 Yna'n gweithio'r ganddo, ac e'r gwasanaeth i ychydig i ddo'r meddliadau o'r bodai sefydliad a ddoch chi ddoch yn hwych ar leisio peirio a ddod o'r ganddo gweithio ar gyfer rhywbeth ei wneud a'n cael ei ddweud o'r arlanc, ac i weld i fynd i debyg o'r лeydd yma, i weld i'r wythig i'r meddliadau a i weld i'r gwneud i'r mwyaf a'r meddyl i'r ddweud a bydd y ddweud o'r ganddo'r grechion o wahanol. sy'n gweld i hyn yn gweithio i ddain, gweld i cyfsofftияu roi, gweld i ddweud gwahodol, i ddweud gwahodol sydd wedi gweld i ddweud gwahodol, ond byddai gwahanol iawn. Ond pharadwyrau bob sydd wedi gweld i ddatgani ardal, yn ddweud y prosecio awr, wedi gweld i ddweud gwahodol, ac mae angen i ddweud gwahodol efallai. Mae'r ddylch yn adeg. Rydyn ni'n ddefnyddio'n gwahanol. Rwy'n brydam i ddim o'r hillion gyda ph whiskanyddiadau a rydyn ni'n ddim gydig am y ddech chi. Rydyn ni'n ddim oed. Rydyn ni'n ddefnyddio ddigon yn gwych, ac mae'r unioniol wedi'u medriwyr mewn lleol. Mae'r neud ar y cwm Yn Mynd Yn Ym Mwy. Pwyr, gallwch. Echmont nawr a llwyddiad. Nid yn fawr i chi eisiau llwyddiad a zeddylu green. So robots, this is what we're here for, is really putting computers in stuff that you make. Here is the first example of a thing that was like these exhibits. I made this with my friend Ben for an exhibition in Bristol. It's a little eyeball that looks at people. That's all it does. That's its whole thing. It looks at people and it blinks. It's not really clever, but people seem to like it. So it's interesting to watch what they did with it. So they often, they loved it when they first saw it for, you know, five seconds or 15 seconds. They were very taken with it and they'd look at it and they'd play with it. They'd go and wave at it, try and work out what it knew and they'd move around and do little dances. Occasionally the cunning ones, if there were two of them, they'd like go different ways and see what happened. In this case it actually averages them so it looks in the middle which is a fail. So they were sort of trying to have a little theory of mind about the thing. Occasionally very few of them would go and look in the back of it. Hardly anybody did that. That's what it looks like. It's not as fancy as the front. But it was really instructive to see the way that people played with the thing and the way they looked for games with it. So it's sort of got me thinking. So this is what's in it. It's an embed which was a weird little ARM platform. I think it still exists. All I've stopped using it. It had a browser-based IDE and it knew the identity of all the embeds in the world and which ones you had and it could compile stuff and it went directly onto your device. I don't use any more. Anyway, a cheap serial camera, I think it was 80 pixels across or something. It's an awful camera and it just did frame-to-frame difference detection looking for where anything had moved and it took the centre of gravity of all the pixels that had changed and it pointed the eyeball at that and that was it really. It had three servos and it had one Christmas tree ornament. So it turns out if you want white spheres of a variety of sizes which you need if you're going to make eyeballs, you have to buy them at Christmas for people to put them on Christmas trees and keep them the rest of the year. I know no other way to get them. They are really cheap so you can stock up. So the trickery is it just uses the centre of gravity of the pixels. It does not solve the problem that the spectator thinks it's solving. It doesn't find people and look at them. It doesn't know anything about people or looking or anything. It's just finding the centre of gravity of all the pixels that are different so it's solving a much easier problem which is good because I'm lazy and with just one person, that's nearly always good enough. So it's a massive cheat. A proper computer scientist could spend a year doing this in those days but I didn't. Next thing is sketchy which really came about by accident. I made a delta robot out of three servos. A delta robot is this structure which has all the actuators don't move around. If you make an arm, you have to move lots of heavy engineering every time you move with a delta robot. Everything that moves is very light so it can be very quick and it turns out that servos are the cheapest actuator you can get and are ideal for making cheap terrible delta robots. Bless you. Next thing to do with this, we stuck a pen on it. In fact we stuck this brush pen on it and wrote some software. The box thing at the beginning is a bug. I fixed it after. Sorry. So all this does is we take a picture with a phone. We do some image processing. We extract a bunch of vectors. We send them to the delta robot and it draws them. But quite often it actually made plausible pictures of people. Usually it sort of actually looked like them. But usually. So people were very engaged with it much more than I thought. Everybody who saw it loved it. If I took it to make affairs, there was always a big queue. And I think it was because it was a weird thing because it was a robot that drew pictures that looked decent. But also it was a thing that for each person it was a little thing about them. It was a personal little interaction and they got this picture to take away. The robot had paid attention to them and they loved it. The bad thing about it is they loved it. So if you take it along to something like that you will spend all day, take the picture, explain what it's doing, watch it draw the picture, hand them in the picture. Then there's another one and do it again. So after about 300 you get bored of that. Implementation was an Android app. Edge detected the image, vectorized it, cleaned it up a bit. Sent it down to an Arduino which had a bit of motion control code. Very simple and just drove the servos around. And so the trigonometry to make the delta robot work which is code we found on the internet, downloaded. Never looked inside it. It was great. Thank you whoever wrote it. It was marvellous. So trickery again. It turns out all these things have massive cheats in them. So it's hard to distinguish people from the background. If you have a person against a brick wall it will spend most of its time drawing bricks. Because the Arduino we used was tiny we had I think a 764 byte limit on the packet that we sent down to the Arduino so we couldn't draw the bricks. The solution is to take a backdrop along so the picture is always the face and nothing else. Totally a cheat. And you also needed to rig the lighting so you got nice shadows across the face. So as long as you've got one eye socket and one side of the nose basically you've got a face. If you've got anything that might be a mouth you're winning. If you've got a chin you're done. It's great. So again it does a terrible job but it's emotionally effective in doing the thing that people seem to like it doing. That really is a theme for me apparently. The giant staring eye I built with Anton I think we actually invented it on the way back from Ohm in a motorway service station on the M25 when we were very tired and we were talking about how big an eyeball we could make and we realised it would have to be inflatable and then I was asked along to a social robot conference they wanted some social robots that were more interesting than the ones they had just to kind of show off a bit so we made this work for them and again all this does is it stares at people. There it is looking for someone. There's someone. It's looking at him very slowly. Oh still looking at him. There we go. That's the only thing it does. Oh yeah the pupil size varies depending on how interested it is so you can kind of tell when it sees somebody the pupil gets big. I don't think that showed in that video. So what do people do with it? Sometimes they have staring matches with it like with the little eye they dance around and try to work out what it knows. Occasionally I only saw this once but it was very funny. They wouldn't notice there was a giant eyeball looking at them as they walked along and then they noticed and then this massive startle. So you know it was emotionally effective possibly in a slightly unkind way. What's in it? There's a beagle bone. I don't know why we chose a beagle bone. It was what there was. There was a 360 degree lens stuck on a webcam. 360 degree lens is very cheap then. I think they're expensive now. They have some failed mobile app that used it. But it's just a lens that takes a whole slice of the world and distorts it into an image that the webcam can see so you can get all-round view from one camera which is cheap. Some Python OpenCV again we're looking for changed areas in the image but in this case we were trying to track the largest blob so you could track a person even if there were other disturbances around it worked. Three servos again and a blower to keep the thing inflated so turns out central heating boilers, combi boilers have a blower designed to run down a long duct which means they generate quite high static pressure which is what you need for inflatables. So when you throw your boiler away just keep the blower. I know a lot more than I used to about blowers for inflatables. The ball itself is cheap nylon sewn together I think it's ripstop. I made a spreadsheet and did a bit of trigonometry to work out the shape of the gaws and sew them together. It's fine. Those are the interesting works. So servos for which way it's looking and another servo to move that little brass shade which is the shadow of that is the pupil. So you move that nearer to the light bulb and you get a bigger pupil. The reason that it's so wobbly is because the servos are not really built to have large moments of inertia on them so if you move it too fast the servo tends to hunt. Also that was a really cheap servo. It's probably my fault. Trickery again. It needs a quiet background. You need to rig it so that there's only going to be a few people walking past it. You can't have a massive crowd in the background and apparently the slowness is forgivable. People really don't seem to mind that it's stupidly slow. I think because large things are allowed to be slow. I think there's a whole trick there to exploit with making things that aren't very good but they're huge so that's okay. Next up the tuner on a stick. This was at Shah. So all this does is spins around all the time. Literally it spins around all the time. There's no control of the spinning and it plays a tune based on where people are around it. So that each note is based on the range to the person when the range finder was pointed at them. The tune, every time it starts up from nothing that is when there's been no people it'll pick a new midi patch and a new set of notes so you get different music when you go back to it. But due to trickery it's always basically music. This got a very affectionate response. People would just be very puzzled by it which is fair. They would wave their hands at it. Once they knew it sensed them they'd try and get really close to get some confirmation it was sensing which is kind of a defect because it doesn't do anything when you're really close. That's a design failure. They'd stand about in groups. Sometimes they would try to play tunes standing around in groups which I thought was a massive win. But my favourite thing, passersby who knew it would come and visit. They'd just wander past on their way somewhere and they'd have a little walk around it and then they'd carry on. Just to have a little place. It had given them a game that they wanted to come back and do again. They weren't just interested in the mystery of it. They were interested in doing the thing that it offered them. Implementation, a windscreen wiper motor I think from an Audi. Cogs made from Lazer Cut MDF which made much better cogs than I thought. It still has the original set. It's five years old or something. It's a slip ring to get the power in. Turns out you can get slip rings cheap on eBay. So that's really just a thing with two halves. Wires on both sides. One side rotates with respect to the other. It just takes the mains into this whole spinning assembly. There's one range finder. It's an ultrasonic. It's reasonably nice, like a 30 quid ultrasonic. There's an Arduino. There's a midi shield. There's like a rat's nest of wires. There's an amplifier. There's a small eBay car amplifier and speakers from a charity shop. That's the inside. Oh, the drum in the middle is the bearings out of an old photocopier drum. Because that was the biggest bearing I had handy. No, cheap stuff. Relentlessly cheap. Trickery is, so there's musical trickery which most of the credit goes to Phil who knows much more generative music than me. So it always picks a scale starting up for a new tune and it picks a subset of the notes from the scale and it's only going to use those notes. So if you play them in basically any order it sort of sounds like a tune. It may not be a great tune, but definitely a tune. And it only plays notes at the times which it's decided are beats, which I think is eight per revolution. So it's always in time. It improvises a little drumline to go with it. So you basically don't get horrible not music out of it. So that is the best of the code, right? It's like find a range, work out what note number you want for that, play the note. There's some other malarkey around it but really that's the core of it. So, not fancy. It's the world of techno. This plays location dependent techno. So for every square foot of the planet there is some techno it will play there. Different square foot, different techno. That was it. It's here, but it's a little broken right now unless my friends back at camp have fixed it. OK, but thanks for trying. All I do with it is fur it up, leave it somewhere on a path. Somebody always finds it and they take it somewhere. They push it around. People look after it. These are the only instructions. Without those instructions people didn't quite get it sometimes. But that's enough. No. Literally I would be sitting there in a lecture and you hear it going go and pass. Somebody is taking it for a walk. So yeah, people would just wonder about with it. People would find a favourite spot where they like the techno the best and they sit there and drink beer. When it broke. So this was really instructive actually. If it broke people would try and charge it or they would fix it. Or they would bring it back very carefully and fix it. I didn't tell them to do that but because it was abandoned to their care they looked after it. So I've taken it to I think 2MFs and I can't remember if there's another one. So it's been in the hands potentially in the hands of many thousands of people by now and everybody has been nice. So I think there's another trick there not a trick really but this thing about you trust the thing to the people and the people will usually just be nice to it which is great. Here it is. I just left it there, this bloke found it wandered off didn't see it for like 4 hours. So what's in it? There's a USB GPS receiver there's a pie running sonic pie the music is actually the sonic pie acid techno sample with a bit of extra malarkey that we worked out a terrible car amplifier again in fact it's not a terrible car amplifier that broke and I put a better amplifier in it which cost £15 charity shop speakers, nice big squishy wheels and an obvious handle it's like maybe 100 quid maybe less there you go it's technology it's terrible so it's flat, 8 volts if you make one of these put some kind of voltmeter in it so that you know when it's not working that it's not working because it's flat rather than because it's broken that is yes here's some code so you see that stuff with lat int percent 1412041 what that's really doing is taking the latitude as a big integer I can't remember how many digits it is it's lots of digits so we're working out some of the bits so we're just using some of the bits to control the seed for the note generator for this little loop in Sonic Pi and then the next line is again choosing a chord using some other bits chopped out of the GPS coordinates so as long as you're using all the bits in the GPS coordinates to do different things you know that you get different techno for any different coordinates you don't have any control over what you get really you can, by choosing the bit chopping you can have some aspects very faster than others as it moves but that's enough people often suggest that I should make it play particular techno in particular places and I think while that would in some ways be awesome I like the simplicity of this kind of it just makes it up from the numbers so trickery it's got a handle and wheels and apparently that means you have to push it a designer would call that an affordance you put the right kind of handle and wheels on people will push your thing around the other thing which is not really a trick but is trust the camp trust the people with your thing especially if you made your thing really cheap mostly out of junk you already had but if you trust the people with your thing they will look after it and they will shower their affection upon it I think partly because they have been trusted with it I'd like to do more of these things that you just leave for people to do their stuff with to see what happens and again not really a trick but almost everybody likes cheap crappy techno it's just a thing you might as well put techno in your thing the giant tentacle so this the eyeball caught me thinking about inflatables and then I started thinking about how you would actuate them and this was a first go after some small experiments at actuating inflatable things which turned out to be quite practical so there it is being run off a joystick there it is a bit later after Robert helped me put all these rain sensors in it so now it plays little games it's either you can't really tell but depending on its mood it's either trying to chase people scoop them up or trying to get away from them one or the other little short range range finders who only knows when they're close and it doesn't know it very well but that's enough apparently it's eating Robert so people got the game with this so some of them did play chase the thing or be chased by the thing they would see what it was doing and they would go along with it they wanted to play the game some of them just wanted to squeeze it like a significant percentage of people just wanted to run up to it and give it enormous hugs so again it's fine some people thought that the end of it was its head and they would run up to the head and they clappered it and waved at it because they thought it would pay attention from that end so in my mind it was always more like a tentacle which has no single focus of attention so this is in some ways a design fail I guess that it didn't respond to the thing that they thought it ought to do if I was energetic I'd go back and try one that had a head and some sensing and see what other interactions you've got a lot of people wanted to talk about how it works as well it was fair enough I liked it so I was very happy too Implementation is one cheap bouncy castle blower they're cheap but they're noisy but they're cheap the valves for this stuff you have to make nobody seems to sell valves for this kind of big bore low pressure pneumatics so I worked out how to do it from bits of laser cut MDF and ventilation ducting from the hardware store compass sensors to measure the bend angles I tried putting sensors actually between the bits to measure the angles that was mechanically annoying and I think Tom in fact had the genius idea of just putting compasses in all the segments because if you know which way each segment is pointing you know all the angles so it knows where it is and it can go to a position it would like to go to and get there reasonably reliably time flight range finders because time of flight range finding is the coolest kind of range finding even if it's not the best and a Wemos an ESP8266 to run it all because it cost £4 and you can get these teeny little screens actually I put these in a bunch of things you get a Wemos and a teeny little OLED and then it can show you a picture of what it's trying to do while it's trying to do it and then when it's broken it's much easier to work out why it's broken and lots of sewing there's the valves the CAD for that is online somewhere I can point you at it if you want to make these things each valve sorry I should say each valve there has a servo in the back that is just turning a big drum that switches the air to go one way or the other trickery all right if you design that thing properly because it's all big rounded curved things that's quite hard you would need computers and CAD and stuff but if you just design a bunch of cuboids and parallel flat face things it'll probably work and it'll come out roughly the shape you expected and be much easier it doesn't matter if it's leaky it's not remotely airtight the blowers are designed to work with big leaky things literally you can put a whole that size in it and it works totally fine and again I think because it's big it's slow it's not very responsive but it is responsive so people tolerate the slowness they really don't mind this was the design process that's a piece of cardboard with some lines drawn on it which I later measured that was it that's not CAD this was just checking if it would work again with bits of cardboard how to build a robot was not my project it was the one I helped on it was built for a TV program it was a semi-puppeted semi-autonomous robot and kind of abandoned in the streets of Bristol and had people interact with I learnt a lot from it I'm not going to tell you all about it now because it would take two hours but it is still online there's a you can find the TV program and find many things again people really looked after this thing I mean it looked sort of like a little alien baby thing and I guess people look after alien babies they would talk to it they would give it these big tactile hug interactions very happily and they were looking after a thing that was abandoned to their care so it was interesting how to build a robot on Channel 4 it's still available on catch up if you want to Google for it I don't know if you can get any of that stuff in Germany I'm sorry if you can't the Trifid is here and currently also slightly broken but hopefully that will be working tomorrow this was EMF camp sponsored it last year thank you very much Johnty big mobile inflatable thing I wanted to try a different principle of actuation for inflatables from the tentacle totally worked this Australian lady loved it more than anything else lots of other people liked it but she really liked it this was a day out at Puppet Place in Bristol which is a big puppet workshop at this point it was manual and it was being controlled by that small child you can see with a controller on the right it turns out she knew exactly how to work it and how to attempt to devour her friends with it this kids were obsessed with it come and see it probably tomorrow it'll be working the sensing and interactions is a work in progress at that point it wasn't really sensing at all now it's got some passive infrared sensing on it quite badly I have a better thing which if it works should be really great we'll find out even if it's not successfully interactive as long as it's moving about people think about it and play with it and speculate about what's going on with it which is interesting because it's a big soft structure it can have quite a lot of complex motion even though it only has really two degrees of freedom and so it looks like it's doing more than it really is people like to play chase games with it if it ever comes down near them they'll either go and grab it or they'll run away from it people like to hug it sometimes they like to stick their heads into the spikes which is hilarious I'll be interested to see what they do when it's got some new stuff so you can come and see the works for this if you want it's one big bouncy castle blower the big blowers have slightly higher static pressure that is the pressure when there's not much flow which it turns out it needs to lift all the LEDs all the way up six valves each has one servo pressure sensors so that it knows the pressure in each of its three chambers so it can servo to target pressure lots of sewing so I got some friends Nick and Izzy came and helped me with the sewing for this it took a long time but I think it was worth it a lot of LEDs passive infrareds now sensing with the jets and nano maybe tomorrow we'll see these were just tests on the bend principle if you like prototypes to see how spiky you make the spikes to get the most bend out of that structure there's valve laser cut hips one servo about that big there's a fixed knit in a field fortunately Dirk and Idy have come prepared with sewing machines the inside which you can't see anymore so the bottom is an air box and then three valves let air into the three vertical cells trickery sort of not even tricks really people love big squishy things they just do if they light up they love them even more more of this kind of thing honestly you can't go wrong with big squishy inflatable light up things massive trick packs up small and then afterwards you'd have to throw it away or keep it in your house both of those are bad unless it's inflatable in which case it goes small so that whole thing goes into a box about this big which you can put in a car and take back to your house that is my favourite trick of all the anemone alright so one week before EMF last year Johnty said we've got a bit more sponsorship would you like to make another inflatable obviously a sensible person at this point the triff it wasn't really running so a sensible person would have said no but I didn't say no because I wasn't sensible so we made another one so the anemone was the simplest thing we could make that would kind of hit some of these points there is nearly full inflation so all this does is the pressure inside changes on a cycle up and down quite slowly and as the pressure rises those little tendral things extrude and as it falls they go back in and the whole thing kind of shrinks a bit as the pressure goes down and it just sits there and does that all the time oh and it's got LED's obviously gets lots of hugs gets people sticking their heads and arms into the holes I did see some small children climbing into some of the lower ones and climbing out it's fine it's really nice to be inside and sadly you don't really get the inside experience because there is only a tiny hole to get through now but it would be nice to experiment some more with that hole making in a flated spaces thing so it's very similar to the Trifid it's a big blower only two valves, the same designer valve two pressure sensors bungee cord to pull the nubbins back in we spend quite a lot of time standing inside it adjusting the bungees so that they pull back in nicely some sewing not nearly as much sewing because it's only really got about six parts plus the tubes we didn't even have a pattern for this really we just sort of laid the fabric out on the drive in a big multi-layer thing and cut the shape that we thought would be about the right shape and it was so no sensing, not interactive still gets interesting interactions by virtue of its kind of physical nature so that's kind of a lesson trickery, right? it just varies its pressure it's the easiest thing to vary with the bits we had on hand that's all it does it's still kind of enough so that's the end of the list of things what's next for me I guess is trying to work out more of the interaction design and the sensing for where people are sensing where people are is kind of annoying and difficult it shouldn't be but it is so I want to do that better the other thing is what these last two have taught me is that people really like these tactile interactions with these things so even though those weren't designed with it in mind and they don't really respond to it in any useful way people really go for it so that's a thing to work on I don't know what the right thing to do to provide that is but it's a thing that people want to have somehow so clearly I should do that so something tactile experiences maybe next year I don't know so some kind of conclusiony things making them cheap is all these things were made really quite shamefully badly if I was doing it for a job well I wouldn't do that for a job if somebody was paying me to do it you would make it nicely and properly and so it could be serviced and so it was definitely good enough but all these things were not really they were the worst thing that might possibly work really the worst I could imagine that might work to try first but pretty much always that has turned out to be enough to get an effective installation that you can take to a bunch of places even if it breaks a little bit and you can try it on people and it does the thing so there's something there about being bold about making your engineering worse and less flashy in order to get more stuff out into the world than in front of people so cable ties and glue gun making things from sewing because sewing is quick and big and using all of the tricks to have as little actual good stuff in your thing as you can have just as long as you deliver the effect that you want so that's all about cheap which I am very much in favour of and this is about good so a lot of these wouldn't have been imaginable without the things that came before there were two eyeballs there were four five I can't even remember and they all kind of built on each other so there's a thing there in fact Hannah's talk earlier had a point about this which said it much more eloquently which was that the people that do a lot that get a lot of good stuff out are generally just doing a lot of stuff the good stuff comes out of just doing and delivering even terrible things because they give you new bits and building blocks for what you're going to do next stuff and lots of iteration and lots of testing on audiences try it on people they won't do what you expect quite often if you had a genius idea that you loved that would make you feel stupid but that's the wrong way to feel because they've just shown you some other thing you can do they didn't do what you expected but they showed you what they wanted now you can do it so you should do it might be great so you've got to try it on they have to be naive audiences they have to not know the plan because you can't trust those they'll tell you it's great but the other thing is the experience makes a lot more difference than the object a clever thing is great but if the people that experience your clever thing in a boring way they haven't gone away with much so designing the experience it's much harder it's work but in the end there's a lot more wind from your thing if you design the experience people are going to get and then they get a good one or if you just got lucky and they got a good one by accident that's also a win but you can't guarantee that so giving people something to do something to learn from your thing something to interact with and come back to gets them more stuff out of the thing the other thing is a lot of these have a simple it's not big or clever it's not highly theoretical it's some simple attractive thing that makes it go people love big squishy inflatable things I know that now especially with LEDs they love bobbly techno they love to wheel things around finding those very sort of primal hooks to base your thing on mean that people will interact with your thing so look for them and they're probably things you like already you can't make things for other people to like that you don't like so work out what you like make things that just do that apparently there's meant to be a conclusion at the end so making an awesome thing is good but making an awesome thing with an involving experience is better in fact I probably didn't say it strongly enough there making a really shit thing with an involving experience is better in the end all you've ever delivered is an experience into some other human's head and that's really the thing that sticks in the world not the thing that you made the thing you made goes away but the experience hopefully is kind of sticking um and I think now that a lot of this is just obvious but I realised I didn't know it before I tried it and I put things in front of people and they did stuff and then I saw what they did and then I made more stuff so I tried it and I fussed about with stuff that probably wasn't going to work but in the end it kind of did so go make some things see what you get there you go thank you very much for your interesting talk and now we can have a look at your friends who are back there they did the fix they intended to did you fix it they do think so since it's most most engaging for the people here maybe put the Q&A first that's my let's do some Q&A so we have two microphones one there one there with people holding them and there's the internet person who's shaking his head if you have questions audience thank you did you ever try swapping the audience detection methods between say the two versions of the eye or the eye and the tentacles do you have comparing the frames of the serial camera method in the tentacle no I never did it's a good question but because the way the sensing is integrated into the stuff it's got a lot of work to do if they were better made there's more modular things you should try that totally try it that's one over there in the middle do you have any tricks for cheaply sourcing appropriate fabric for making these inflatables so I found a fabric that was £3 a metre on eBay in fact it's a six ounce proof nylon I can show you some we'll go and take a look at the triffid but that's the cheapest stuff I could find it's strong enough for these sorts of pressures and it's reasonably airtight it's not completely airtight but you don't need that okay any more questions wave like crazy wave like crazy the lights are very bright what does the internet say the internet says nothing at all can answer a question nobody asked but oh go on there's one thing you said you choose the cheapest crappiest stuff material that you can that you can find how often does it happen that it turns out that those things are too crappy surprisingly really rarely so I had to replace the amplifier in the world of techno because the connectors just broke off the board and it was too shit but it got through like three hacker camps before that happened so even then it's very rare that I just have to throw a whole thing in the bin because it's too terrible really isn't a thing it's kind of amazing even though I've been doing it for a while so yeah so that's another one yes so the world of techno how do you deal with the drift and GPS locations the techno changes I literally pay that no attention more questions from the audience there's one over here in the front all of these are quite nice and squishy have you considered making something that scares people no not really but now I think I should ok more from the audience more from the audience Jackman do you have a question well I have an answer so if you want to make things like this there's mostly source code and some CAD files and construction of detail online you can probably find it by googling for the names of the things or off my website or send me mail all of those were collaborative projects are we happy to collaborate on more of that kind of stuff or just to tell you about it in case you accidentally make something awesome so yeah send me mail then I would say thank you again Jackman and a big round of applause for him