a specific station for airplanes or ships or even for a specified altitude to check the ground coverage. As aviation is heavily impacted by the weather we include a precipitation layer which we can do thanks to the open data policy of the German weather service. It can be very interesting to see the planes getting diverted in real time to circumvent areas of bad weather. There are some more features and functions built in but instead of showing you every little detail you can explore the map yourself later on as it is publicly available. We know some of you are eager to see statistics about collected data. So here we go. You can clearly see how the amount of flying airplanes decreased during night time. The daily peak is around 3pm while the low is around 3am in the morning. Also you find that the least aircrafts are flying on Saturday nights. We think that's because the big cargo hubs by DHL, UPS and FedEx are closed on Sundays so flights in the night decrease even more. The difference between night and day time is around 400 planes in absolute numbers. You even see the up and down from night to daytime in the network graph since the received data correlates with the flights. Back to these two graphs. When you take a closer look you see that the quantity of planes not sending a position is about a fifth of the total planes. So what we need is multilateration to track them. Compared with the quantity of planes without position, we can track only a third of them with multilateration right now. So there's a big potential which we can't use. There's only one solution for that. We need more stations. I think I don't have to say much about this. Does anybody know what happened in March 2020? Anyway, besides the incredible drop in March, there are two interesting things in this graph. You may ask yourself where these peaks on any other day come from. Well, in the spring, it was the weekend hobby pilots who pushed the numbers. In the summer, this gets mixed with the holiday flights, which were possible again at that time. Last but nicht least, you see the same effect with the fall holidays in Germany starting mid-October and ending at the beginning of November last year. So here are the statistics about ships. The first station in Kiel is feeding us data since the end of May of this year. Then three stations near the Rhine in Düsseldorf, joined in July and August, hence the increasing number of received ships over the first half of the graph. You can clearly see the peaks on the weekends when the most pleasure crafts are underway. Other than that, there's not the same day and night effect like we've seen with aircrafts. We think the reason for that is that ships also send their position when they are at anchor or moot. Aircrafts have to turn off their transponder when they reach the gate. Also und I think that's the more significant reason. There are no restrictions like the flight ban at night for ships. All received ships are feeded into a database. Since we started receiving data in May of this year, we already saw 8800 unique ships in total. It would have been even more if we wouldn't have restructured our database structure in September. And also if we wouldn't have forgotten to rewrite the the insert query in the feeder script to these changes. Pace palm smiley. Some of you may ask yourselves what technologies we used. Well, we tried building it as independent as possible, but at some point we decided to use some small frameworks instead of reinventing the wheel. First of all, we use Vanilla JS for the frontend and CSS for the styling. All the plane and ship icons are made in SVG. For the map we use leaflet JS, which is a great toolkit for building interactive maps, even more so when extended with some plugins. Speaking of which, we use leaflet real time for all the real time action on the map. Leaflet heat is used for the generation of the heat maps and leaflet active area is used for some UI enhancements. For all the interactive tables in the sidebar, we make use of Tabulator. In the backend, we use bash php jq Maria DB and SQLite. But let me give you a brief introduction on how we process the data. So this is what a simplified version of our Dataflow looks like. We are starting in the top left with our Raspberry Pi. We name it ADSBerry Pi. And we start with the incoming data from the SDR running into the Dump 1090 process, which then will feed the local web server with the stations let info. And it will feed the ADSB service running on the Pi, which will then forward all the JSON file with Plains and Infos to the web server in our data center. The web server will then push all the info to the live data service, which will combine the data of all incoming stations, and will combine it to an output feed, which will then be served by a web server to the clients. But there's more than that. Multiliteration works, that you have installed the MLAD client on your Raspberry Pi, which is feeded with the station raw data. The client then pushes the station raw data to the MLAD server, which will use the data from the Raspberry Pi and other MLAD clients to calculate MLAD Positions, and will feed it back to the Pi, which is the blue arrow MLAD back feed, which will then use it to feed it to its own web server. The MLAD server will also feed it to a service, which will create an MLAD JSON, that will be feed it to the live data service, where it will be combined and be available on the web server. So our clients will be feed it with LAD and MLAD information. You see, the live data service also feeds the stats service, which uses a database, which is a MariaDB to generate history and statistics data, and will build a statistics JSON to serve it to the web server, and then to the clients. But there's more than ADSB, AIS. We're starting on the top left again with the AIScatcher, which will decode the AIS data from the SCR dongle and will push the NEMA feed into the GPSD. And the GPS pipe will take the NEMA feed from GPSD and convert it to a more or less readable JSON file, which we will then feed into our AIS service on the PI. The AIS service will push the information to the web server in our data center, which will then feed it into the live data service. The process here is a bit different, because the live data service will push the data into a database, and our live data out service will generate the feed for web server and client from this database. That are two different approaches on handling the data. The AIS data is with database in place, while the ADSB data will work fully without the database. And the database is only for history generation. You may ask why we use two different approaches. Well, this is due to the fact that we implemented ADSB a long time before we implemented AIS. And from the things we learned from ADSB, we created a better process for the AIS. Maybe in the future we will work on the ADSB data handling, but at the time we don't know. So, we have shown you all the little features we have implemented. We have shown you the text stack, how it works, the hardware and so on. The only thing left is to show you or tell you what we think is next. What we like to do is implement more icons for different airplane types and photos of airplanes. So you have a nice, better experience in the UI. The statistics is a weak point until now. So we want to improve on that and build a completely new statistics back-end and front-end to give you more and more beautiful statistics, because we really like statistics and graphs. Another kind of weak point is responsive design. So while you can open the website on a mobile device, depending on device and operating system and used browser, your mileage may vary. So there's room for improvement too. We can never have enough stations, so we definitely want more stations, but we cannot do that without help. And maybe sometime in the future we want to experiment with another data source form. At this point we have free, normal ADSB, then AMLAT and then AIS. But who knows what we'll find and maybe we'll implement it. So how can you participate? You can, of course, build a station or you can feed us data from an already existing receiver, which is feeding to another service. Right now, most of the receivers are located roughly in the western and southern parts of Germany. There are also people feeding data in Erdiel and in Washington. If you live in an area without CoverWedge and want to feed us data, we would be happy about that. But we would be also happy if you would feed us data from a receiver, which is just two kilometers away from an already existing one. Since this would gain CoverWedge in terms of low-flying aircrafts nearby and it increases multilateration quality. In any case, have a look at our instructions on how to build a receiver and how to feed us data in our GitHub repository. And if you have an idea or a question, talk to us. The map is available at adsb.chaosconsulting.de and if you want to feed us data, send us an e-mail, or even better, join us on our Mumble server. We will hang out there most of the time during Congress. We also have a Telegram group, which you are free to join. Last but not least, the source code of our website is of course available at GitHub. Thanks for your time and we are looking forward to answering your questions in the Q&A now. Thanks both of you for your awesome talk. There were a couple of questions from people on the Internet. I'll just pick a couple of these. First one would be, what measures are implemented to protect the privacy of pilots who are identifiable by being the sole users of their own aircraft? Yeah, good question. I think there is no such way of doing that as the data is unencrypted in the Internet. And everybody can receive it as shown. So you can't really do that. If we would do it, it would be too late for it because the data is already out there. All right, thank you. Then it appears that another group is doing something very similar. Their question is, we've got a Hacker RAF on a Raspberry Pi 4 on the roof, but how do we do that? So, we have more on the roof, but have huge stability issues, especially with Open Web or X. Did you experience anything alike? We are not using the Hacker RAF, but we are using RTL-SDR dongles. We have seen stability issues with older and cheaper dongles. We don't know if it is because the dongles are some years old and they were very cheap dongles. But we changed these dongles for newer and more expensive ones and haven't seen these issues again. But we can say for sure. All right, thank you. Given that AIS is not that popular in Inland waterways, apart from the Rhine Donau region, is there anything you can recommend for tracking inland boats? I don't know if I got the question right, but there is AIS on different canals in Germany, so Dortmund-Emskanal and so on, where I've seen it. So you can track ships via AIS, even if you're not that near to a canal, if you're in a good spot, maybe on a hilltop, and yeah, use an antenna and an RTL-SDR-dongle and check for yourself if you can get some messages. All right, thank you. I hope I get the next question correctly. How often do people complain that you do it unfiltered and so on commercial? Never. Straight answer, thank you. Until now. All right. The next part is a bit of a longer question. We maintain a marine traffic station, which I kind of inherited from a former colleague. Is there anything else that you dislike about marine traffic, other than that only marine traffic is tracked? And could I use our setup antenna to send data to your project and to marine traffic at the same time? What does that make? Okay, I'll take that one. It depends on the type of station, if it is a ready-billed station, I don't know what it can do, but with a homemade Raspberry Pi-based station, there's a very good chance that you can use it to feed more than one service by installing one and adding the scripts or feeding clients for another service. Yeah, what do we like about these services or what we don't? It wasn't really a question in terms of AIS because we started with ADSB and at the time when we started, there were, to our knowledge, no good non-commercial services. There may be one today, but when we were doing ADSB, it was quite natural to add AIS at some point. We never thought about it, if it was a good idea or if it is needed, we just wanted it. All right, thank you. Next question would be, you mentioned the mobile setup. How much power and mobile data does that one need? Well, power-wise with the Raspberry Pi Zero around, I don't know, we can power that with a power bank and, I don't know, one day should be possible. Also for the data, for the traffic, I think it's around one gigabyte, but maybe JJ knows that better than me. So a gigabyte a day. Yeah, it depends a bit on how much traffic you have, how much messages there are to decode, but with a decent, I think, 20,000 milliamp power bank, I got about 30 hours of runtime for an ADSB-only setup, and data depends on what you see and what you get, range-wise and plane-count-wise, but with two gigabytes a day should be fine. So it's not a thing you want to deploy on mobile data for a long time, especially with German-capped mobile tariffs. All right, thank you. But you can fine tune that down to maybe a three-second interval or whatever. Thanks. Another question is, do you have more information on how your MLAT implementation works? Basically, it's the normal multilateration server and client, but we also implemented an S-Tunnel, so that the connections from the clients to the server are encrypted, and that's it. All right, thank you. Next question, I hope I get the acronyms correctly. Where do you get the registrations for the ADSB ICA-Adresses from? Is there an open database available? We use a database from 2018 that was licensed under the MIT license, and we update that regularly, so we are also spotters, so when we go spotting, we update the database with our own data. All right, thanks again. Are there plans or projects to explore the new ADS-C? Not yet, maybe in the future. All right. And I think that is actually the final question. How expensive of one ADS-Barre pie to build? That depends. There were times where you could build such a setup in an autoworthy enclosure for about 100 Euros, but at the moment it's quite hard to get hardware and it costs a bit more, but if you want a good ADSB-Only setup, you can work with about 120, 130 Euros in Germany with a small little test setup without filter, without casing and so on. You can even go cheaper, you may have a Raspberry lying around or use just a laptop for a day or whatever, then you can buy an ADS-B-Dongle for about 25 Euros and start the journey. Full setup with ADS-B, AIS filtering, good antenna mounts and all this stuff all around 200 Euros. All right, thanks. Are you willing to take one more question? Yeah, sure. Thank you. Is ADS-B also used by smaller planes? The person asking seems to live about 10 km from a small airport without a commercial service and is wondering if he or she could receive information from the GA planes that fly over the house. Yeah, sure, not every plane. Many of the smaller aircraft do not send their position information, so you can see the beacon, you can see it in the list and if you have AMOLED coverage, which means at least three stations receiving the plane, then you can get a calculated position. There may be smaller glider types that use other services such as Flarm or Open Glider Network. But many small planes use ADS-B in some form, so there would be something to receive. Thanks a lot. Just as a common question that popped up a lot of times was how can I contribute to your project and it seems you answered these with the second to last slide. But up to this point, that question came like three times. Thanks again for your great talk. Thank you very much for answering all the questions. It really is appreciated. People watching the streams, you will now see another edition of the C3 News Show and coming up afterwards are the lightning talks of today, the first round of the year. Thanks a lot. If you have the time, please provide feedbacks to our speakers, that helps to improve. It also is some kind of appreciation if you actually do provide feedback apparently. Taking from the questions, a lot of you like the talk. So again, thank you for the great talk.