 This is literally a follow on. I think he was asking the question and I'm trying to give an answer in a silly way. So we have, you know, we have things like this, they fly around and they crash and sometimes they're going to planes as well. So what do you do? How can you actually control them? So people have been thinking about it quite seriously and there's a lot of money to be made. So if you look, you know, the first one, Airware is an XMIT guy. He used to be a pilot. He left his job and he raised serious amounts of money from ECs and I think now he has like 20 or 30 people working for him. They have a piece of hardware and then a bunch of software that platform is there. That's true of Skyward and Drone Deploy as well and of course NASA, the guys in MS has actually got like a project looking at airspace management. So for me the question was can you actually try to do this? We thought much money and limited amount of hardware. So like our previous speaker, I've been thinking around quite a long while and I got into this business like really early when 8 p.m. came out and used to be on the other company for a while. So I have things lying around the house which is like GPS light poles and stuff like that and I thought okay, can I actually try using first GSM and somebody in the audience actually gave me some vets to connect it back. I want to say who it is and if you recognize the person, you know who it is and I tried that and it failed. Then I went okay, this is the next iteration. I should try doing a bit more bandwidth. Essentially it's like I don't get the data back and it's just dropping off so I don't know what the hell is happening and then suddenly it goes somewhere, disappears and reappears. So how can I actually improve that a bit? So I went okay instead of actually having a GSM kind of thing, maybe I should try a bit more bandwidth. So I went 3G. Okay then I was tinkering around with my choice of flight controllers and initially I was thinking of 8 p.m. mega and then I went okay. It's a bit harder. I mean as more later versions came out, the space left on it was pretty much none. At 2.8, I mean 8 p.m. 2.8, you could only run like other copter 3, 3.1 wouldn't even run. So I went okay, if I need to allow other people to tinker with this, I need to use the platform that supports whatever that's available. You should be able to download it straight and then play with it. So I went okay, I'll do big shock. Yeah, this is the picture, kind of what he was kind of describing. Essentially it's the same stuff. All the things he was describing in very great detail. I was trying to describe it in slightly less detail. The bit in the middle is the bit that actually controls all the map in there. The extra bit that Gavin didn't have where the XP bit essentially is sending the data out and taking it on the other side. I mean I'm not really using that. What I'm trying to use is like a maverick kind of thing where I can send the data out and I can actually receive it somewhere. Okay, so this is a standard thing that we tend to use as in the as in the people from the other copter community and other people to just plan the flights. So I took this and I went okay, can I actually take this apart and then make it into like three two or three pieces and then use that to actually control multiple drones. This is an experiment. Okay, so everything is an experiment and sometimes it fails. So this is the attempt that was done. So this was what I actually talked about like a long while back when I was given the thing by the person. So I took a big shock. The next iteration, the first after the first failure, I connected it up. I put some way points, send it out. The offense debt, it failed. And I came back and I went, okay, can I do this slightly better? So when back to big shock, change that, I got a 3G modem. So I can send some data out, plug it into a Raspberry Pi, connected it back to big shock and then send it across IP to something, which is like a server, which I'm going to show and had some head into it. So there you are. So it's like in a real, if I had lots of time and nothing else to do and lots of harder lying around and if I had the ability to write whatever code I wanted to write, it doesn't exist. I would really like to have like security mechanisms. What essentially what that means is like I should be able to bootstrap a firmware and then check, is the firmware valid? And then check the positions as I'm way starting. And then I should be able to either prove to a third party where I am and my calculations are right. And at the same time, I could even prove to a third party that I am within a, if I'm within a geofence, I can prove that I'm within the geofence or if I am outside a geofence, I should be able to prove using the normal protocol that I am outside of it, that I haven't entered the space. It's just kind of interesting. Both cases are interesting. If you are inside, you're not leaving. It's kind of interesting where you won't actually bound it. If you are saying, you know, you shouldn't go into an airport or any areas like that, that the other way around is important. And the thing is like, you know, if you reveal information about, you know, where you are, then there could be lots of instances where you could reveal information about what you're trying to do when things like that. It might or might not have implications. As a Psyphepank, I would really like to have, you know, as much control as possible. When somebody wants to know exactly where you are, give them the information. You don't have to tell them what you're doing. That's the whole idea. So, yeah, that 3G link is kind of a tricky thing. The problem is that it's, how could I say? You can only try it in places where there are proper connections. So it's kind of tricky. You can't go to some place where, you know, you just have GSM, more stuff like that, and try to fly it. Oh, if you fly it really high, it just fails. It straight fails. And it's really, it doesn't have a backup plan in that sense. You can do an RTL, a return on land, but beyond that, you don't have much of a backup plan. So, yeah, so the survey, I really wanted to implement that bet, but I never got to it. If somebody gets really interested and wants to implement, I'm happy to help. Come talk to me. I can give you the source code. I can give you some bits of hardware, so the known broken bits, and you can go give it a try. So, okay, so this is how I set it up. So, okay, I'll show you the thing. So it's literally, you know, bind, okay, then the next thing is like literally bind it. So literally bind the whole thing app. I have the whole thing, essentially I created a whole set of steps to how to bind the whole thing up. Then you can create boxes, and once you have the boxes, and when you fly it, you can actually prevent them from going into bed. That's essentially what it was. What I can't do right now is like a failure can actually make this thing go into the bound. What I mean by that is like, if my hardware fails, I don't have any mechanism to prevent them from, you know, taking a path into the thing rather than out of it. How do I describe it? So I don't have graceful failure. That's what I would describe. My failure is real failure. If it fails, it fails, and it goes wherever it goes. Right, so yeah, that's essentially what it was. It's like I took the other copter thing and put it into the hardware, and then, you know, ran this bit. Well, this is how you bind it. So this is like a Raspberry Pi that I was stuck in, and I just wired it up backwards. And then what happens is it's like, you know, I can send the messages across. So it's, you can talk to the pick sort directly. As if you have a, you know, telemetry link. That's essentially what you're trying to do. And then what you have is like, from your platform, you're sending out the messages. You're saying, okay, you know, go within this, and whenever you go into the box that you're bound, you say, don't go in. Practically as simple as that. So yeah, that's the picture. Somebody help me with this one, okay? Literally, I can't do anything. This isn't looking. Okay, that's all the rubbish bit of code. Yeah, that's more or less what I did. So I have the source code. It works. I've done simulation, most of it. Done couple of flights. Again, I always have a human at the end of the loop because the code is written by people like me. I don't trust it. There you are. Feel free to ask any questions. I'll try to answer.