 Thanks everyone. Thank you for GUI for inviting me to talk about what we do at Subnero. So my name is Chinmay. I work for a company called Subnero and I'll be sharing with you what we have been doing with the underwater theme of things. The talks before me were very interesting. I think set a lot of stage for the interesting challenges there underwater. One of the things that I always feel, because we live on land, we always don't think of the different things that change when we go underwater. And as we go and explore things, as we listen to the different talks today, we realize that things are very, very different underwater. And we need to come up with a lot of interesting ideas, like what Vignesh was talking about, different kinds of robots, different kinds of sensing, to be able to do things that we want to do, like exploration underwater. So what we do is wireless internet underwater. And the question is why? Why do you care about wireless networking underwater? Well, everybody talked about this. A large part of our world is actually under the ocean. And if you want to do things like measure, monitor, sense, or the robots we talked about, if you want them to talk to each other, or if you were talking about the coral reefs, if you want to go down and place some sensors there and try to find out what's going on, try to trace what's going on, we need to bring this information back to where we live and our internet and what we want to be able to see it and track it and monitor it. And then based on that, take actions or maybe send some robots down with it. And that basically needs internet connectivity. And one of the biggest problems is that cables, which is what we generally use for internet, when you sort of stretch that to the ocean size of things, it gets a little unwieldy. This is a ship that's used to lay cables. These things are massive. Laying cables is a huge effort. It's expensive, it disrupts a lot of the ocean, and it's a real problem to basically lay cables to somewhere deep inside at the bottom or somewhere far in the sea to be able to get any kind of network or power or anything remotely useful in these areas. This is a reel of cable in one of these ships. These things are tons, like multitudes of tons, super clumsy. And one of the biggest problems is they break quite easily. You're talking about robots and you're talking about giant sea animals. These things are considerably clumsy with respect to what the ocean can throw at it. So a breakage of cables is quite annoying. Then we think about it like, wait, we have wireless internet. Everybody has a phone in their pocket here and you're still connected to Facebook, Twitter, whatever. This works on LAN. Why doesn't it work in water? I mean, it's so simple. So we have not only does it work on LAN, antennas, which is what we use in the terrestrial Wi-Fi or wireless internet, it works in satellites. And in fact, recently we had transmission from Jupiter, which is thousands of miles or thousands of kilometers away. And you're still getting data from that. So it works very easily in air. So wait, why does it not work on LAN? That a cool thing about the internet we have on LAN is that you get, this is from my house, you get many megabytes of data per second. So it's really fast. It can go really far. It's really fast. This is the kind of network we are used to. But in the water there's a problem. Now, this is a graph I'll try to explain to you. It basically tells you how much power a radio frequency signal loses as it travels underwater. And each curve is a different distance it travels underwater. And if you just look at a black curve, which is five meters, only five meters underwater. A radio signal, let's just take a random frequency at this. A radio signal in five meters loses a bit more than 100 decibels, which is something like 50,000 times of its energy. So it basically reduces drastically in the amount of energy just traveling five meters. Point of this graph is to say that any kind of radio communication is more or less impossible underwater. The radio signals lose power tremendously fast underwater than they lose in air. In air they can travel to Jupiter and back, and they still have enough energy for you to detect it underwater. Just within five meters there is basically no energy left in that signal that even our best and the most sensitive detectors won't be able to detect this. So this is definitely a lost cause except in very, very specific cases where extremely low frequency signals are used, especially in submarines back in the day. So old military submarines used to use these, but there's a lot of problems with those, and that's not considered what's top, like, the best case research, and that's what people are doing today. So what are the options do we have? There's lasers, because, you know, why not? Everybody loves lasers. There's a problem with lasers. This is the kind of water you generally see outside the coast of Singapore. Is that true? So you have dived as well. So lasers don't really go very well in this. The water is way too murky. So even if you point a laser in this kind of water within a few meters again, you don't detect it anymore. So same problem like electromagnetic waves. Lasers do not work in all waters. If you have a very pristine ocean, you know, off the coast of Thailand or somewhere nice, yeah, sure, you might be able to use lasers, but they don't work in all waters, and the second water gets murky optical transmission is not a use case. So what are the options we have? Sound. Hey, dolphins use sound, right? This is like biomimicry. So wait, why can't we use sound? Sound for communication is something that, you know, has it been done before? Well, do you guys remember this? This is telephony over, you know, sorry, this is networking over telephony. This is something that we have always done, and this is something that we are very good at doing. We know how to do this. We have deployed it over for the entire world, and it has worked. This is how our internet used to work back in the day for those who remember. There might be people who might actually might not have heard of this before. Anybody hasn't heard of this sound before? All right, cool. All right, I don't have to explain this. If you go back for a second, this is a cool diagram. If anybody wants to go check it out, somebody did an explanation of what each of the different parts actually has. I always listen to this sound and I know this tone, but this explains what each tone actually does, which is pretty cool. If you're kind of nerdy, it's a fun website to check out. But it's not so simple. There's a lot of challenges using sound to communicate underwater, and I'll go through some of them. One of the problems we have is bandwidth. I showed you the meter earlier with my house getting 89 megabits per second upload rate. That's the kind of stuff you get in wireless networks today. We are nowhere near that in acoustic networks underwater. We get much, much lesser. This is something I was plotting just a few days ago at work, and it's a model of how much data you can transmit at different frequencies. Again, the thing to note is at something like 200 meters, you might get about 100 kbps if you're really lucky. This is again a theoretical model. Practically, you have to discount a little bit of the data for correcting errors and doing other kinds of things. But look at something like 500 meters. You can barely get 3-4 kilobits per second. This is the kind of maximum data rate you're looking at. We're not looking at underwater communications, at least for now, for using Instagram and watching YouTube while you're diving. Although, I was just talking to some people and they were telling me about how, depending on how deep you go in diving, when you come back, you have to do these compulsory rest stops. They're bringing magazines to read while they're coming up because it's really boring. It takes a lot of time. They're saying if this stuff works, YouTube videos would be a great thing to do on the way back. We're not there yet. Maybe someday we will be. But for now, this is the kind of data rate we're looking at. So there's a lot of challenges in trying to get this to work. But maybe with some new technology, some new ideas that people are working with, we might be able to figure out. The other big problem in the ocean is noise. The ocean is really, really noisy, especially for acoustic. Many animals use noise. This basically ships in the propellers and we talked about how it's horrible that they end up destroying or damaging some animals. They also produce a lot of noise. But trust me, one of the most annoying noises we have in the sea is this shrimp called a snapping shrimp, which around this area of the world we have a lot of. And correct me if I'm wrong, but what it's doing is it's snapping its claw really fast to make a bubble, which then cavitates and causes a shockwave, which then stuns its prey and then it goes and eats the prey. So it's basically using this bubble cavitation mechanism to stun its prey. The problem is when the bubble cavitate, it makes this super loud sound and it sounds something like this. If you can play the sound. Some people describe it as eggs frying. It's a very appetizing sound. And this is a waveform of how it looks. So for most people, you'll be like, okay, that's just another audio waveform. But for people in communications, they realize this is the worst thing you can ever have for communication because this sound is basically affecting every single frequency there is in your band of frequency you're trying to transmit. So whatever frequency you're trying to transmit at, you know, earlier we saw these graphs and we're like, oh, if you choose this frequency, you might get more data rate. If you choose this, you'll get less. Well, the snapping shrimp is going to destroy or affect or put noise in every single frequency there is, which means it's super annoying to sort of work with these around. And there's thousands of these in our waters, and they always snap all the time. So when you go and put anything in the water, the receiver is always going to listen to hear this plus whatever you send, which means it has to somehow figure out how to filter this out. So this is another challenge that people are going through. Snapchat, that's a good idea. The other problem that we don't think about enough is delay. And one of the things that's sort of very fundamental is physics here. So we use radio for our wireless communications in the air. And radio is really fast. It's electromagnetic spectrum, so it travels at three times into the power of eight meters per second. Very, very fast, very big number. Sound is much faster in water than in air. It's about five times faster in water than in air, but it's still not that fast. So if you think about it, if you are trying to converse with something that's one and a half kilometers away, sound is going to take one second to go and come back. Now this is in networking, this is the concept of latency. If I say hi, how long does it take for you to say hi back? If we are one and a half kilometers away, the least it's going to take is two seconds, right? One second to go and one second to come back. That's at least two seconds before I hear a response from you. Now if you go to a website and you click a button and you don't get a response from it for two seconds, it's really annoying. You're like, hey, why does it work? So there's an inherent issue with latency. It's something that we don't think about. If there's any gamers here, I'm sure the gamers here care about latency because when you're playing video games and you try to click on that button to shoot someone with whatever counter strike and you're too slow because your latency is too high and you hit the network for being slow, that's exactly what latency is. It's basically the time taken for your signal to go to the other side and their response to come back. And that gets really, really annoying because a lot of our networking protocols and the way we do networking is designed for electromagnetic waves and it's designed to, assuming, very low latency. All that knowledge that we have built up, all that technology that we have built up has to change because now physics has changed and now if you're using sound and water, we have a lot more delay. So this is what I was talking about. Like, if you send something, it takes some time to come back. So this is time going that way. And we have to figure out how to change our networking protocols and the way they work to take care of the delay because in the time in between nothing is happening. That's a waste of time as well. So you're not only having more latency but you're also losing the amount of data you could transmit. So there's a lot of interesting challenges in this part. It's a very new area. There's a lot of research going on in this area. In fact, some of the work that's been done here has been done at a research lab at NUS, which is who we collaborate with a lot. So there's actually an interesting thread throughout a lot of speakers today. There's a lot of NUS connections. They have been working a lot with this and they've been trying to develop a lot of ideas around how to solve these issues and how to make underwater acoustic communication a viable system for a bunch of use cases. So that's what we do at Subnero. I'll talk about a couple of things we do to mitigate some of these solutions. It's basically just taking all the networking knowledge we have, choosing some of the best bits of it and using it and then reinventing a lot of really interesting ground-up things. So we try a lot of custom software. We do a lot of custom networking. So there's all the network stacks and all the, you know, if you've heard of TCPIP, a lot of the basic ideas that were based in those, you have to sort of go back, re-observe them, see if they make sense, otherwise, you know, put in new ones there. So we work on things like that. And we also do a lot of custom hardware. So these are some electronics boards that Shan here has built. So basically, because a lot of the off-the-shelf stuff doesn't work here anymore, we have to build our own, design our own. So what we make, or finally, are our modems. So this is how they look like. I thought you would like to see what they seem. They're really big now. They're not as tiny as, you know, these guys. But hopefully, after a few years, we will have, you know, much more better technology and we will be able to miniaturize them to something that's a lot more smaller. But if you want to, you know, snapchat in the ocean, for now, you have to carry one of these as a diver. Not the best, I would say, but maybe someday. So some use cases that what we are envisioning is a connected network underwater. So, you know, vignacious robots and your coral sensors could all chat with each other, not snapchat, but just at least communicate with each other and help relay the data to, you know, you don't have to go diving all the time. You could just sit in your office and look at what's going on and how the corals are doing or how the robots are doing and change them. Or even, you know, ships with, you know, trying to figure out if there's too many fish in the area and trying to navigate away from them. Basically, the second you have network underwater, you can do a lot of things that you do on air or on the earth underwater. The other thing, you know, we talked about the OpenROV earlier. Hakim Shoras is OpenROV, which is super cool. One of the things if you noticed, Hakim's one didn't have it, but they always got the cables and the idea of remotely operated vehicles is that there's always a cable attached and cables are really annoying. We talked about this earlier as well. There's always these tethers, so this is a new version of the OpenROV that's coming up. Hakim, do you want to build this? I'm just curious. But it also has a cable, but imagine if you could actually do this without a cable. If you could communicate it with it wirelessly, tell it where to go, get video data back from it. That's something that we are working towards and hopefully one of the days we will be able to do that. The other thing is, you know, these big installations, whether it be oil rigs or any kind of big structures in water, maintaining them, you know, looking at them. You know, we're talking about biofouling earlier. Cleaning them is quite a big hassle. Having sensors to know whether, you know, if they are broken, if there's this fouling, if you need to go clean them is something that's super useful. Again, if these sensors could, you know, communicate wirelessly without needing cables, that would be super handy as well. So these are the kind of use cases we are looking at. And that's all I had to share about. This is how we try to communicate underwater. We are still a young startup. We're trying to figure out what we can do. So if you have any ideas, queries, questions, you can come meet me or Sean later and we can talk about, you know, network communications and underwater stuff. Thank you. Thank you so much.