 I'm very happy to introduce Lila Fisch to you, who is giving the talk on maritime robotics. She's active at CCC Munich, for example. You know her probably from the last congresses. But now she went to the coast to research maritime robotics, and hopefully at the next camp where we have some awesome lakes, we see more interesting technology also in the sea and in the lakes, and I'm very much looking forward to that. Thank you Lila Fisch, your applause. So over the last two, three years I've been getting more and more interested in the area of maritime robotics, and as someone who's been visiting congress for quite a while as well, at some point I sat down and was wondering, how did I miss all these amazing talks about robots in water that must have been here at some point? So I had a look at the things that we talk about here, and we talk about a lot of interesting robots that go across the land, fly through the air. But there really were no talks about robots in the sea. So here I am at a talk about gated communities, and indeed the oceans are gated communities. A lot of them talk about the ocean floor factory thinking about getting oil and gas from the sea, military operations, or getting more resources from the sea. There are also biologists and geologists doing research, but the big investment is coming from these sides. So here I am now to call thee to the oceans, whilst we still have much more things on this factory floor than just machines. So there are currently still fairly interesting species being discovered at the ocean floor, from dragon-like to cuddly, and we know more about space than we know about our own oceans. We can say we have mapped all of our oceans, but the resolution at which it's mapped is kind of ridiculous. There are some fairly big dragons that can still hide in this resolution. And besides finding those dragons at the ocean floor, well, there are a lot of interesting challenges and other things to do in the sea. I tried to find a picture that is really impressive showing a broken maritime robot, but there's one big problem. If a maritime robot breaks, it tends to not come back to you. So you can't really take any impressive pictures. This one was attached to ropes, so when it imploded at 50-meter depth, because we had slightly under or overestimated the wall thickness of this ball, we got back our electronics washed in salt water and a hull that was not really usable anymore, but at least I could take an impressive picture, and I found that car window motors are really, really good motors, because they still work even after being soaked in salty water. Then there is a lot of interesting data to find. There is open sea map, which is trying to map ocean depth, for example. And then all boats have to carry AIS beacons, which give information about them, which can be turned into maps of what boats are moving where at the moment. So anyone who likes Flight Trekker, you can also check out this to watch boats moving and oil boats standing around in the sea waiting for prices to change. And then you can filter these things. So for example, you can find out where are boats that are only fishing boats, and you can see what our fishing fleet is currently doing. Then there are a lot of shipwrecks at the ocean floor. So of course within the laws, you can go and check out what might be hidden in your nearest lake. And they might also be quite, or robots will be quite useful for monitoring what is put into our oceans to do regular monitoring and find out what chemicals are there without risking human injury. So to first get started into this whole area, I tried to categorize the robots that I'd come across over the last years into a family of robots. The simplest thing you can do is you can put some sensors on a robot, put it into the sea, talk to it via satellite or via local network. And then just get information about wherever that sensor boy is drifting or where it's anchored. You can add more sensors to it. And then at some point, you want it to move. So you turn towards the traveler, which is something that uses its environment effectively to move because battery charge is something very valuable. So you generate some energy from the sun to get sensors working, and then you try to use your environment to start moving. So for example, sailors will like this, or this one here uses wave energy. It will not be a comfortable boat to sit on because it actually exaggerates wave movement to propel itself. And this one here again uses wave energy to get moving. And it's already starting to move a bit under the water, and then you want to get some more data from under there, right? So you go into the underwater airplane section. So you all know airplanes, right? They have big wings, and they use this to fly upwards. Now, the underwater airplanes have small wings, and those small wings are in water. So they have to be kind of in different sizes and all. But still, this is what helps control the vehicle. And like in airplanes, if an airplane stops in midair, you have a problem. If your underwater vehicle has to stop if it's an airplane design, you also have a slight problem. Because as long as it's floating, or as its basic status floating, it'll start crashing towards the sea surface. So these skies are pretty useful. They use very low energy. But you have this condition that you can't stop in the middle of the sea. So if you go to the ocean floor for a longer period, you'll use one of those, which uses a propeller to propel itself. And you get to the limit that you can't get any closer because of exactly this problem that you don't know what's ahead of you. And you don't have much time to react because you can't stop. So that's when you get to the diving box. The diving box has a lot of sensors. It has a lot of actuators. So it can move in any direction. But it's not very energy effective. So you only have a few hours where previously you could have a vehicle moving for days or months. And these come in various sizes. And sometimes they have a wire attached to them, a tether. And then you can use them for a bit longer because you can also send electricity over them. Or you can have a human at the top controlling the vehicle. And then last but not least, we've got the wildcards, which are robots that are often inspired by animals or designed for very specific purposes. So we've got three animals that you can clearly recognize here, I think. And then the top one here is a quite interesting one that was specifically designed to collect oil spills. So the sailing boat was designed against any instinct for a sailing boat, but to drag a long net behind it. So onto some basic principles that you need to think about besides building a robot. If you want to put something in the sea, you have to think about what its weight is relative to the water. If its weight is lighter than the amount of water that would fill its space, it'll always come back to the surface, which is a quite useful thing. If it's called buoyancy negative, it'll go down, which is quite good if you want to get a vehicle to high depth. Because then you don't have to use energy to get it down there. But you have to keep in mind whatever happens, it needs to come back up again. So you need to be able to get rid of some weight. And then you have the magic point in the middle where the force of the weight and the force of the upwards force are the same. And then you get basically a lovely vehicle that is floaty and very fun to control. But overall you always want to be around that zero point because otherwise you need to use a lot of energy. So my general rule when designing or looking at a vehicle to guess its weight is I take the vehicle and replace every bit of the vehicle with a bottle of water. And then I know how many bottles of water I have to carry or how much this vehicle weighs for me. The next problem is that you have pressure there which is higher. With every 10 meters of water you get one atmosphere more. So that's what we fell over when we tried to get our vehicle to more than 50 meter depth. The most important factor in this is how big the vehicle is. And that directly influences how thick the walls of it have to be. So if you want to go to really high depth you need to have specifically designed very thick metal or glass containers. To give you kind of a scale the Mythbusters did a very entertaining episode where they tested ping pong balls which make 25 meter depth. Biologists like to put styrofoam cups into their test containers and those usually come back from about 1,000, 1,500 meters very nice and tiny. Now our oceans are four kilometers deep on average and the maximum depth is 11 kilometers depth. Yes. So in the worst case our vehicle implodes and everything that was there before is replaced with water and you suddenly have a very heavy vehicle attached to a tether or not even attached to a tether and then definitely not coming back. And this happens occasionally so sadly this vehicle that was quite capable of diving about 10,000 meters deep did not come back and the guess is that an implosion happened somewhere. So now we've taken care of the basic problems of having something that needs to move on or underwater. How do these things differ from the air and land vehicles that we know in communication orientation? The sensor boys and travelers are quite similar. You've all or some of you might remember the talk about Iridium and SPD which is a quite useful service when you run out of all other various wireless communication methods. Now underwater you have one big problem. Wireless communication doesn't penetrate seawater very far as in it goes a few centimeters for it. No chance of using it for your vehicle. Then if you look overall the electromagnetic spectrum you see that the sea is blue. So let's look at this wavelength. We could use a laser for example to do some 3D scanning but there's one big issue with the sea. The seawater isn't clear like the sky. There are even some animals that have a basic defense against being scanned by lasers. So just imagine what some crazy waves will do or some mud in the water. So hello acoustic waves. This is how communication and orientation in water works. This is my hand. I'm holding it into the water. You can see the water surface and you can see where the sonar beam detects my hand. Now there is one big problem with this. Speed in water of sound is not the same as speed of light. So what we're used to get instant pictures, if we want to look 700 meters far in water we have to wait one second for the sound to travel 700 meters and then 700 meters back to us. And similarly if we have an opening angle of one degree which is quite good for a sonar, we get some pretty bad resolution if we move a bit away from our vehicle. So with 100 meter distance our vehicle just disappeared. And we also cannot tell what is ahead of our vehicle that way. So if we look back to how far our ocean is mapped, the 100% mapping actually comes from satellites which look at the ocean surface and then the second level kind of 100 meter resolution. This is boats. So they will get about a 200, 100 meter resolution. So they've covered 10 to 15% of our oceans. And then we've had a couple of AUVs that got close to the ocean floor to give us a bit of a better resolution in some areas. And we cannot just use the sound to orient ourselves. Sound is also generally a very interesting thing to look at in our oceans. So if we just listen, there are a couple of interesting things we hear. We can notice how sounds that we always hear slightly change it depending on weather and other conditions. And the lower frequency, the further we can hear something. So there are these stories that there are things that were heard that nobody knows where they came from or what they might be. And there are some interesting nice services where you can listen to the sea floor. But there are other things in there that we're not so much supposed to hear. And that have to first be looked at and then maybe removed. And then sometimes bits are accidentally removed. So we can't tell which bits have communication in there that we're not supposed to hear. So those hydrophones are coming online and offline sometimes, but you can listen to them on the internet sometimes. And then researchers also give you access to some things. So I've got three examples for you to listen to. So this is what a whale sounds like, sped up 10 times. Then this is more of a not so animal like. So you might recognize the sounds from vehicles slowing down. So these are larger things moving around in oceans. In this case icebergs. You can also hear icebergs coughing, for example, like that. And then here is human activity. Those are basically very loud explosions at the sea floor, trying to find out what's going on underneath, finding oil and other things. Now, in my description of this talk I wrote somewhere, it's not all that difficult and expensive, but I've just been telling you how these things are complicated and I can tell you commercial sonars are quite expensive. But we've got a chance now with electronics getting smaller and smaller. This reduces the size of the things that we can build. And therefore we don't have to carry a 60 or multiple ton vehicle around anymore if you want to do tests. So we've got high energy batteries now, which I've got here one. And then I can just use a phone to use as a computer. We can use small, ah. So yeah, it's just a battery, which is nice and compact. And a small phone. And then these were used quite creatively by students in a robotics competition. They are very useful for soaking up water that's been spilled. And this makes a quite good waterproof container for your electronics on your robot and your batteries. So this is done all I wanted to show. And another couple of interesting things that you can see from people starting to build their own boats and robots is drinking bottles make nice, fairly pressure-proof containers for smaller depth. Syringes are good for building engines to change the weight of your vehicle. So for example, the glider that you see at the center here is one of those. And then on the sensor side, it's a bit harder still to find things. So there are things called fish finders that people use on boats to find fish, but they should be usable for sonar detection as well. And then under the words, a crow in audio or whale listening, you'll find hydrophones, which are microphones specifically for listening to underwater sounds. But those still need a lot of work to be really useful in boxes like this. So they need reversing and looking at additional amplifying circuits and analysis and things like that. But it's ready to play with, and things have gotten small, so I think you guys will hopefully have some good ideas of how we can build smaller and better vehicles with everyday hackable objects. So we're talking about communities. If you've gotten interested in maritime robotics from this, there are a couple of forums on ROVs and ROVs are the box-style vehicles, but they also sometimes talk about the flight-style vehicles. Some pilots from Aero and Land vehicles have been adapted for underwater work. There are awesome competitions, mostly for students, but what's really interesting is they publish their papers so you can find the cheap solutions and the simple solutions that are effective and that students came up with creatively in those papers. There are two quite interesting challenges to watch. There's the microtransit challenge, which is about sailing the ocean on a two-meter-long boat, which has not yet been completed for the last couple of years it's been going, and especially the ones that start on the UK side usually don't even make it to the finishing line, which is quite interesting. Think about how difficult clearly it then must be to build a sailing boat that can autonomously navigate our oceans. And then quite recently, the Shell Ocean X price came out, which is basically about mapping a large area of the ocean floor in a very high resolution and a very short amount of time. So it's going to be quite exciting what happens with those two competitions. But before I send you all out to build your own robots, a few things about being responsible in this area. I do not know what the legal rules are for putting stuff into the sea. Many other people are not so sure about that either, so an ROV turned out in a nature-protected area and this is what happened. It was not from the beginning quite clear who would have to pay for getting the vehicle out there, who was responsible for the damage to the coral reefs that was done. So a lot of these things, especially about robotics in the sea, are not fully clear yet. Then just please do not let her think about how to get your vehicle back if things go wrong and think about careful testing before you rely on your system being autonomous. I'm used to using some materials that, for example, may have their sticker on them and then I normally use them in the lab, but if I think about putting them on a vehicle, I should check for these things and so should we all. There are issues with marine lifeforms moving from, for example, one lake to another invasive species. So before you go to a lake or switch between different experiment locations, look up these things. Look up how to clean your materials so you don't transport something by accident. And finally, something for you to consider when designing your circuits. Salt water is conductive. If you get your hands wet in salt water and then you touch your robot and there are voltages that are higher than, I don't know, a few tens of volts, keep in mind your skin resistance is significantly reduced and the only thing that protects you is the body resistance, which is significantly smaller. So I am actively looking for more people to discuss these robots things with. If you know about more places to listen to hydrophones, if you're building your own robot, please come talk, let's hang out on IRC, let's discuss things on Twitter, let's exchange more exciting competitions and events that are happening. So I have a few more minutes for questions now. That was really interesting. Thank you. We have a signal angel who hopefully has some questions and we also have microphones here in the room so if you want to know some more, come here to microphone one until six over there in the edge and we have the first question, please from the signal angel. Topsec Marx is asking, why not filling the robot with oil to prevent the crushing for weight control? This is done for some things so you'll find for example that batteries are filled with oil but you usually also want to modify things or replace broken things and similar things. So you can for example embed your entire electronics in wax as well, which is a lot less annoying and will not leak as much. But yeah, basically you want to take things apart occasionally. Do you have other questions on the internet? No. So probably people will come directly to you to talk about that, also email. Anybody in the room who wants to have their voice heard? Yes, over there at microphone four please. Hello, have you considered the Bluetooth sonars which are available for phishing? I've read that there are sonars which you can connect to over your phone I think which will speak Bluetooth. Yes, those are being discussed in the ROV forums as well. I'm currently more interested in a about one order of magnitude cheaper handheld sonar because I think those Bluetooth ones were around 400, 500 euros per piece. So as an order of magnitude sonars you start thinking in thousands of euros. The handheld fish finder cheap devices you can get for less than 100 and then the Bluetooth ones got significantly more expensive but there are people who are interested in using those. And then also as an order of magnitude the hydrophone that I've linked there is a bit over $100 so that's kind of you either get something fairly cheap but you have to put in a lot of work or you get fairly high prices very quickly. Wow, so we have microphone five please. Yes, you mentioned the extreme pressure on these steps in the water. So how do you test for it? Do you just throw it into the ocean or is there some way to simulate this at affordable prices on land? This is a very nice question, yes. So basic tests is just put it into a swimming pool there are training facilities for divers if you're lucky you can get into one of those. And then the next level is what we tested in was a self-built autoclave which is of course something that you don't really have standing around but maybe someone you know who does materials testing does and then more dedicated other really expensive test chambers at oceanographic institutes that you won't really get into. So that's all I know about testing but if there are creative other ideas. Thank you. And microphone one please. So these environments seem somewhat similar and salt mines. I mean some of these salt mines have very salty lakes. Do you know of anyone interested in exploring the salt mine lakes or do you actually test in them or like... So I've not come across this but I'm just someone who's looking into this area quite freshly myself. If you're interested in finding out I would probably go to the open ROV forums because those have a lot of people who use them and then tell the story of what they do is my impression. So it looks like an interesting thing to do. Absolutely. We have one last question from the signal angel please. Thank you. How do you navigate underwater since you probably can't use GPS? Mostly you just try and remember what you've done last. So you've got your local navigation where you use a sonar and you just try to recognize maybe landmarks that you know beforehand or just you react to what you've got around you or alternatively you try to have a really good IMU that will not make... increase the error too much over time so you just remember every single step you made and you hope that you don't get into any streams of water that will take you completely off course. And then there's an additional thing that you can do which is you can have... on bigger experiments you'll have a boat at the surface which will at the start talk to the vehicle when it's reached its goal depth and then it'll do some like 20 minutes of the vehicle moving in circles looking where the vehicle appears in the sonar of the boat and then you can kind of say okay this is your zero position of you now which of course is not something that you can do with a smaller vehicle but this is the principles that are mainly used. Great, thank you. Thanks Leela Fisch, everybody else can come to you afterwards or tomorrow. Thank you.