 Mae Llywodraeth yma o deithasio yn Blomain i Gweithwyr i Llywodraeth, ac mae'n meddwl i'r dda i uswght hypniadau coffacksig rydyn ni. Diolch. Rydyn ni i gafod diorelu'r gweithwyr i Llywodraeth, mae'r iawn yn fath o rhan o'r rhan o hynnyseidio'r gweithwyr gwybod llwyddiad Cymbl yw rydych chi'n rhan o'r Llywodraeth, a'r holl o'r llwyngau o dda, i chi'n mynd i chi fajaen. Ariwch. I'm electronic engineer based in Cambridge. I've got a Twitter that I occasionally rant on. The robot's got a Twitter that it occasionally rants on, less occasionally. And then we've got a Facebook page that we try and post updates on where we are with stuff on. You can ask questions, we're pretty reactive on it, so you've got any problems? Questions, whatever. So at the beginning, I watched Overwatch right when it started. The very first series with Jeremy Clarkson, which is a bit weird. And these terrible robots trying to kill each other awfully, but normally just breaking down. It took a bit of time before I got round to actually participating. I kind of started building robots, taking part, fighting robots around 2014. So this is once I've got a job and I could fully fund this terrible, terrible hobby that eats all your money. However, one thing can be said, everyone at robot events is lovely. They will help you fix your robot right after trying to kill it. They will suggest things to fix for the next time. It's great community, it's full of really lovely people, some of whom might be here today. And yeah, no, I'm really pleased that I got into it. So, as a result of getting into it a lot and making some good friends that are moderately local to me, we eventually thought to ourselves, you know what, it'd be kind of fun if we made a heavy weight. Because, you know, we competed with the smaller weight classes, so one and a half kilos, 13.6 kilos. So I thought it might be fun to try 100 kilos. I mean, how hard can it be? So, the reboot hadn't yet been re-announced at this point. So you have live events that happen around the country, so we're like, right, we want to do live events. So, but we were kind of concerned that RoboWars might come back because we'd heard some inklings of it. So we thought to ourselves, right, if you want to go on robot wars, you need to meet the looks unique. You can't just build a cheese wedge flipper, which is the predominant style in the UK at the moment, or was then. We didn't want to make a spinner because you can't run spinners in live events because they're too dangerous. Because, you know, they're spinning a big bit of metal round really, really fast and they drive into other guys who made a metal and then bits of metal fly off at like a couple of hundred miles an hour. That's not fun. So we didn't want to make a spinner, we didn't want to make a flipper, so the one we kind of came up with was a control robot. And after a lot of random sessions where we each egged each other on into worse ideas, we kind of came up with this. There were other ones, which were different. So we went with control because it's pretty rare in this country. Most combat robots in this country, as I said, were flippers. So we thought control robots were quite interesting because they'd be moderately unique. So that's what we started designing on. And this is a midway cad. It's not the very original cad that we started with. And it's not the cad we ended up with. But it's somewhere in the middle. So you can start to see some of the shape of it. So we've got folded side pods. None of us were too keen on welding at that point. So we were like, we'll fold everything. It'll be easy. And then we've got the claws on the top. The idea would be those would come down and clamp you and then the centre body would lift you up. And we had this weird idea that it wouldn't be fun if the whole centre body could do a full 360. But that took a bit of engineering to kind of try and do. So that was a bit of a painful choice. So we decided to call it Deadlock because the software engineer in our group of three was a lot more forceful with naming things. We tried a couple of names and then he was like Deadlock. And we're like, I guess. And he's like, no, it's Deadlock now. And it's stuck. So a Deadlock is a software term for when you've got distributed processes. And they all end up waiting for someone else to finish, but that someone else doesn't ever finish. So it's a Deadlock. And we thought that's quite nice because we're going to try and lock around someone and then make them dead. The robot, not the person. Just to make that clear. So I was going to say, yep, we clamped with the top jaws. They only exert a tonne or two, which sounds like a lot of pressure, but now everyone's made out of these really, really tough steels like something called hard ox. And that tonne's not enough to get through it. You need lots of tons. But we don't need to get through it. We just need to hold them so they can't run away. And then the whole, as I said, the centre body lifts them up, lifts them off the ground, and then we can do fun things with them. The goal there is to make it very fast. So we had a lot of fun getting the gearing for that right. And by that, I mean, we got it wrong a lot. And then we got it right. And then we broke everything else. So we did actually manage to sort the engineering out to get the centre fork to rotate all the way round. The reason, the way we did this was, going back a side, we've got all the motors in that grey body in the centre. And then the driveshafts come out really funnily into the red drive pods. So OK, so once we get the CAD, we were like, right, CAD is what, so computer aided design is one thing. You've got this screen, you're looking at your screen, you've got this model, everything fits in, you've got a place to put all the bolts, the motors all fit in, the batteries all fit in. But it's all on a screen, you can't physically touch it, you can't look at it, you can't figure out whether it'll actually go together how you imagine. I mean, the stereotypical one is a bolt that you can't physically put in because there's like a bulkhead in the way. So we thought it'd be a really good idea to make use of the local facilities at the local hackspace and laser cut the robot out of cardboard. So that way we've got an actual physical model that we can see, that we can touch, we can manipulate, and a glue gun is about the same size as a welding torch. So the joy of that is you can actually, anyway you can get the glue gun in, you can get welding torching, so that kind of confirms that it can be built because it would suck if you couldn't build it. So I made the cardboard model and from that we made a few changes. The notable one is we made it shorter by five centimetres because we felt that it was a little excessively long, and yet it's big for a robot, but you have to pick up other robots, right? So at this point it was about, season two had been announced and they were looking for people to enter. And we tried entering but we hadn't quite finished some of the finer points of the CAD so it probably wasn't a good idea that we got accepted. We didn't, so that was fine, bullet dodged. The second time around for series three we entered and we got accepted. And you get accepted and then you get told, by the way you need to be in Glasgow in six weeks. So it's like, oh, okay, we can do this. So we ordered the metal, which was not cheap, and a week later it all turned up and it all looked kind of, it all kind of lookedish right. But then as we started trying to put it together, because as I said it was all folded, so it should have all just kind of dropped in place, maybe neatening up some of the holes, but then it should have bolted together quite nicely. No, not even slightly. The supplier that we used had cut everything wrong and it bent everything wronger. So they missed some of the key important holes for bolts and stuff off, which is annoying because it's wear plate, it's really tough. We don't want the opponent to drill through it, which means it's very hard for us to drill through it. So that was very painful. And the other big problem we had was every single bit that had a bend in it, the bend was wrong. Every single bend was wrong. Some things had bent backwards, some things had bent in the wrong place, nearly everything was bent to a completely different angle to the angle we needed. So none of it fit together. So the only way we could get it to fit together was to literally cut it, grind it all a bit and then weld it back together again. And as I said, wear plate, this wasn't fun and it was very difficult. So, yeah. We managed to kind of get it somewhere there after a lot of pain and a lot of evenings of hard work, long days, long evenings. So as you can see, yeah, that's the metal robot. It looks very similar to the cardboard one. Because the hope was that we'd buy it, we'd get all the metal done, that would take a week to come back. Then it would take us maybe a week, week and a half to fit it all together, get it welded up and then we'd have the remaining two and a bit weeks to realise that we'd made a mistake and fix it to get the drive working so it could drive around, to get lift and clamp working. So we had all these great kind of plans that immediately flew out the window the second the metal came back wrong. But we did eventually get it looking like the robot in the CAD and we did kind of get bits kind of working like the motors would turn, the wheels would turn for a while and the lift would lift up a bit, the clamp would clamp a bit. So, and we drew close to the deadline where we had to be in Glasgow. So we tracked it on a pallet in the middle of one of my friend's lawns. We sprayed it grey, we then picked the pallet up, stuck in the back of a big lorry. That got tied down, the pallet got tied down in the back of the lorry, filled the lorry full of all the tools that we could think of, we could need and basically just the entire garage worth of tools. And we drove to Glasgow while the primer dried in the back of the van. So we got there on the Wednesday which is when you're meant to get there. We had a couple of days for our scheduled Friday on the Friday. So we spent literally every second, every waking second we could, working on it or thinking about what we needed to do the next day. So they opened up the big warehouse where the Robot Wars Arena was at 8 a.m. and we were there. We took quick lunches and ate the free muffins. Food was good actually, totally worth it. And then they kicked us out at 8 p.m. and that's when we left and we were down with the heads-in robot trying to get it to work, trying to get the drive to work for more than about five seconds to get the lift to actually move as fast as it should have been moving. And we were down as one of the first... Well, we were down to be in the first fight of one of the later episodes, hence the starting on Friday. And we were constantly like, yeah, we think we might be able to make it but every time we fixed one problem another problem would crop up. Sometimes they were masked by the previous problem sometimes it was just something that we didn't think was a big issue. Turned out to be a bigger issue. Finally, we had to go to the production team. Sorry guys, doesn't look like we can fix this within before Friday. And they're like, cool, right? So we got put back into reserves and we got replaced by a robot called Coyote. Lovely bunch of guys and a girl, great robot but they were ready and we weren't so they took our place and they went on and had a quite good time. So we kept trying to work in the reserves section but unfortunately that was, still couldn't fix it in time. Production staff throughout the whole thing were very lovely. They were all very understanding, very supportive. Our main point of contact was like, it's such an awesome robot. It looks completely different to everyone else's robot. It does the thing that no one else really can do. We're really keen to see you guys come back for series four if and when that gets announced. And we were like, yeah, we can do that. So we were all kind of like big set back but we can move on. So that's a little picture of deadlock as we were leaving the Robert Wars arena and you can see the pits and you can see the big grey primer. So we left sad, tired but kind of a little hopeful for the future because we figured yeah, we'll come back for series four because this is, Robert Wars is doing pretty good, right? And it'll be fine and we'll show our stuff. We'll look really cool and then carbide or killers later on. That's okay. We'll all have fun by that point. And then the BBC don't renew it. So that was pretty heartbreaking. There are still live events. I think the next one we're scheduled for is Steven Edge in mid-October, I think. So that's that to look forward to. We've still been working on it. We've still been trying to fix up the drive which has proven difficult. Trying to fix up the lift. We're quite keen. We think we've got the lift sorted now. The chain, the very complex chain path is all constrained, it doesn't slip. And we can lift about a ton in about a couple of seconds from ground up all the way up. So that's quite fun. We've done a few live events with it and it's kind of worked but the drive's been a bit spotty. We think that fixed that like a week or two ago, right? Yeah. Totally fixed. Totally fixed. Haven't heard that one before. But we'll see. We'll see. So we're going to try and get it running at Steven Edge. Try and have a good fight. Hopefully not get the stuffing kicked out of us. And once that's done, get it painted under. See if we can't go to, say, America. Because while the BBC's cancelled robot wars series four, America has just restarted their version of it called Battlebots. So they're currently showing series three. And hopefully maybe we'll get on their series four. Crossing the fingers. Also China's got a couple of TV robot wars shows starting up as well. So we might be taking it to that. Yeah. So any questions? I'll be outside. Or just try and flag down the white Sinclair that's zipping by at 100 miles an hour. Sorry, that was too fast. OK, we might have time for just a couple of questions. If you put your hands up, I'm going to throw this microphone over to you. My name's Lucy Rogers. I'm really pleased to see that there is life after robot wars for the robot ears and that there's plenty of things that you can be doing and potential. Any advice for an ex-judge? Well, I guess the first one would be there was Hebercon yesterday, which I'm sure could have dealt with a firm judging hand. I'm not sure on the status of that one. There are other live events going on around the country, robot combat related. But no, I'm sure there are plenty of opportunities for someone with as distinguished CV as a robot wars judge. Oh, of course. Yes, I forget. So one of the joys of being in Cambridge is there's a rather rabid Raspberry Pi community. And two mad selfless gentlemen called Tim and Mike run a competition called Pi Wars every year. And Lucy is one of the judges for this. And basically, there's just these, you build a robot with a Raspberry Pi as a base and then you have to complete a number of tasks. I can't quite remember what the tasks set for this year are, but when did the deadline for entries close? Next year, then. But it'll be awesome. Or there's one over this side from one of my teammates who's going to mock me mercilessly now. Hi, can you tell us a bit about the motor controllers? Specifically, you're ever going to actually finish it. So when we started designing the robot, one of the things we noted was in order to get everything packed into a small space to make it compact, we were going to use these big brushless motors, which are a little bit more efficient than the standard DC motor that you might be more familiar with. Because we're using big brushless motors, you need big speed controllers for them. And you either buy these 300, 400, no, 700-pound ones, which are 700 pounds, which is a lot of money, or you could try and get away with buying these 100-pound ones. Yeah, there's a reason one. There's a price difference. The 100-pound ones are not very good. And I foolishly volunteered to try and make better ones because I'm an electronic engineer, so I know what I'm doing. Yeah. How hard could it be? I'm starting to learn to hate that phrase. I still say it by accident, though. So, yeah, I'm still working on big speed controllers. Thanks, guys. Yeah, yeah, I know. Still haven't committed to a deadline, though. Do you have any videos here or online? Because I'm really anxious. So we should have some videos on our Facebook page. That's probably the easiest way of me trying to hunt through my collection of very poorly organised videos. Occasionally, we tweet videos as well, but the Facebook page is the main kind of thing we try and keep updated. The other one is the live events. Sometimes they'll put out videos of fights. So you'll see a great robot briefly fight and then break down in an arena with a poor lighting, but we'll get better. Thank you. No worries. Hi, Rob. Hi, Rory. So what lessons were learned from this particular build, Rob? Ouch. I guess the big one is... The big problem with robot wars is it's a massive investment for how much money you've got to spend on a robot. I'm not going to tell you how much that one cost. I don't think that's a good idea to say that. So it's quite hard to justify buying all the bits to build a robot before you've got on the TV show, but at the same time, six weeks to build an overly complicated, well, very, very complicated first-of-a-kind robot probably was a bit much, maybe? So, yeah, definitely... I mean, we started small. That's the usual advice for robot wars stuff. You start with smaller weight classes and work your way up, but yeah. What's your opinion on precision-folded hardhawks instructing a robot? OK, can we hear it again for our speaker, Rob?