It looks like you're having connection troubles. Reconnect now.
02/04/2024 17:17:42
37c3-talk-57905 Latest text of pad 37c3-talk-57905 Saved Feb 4, 2024 |
Hallo Du!
Bevor du loslegst den Talk zu transkribieren, sieh dir bitte noch einmal unseren Style Guide an: https://wiki.c3subtitles.de/de:styleguide. Solltest du Fragen haben, dann kannst du uns gerne direkt fragen oder unter https://webirc.hackint.org/#irc://hackint.org/#subtitles oder https://rocket.events.ccc.de/channel/subtitles oder https://chat.rc3.world/channel/subtitles erreichen.
Bitte vergiss nicht deinen Fortschritt im Fortschrittsbalken auf der Seite des Talks einzutragen.
Vielen Dank für dein Engagement!
Hey you!
Prior to transcribing, please look at your style guide: https://wiki.c3subtitles.de/en:styleguide. If you have some questions you can either ask us personally or write us at https://webirc.hackint.org/#irc://hackint.org/#subtitles or https://rocket.events.ccc.de/channel/subtitles or https://chat.rc3.world/channel/subtitles .
Please don't forget to mark your progress in the progress bar at the talk's website.
Thank you very much for your commitment!
======================================================================
good evening everyone and welcome back to fire shunks before we start uh a quick block of announcements you can send us questions requests or memes on Mason by posting uh with the # sh23 and public visibility you can send us questions on IRC uh in the channel sh23 on ic. hack.org we've got a uh social viewing experience in the aquarium on jitsy uh the link for uh to that can be found on fire.org uh we have a YouTube and a twitch channel uh under name fire shunks uh where we also have a live chat where you can ask questions that we will pass on to the speakers uh our main streaming platform is streaming. media.ccc.de um where we also have a relive where you can rewatch past talks in case you missed them um we're a bit difficult to find on 37 C3 that's because we are about 30 kilm under the parking garage please bring your own drill we're happy to present the talk encapsulated electromo electromyography with mayo and Ras raspy by C Chan and Alex Le cchan is a hacker and CS person from renela Polytechnic Institute they are now studying at the univiversity of Maryland where they are very good at computational text analysis and resuscitating abandonware Alex leech is an artist educate and technician they currently serve as assistant director of the HCM program at the University of Maryland where they have taught courses on programming interaction design and Fabrication since 2019 their recent projects include interdisciplinary outdoor installations like aqual a capacitive Stone installation and Fiddle hex let's talk 10-year-old Tech de Maya Al was one a really strange way to control a computer and then became a way to do fine Grand biom muscular electrical detection research this is a talk about how to make a Mayo how to hook a Mayo to a Raspberry Pi freeb in 2023 and from there how to have the armor communicate over serial to other devices hello welcome to my talk for fire shunks 2023 encapsulated electrom myography with mayo and raspy uh pneumatic sculpture with invis
ible controls this talk uh is brought to you by me Alex leech the talking person and my collaborator C Chen who was the lead programmer on the project so introduction who are we C Chen is a hacker and computer science person who is responsible for the python script repair and the myof functionality on the Raspberry Pi my name is Alex leech I am a sculptor and a human computer interaction person my job is to build actual sculptures that move uh so I do the mechanical engineering and to do the talking on the project as well as to come up with some of the debugging for the scripts we work together as a team to put this together in the fall of 2023 as you can see it is a project that involves a lot of pneumatic balloons made of disposable objects so this is an emotive sculpture project the goal of the project was to make the audience feel something and the secondary goal of the project was to fix up some abandoned wear that we had wanted to work with for a while so I have had a bialic myo armband since 2014 when the project was still kicking and was still an independent project uh and I wanted to see whether it could be put together so that it would fit in a small beautiful display package the actual sculpture was inspired by Soft Robotics and some artworks that elicit reactions from the audience so there's a common thing in HCI where people are like I want my project to make people feel something I want them to laugh and be cheered up or I want them to be you know I want to be able to discuss them I can read an expression and show something that will change people's feelings but we have a whole discipline that does that it is called Art so that is what this sculpture is for we wanted to do this by encapsulating some Hardware in this small box you see here so this box is laser cut and these parts are all made of disposable stuff in order to really emphasize the DIY or rapid prototyping uh approach to project production uh and what we went with was the again phalic myo a
rmband uh loaded onto a Raspberry Pi model 3B plus and we wanted to work with this project called uh programmable air by a guy named Tinker mind at ITP NYU so the programmable air project is part of a Soft Robotics prototyping toolkit which can be used for things like grippers or uh mostly grippers actually vacuum holders or also aesthetic devices which is what we were interested in using it for the key elements of this project which is a kinetic sculpture are that it is made of disposable elements or trash in order to capture that DIY rapid prototyping aesthetic that means that it was designed around 3D printed parts that were intended to use bamboo skewers or coffee St stir sticks or uh other types of common work in order to assemble a thing that was greater than the sum of its parts this is also why we went with condoms because when we were working with latex balloons we found that they popped a lot where condoms generally just do not break if you do not do something really absurd to them they can stretch and then recover their shape really effectively uh and they also work really nicely with themes of feeling or sensation because their form obviously gets to mimic really organic shapes all the time this was also inspired by the corpse flower the Titan arum which is a flower that blooms about once a year and it smells very strongly of Decay compost so we wanted to bring all those themes together to make a sculpture that both rose up and stretched towards the surface of the Earth or away from the surface of the Earth and then collapsed backwards in a really dramatic way but without any visible interaction with the sculpture itself we want it to do this sort of off from to one side from the person who is performing the control of the sculpture so in order to do that we needed these controls to work pretty seamlessly we needed them to work like magic the idea was that when you did a gesture in the armband the sculpture would respond with invisible or seamless contro
ls so we have pneumatic we have pneumatic actuation within the actual sculpture when we fill it up with air using the programmable air but the controls have to come in from somewhere else and they have to be kind of hidden the seamlessness contributes to the magical effect of uh suppressed infrastructure infrastructure that's hidden under the surface of things and in the meantime the Arduino platform inside of the programmable air maintains peristalsis which is uh an internal pressure for the breathing effect in the sculpture itself we chose the myo armband because it did not seem to be a great choice it had been sitting on the shelf it has a lot of challenges so a little bit about the thalmic Mayo project uh the phalic Mayo came out in and around 2014 it was out of waterl Ontario the company that put it together put it together entirely on their own inh house which was exciting at the time a lot of people were prototyping hardware and then the company got bought it got bought by company called control labs and control Labs got bought by facebook/ meta which means that you can now buy myo armbands but there isn't a lot of support for the myo armband software packages that go with them they're all pretty old they're about 11 years old they're not out of date exactly but it can be tricky to find them and with the internet being the way it is there's no guarantee that that software is going to keep being reliable as we move forward into the future so we thought that was an interesting opening for some hacking or at least to figure out how people are working with this the programmable air prototyping project by Tinker mind is great so this uh is a little project for making sure that soft robots can get tested out in a reliable and consistent environment it comes with two pumps one for suck one for blow you can set it up to do valve controls and differential pressure controls it has a little pressure sensor and an indicator as to whether to how intense the vacuum or the
suction or the inflation is in the project and three valves to be able to control three different ways of putting that pressure into a pneumatic system and the thing is this is actually great it mostly just works straight out of the box it works as depicted although most of the examples are uh a little bit less well-maintained than one might want I do not really feel the need to explain a Raspberry Pi model 3B plus to this audience but in case you're wondering we did have to go buy a bunch of new SD cards for it uh it seems like the Raspberry Pi software needs a lot more than 4 gigs to run these days which is nice for it I think so here are some caveats about working with a piece of abandoned Hardware or semi-abandoned Hardware or like Hardware in some gentle forms of Decay the first thing is the Mayo has a lot of caveats to how it works it takes a bunch of time to get used to what it is doing it has to be plugged in to charge it has to be unplugged to turn on so in order to work with it what you do is you plug it in you plug the dongle into whatever platform you're working with then you unplug the Mayo from it and you leave the dongle plugged in and then it can pair the control code that you're working with the python control code by default the myo armband will lock up and stop receiving gesture inputs if you double tap so you have to edit that out of any Library you're working in in order to make sure that your gestures don't accidentally turn it off uh as mentioned the Bluetooth dongle absolutely has to be plugged into whatever platform that you're working with otherwise you will be reversing how to pair one of these to a main computer it's simply much better to not lose that dongle to start with and then the last one which is really interesting the Mayo stores all of its gestures on board so all gesture registrations have to be pre-built in a Windows laptop with the abandoned thalmic software thalmic uh myoc connect uh this is important because the Mayo armband
by default comes with a library that was tuned on some pretty beefy forearms some uh some large masculine dude's arms and so if you have smaller arms or if you have much beefier or forearms your Gest won't register the same way into the myo armband we also had some problems when we were using this for performance uh where a differential a differing stress level in the presenter would cause gestures to be recognized in different ways so that was really interesting and that was one of the caveats we found from working with it it is still really reliable I would rather use a myo armband than build my own gesture recognition armband I do not really want to build my own electromyography project and I think this is a good answer to to that the build quality on these is really solid it's probably worth 150 bucks to not have to build one yourself but working with it is a really particular experience and we probably could put more effort into making that uh work properly for the next time we worked with this so how does it work once we turn that armband on what are we actually doing so over here we have the myo armband it pairs through Bluetooth into the Raspberry Pi 3B Plus using our custom myoc serial. py library that we wrote uh we wrote we wrote a script to control it uh the Raspberry Pi once this is turned on myoc serial. py detects over the port 9600 sorry over bod rate 9600 that the Arduino Nano is also turned on and listening so all three of these things have to be listening on 9600 we tried listening on different bands but it seems like the armband itself really prefers 9600 as a b rate and so it is best to Simply set up whatever you're listening to over here so that everybody can listen on the same B once it's turned on and going this piece of python scripting detects whether there is an Arduino Nano available and then it will send the Arduino Nano poses that ad hears from the myo armband as they come in so the myo arm band detects poses like fist or finger spread
or any of your gestures and then it pushes those out over the serial port to the Nano which is also listening on that band what did we hack to make this work so we did not write this code raw from scratch although we kind of did so my collaborator C wrote most of our scripts directly in the Raspberry Pi in order to make sure they were working using the Nano platform there are other ways to do this but it was ultimately simpler and somewhat easier uh given the restricted number of USB ports available when the Mayo is plugged in to do this directly inside of the system and then extract the code out afterward words to figure out what we had to edit we went really spelunking into the depths of the internet we found a 6-year-old Library called Raspberry Pi by shaan on GitHub which was based on Zoo's Mayo raw output work almost every piece of mayo work that exists that is not based directly on thalmic labs's work is now based on Zoo's myor raw outputs it's a very legible project and I highly encourage you if you want to work with this to work with that directly but we found it easier to consider what shaan had done in order to put together a myob based control for a Lego C where it turned left and right so by comparing the two libraries and reading them back we were able to write our own to output something that would control an a Ras that would control a programmable air through the Raspberry Pi we also had to go into the programmable A.H library because it automatically turns on on port 115200 in order to listen on a serial Port so we had to remove all calls to the serial Port from the base library in order to make sure things were normalized to uh 9600 bod on the serial Port I keep saying Port because my usual job is just working with the internet and it is difficult to remember which words to say one so here we go myoc serial. py we store the pose on the band itself so this was the most interesting Discovery was that once the Mayo is properly tuned and calibrated you
can turn it off and leave it on a shelf for months and it will keep those poses it will keep what gestures it expects from your muscles and what those gestures turn into and outputs them into uh a readable serial call the myor also shows a b rate of 9600 for serial Communications so this was an early block on working with all of these libraries together was they would turn on properly but we couldn't figure out why they weren't talking to each other and it was because their serial port preferences or their serial bod rate preferences were buried inside the libraries and themselves so a big piece of this kind of work is elevating connections that are hidden inside these libraries and outputting them properly from there we convert these in andout commands so that the Arduino listening on that same B rate can translate them into an actual action this was a major piece of work like putting together myo seral figuring this stuff out was a pretty major and serious piece of work and yet it doesn't seem so serious when we talk about it so here's a demo video of this working and let me walk you through what is happening here so here we have uh my collaborator C is spreading their fingers which causes our robot to collapse in a really dramatic way then they're resting for a second which kicks in the peristalsis and then they're clenching their fist at which point the robot comes up and becomes very buff and and and inflated and strains at the axis they relax their fingers and we can see the robot drops off again just collapses again into a mass of sticks and condoms and nothing and then as they relax you can see the peristalsis kicking in and then when they fist you can see that that pops the robot back up again with the strain of of the Fist so that seemed really aesthetically compelling uh I thought that was a really big success but the big trick is that this effect this armband and gesture can happen when the person isn't in frame with the robot so they can up to 30 ft aw
ay which is very convenient when you're trying to put together stage effects or special effects or room scale effects or secret theater effects uh you can have somebody who this band is tuned to controlling what's going on in the room kind of secretly which is really helpful for parties or uh interesting installations in spaces where you want people to have a consistent interaction but you don't necessarily trust that the audience themselves can fire that interaction consistently uh this is a problem we see a lot in video games and in other kinds of public installations which is that when people are trying to interact with devices if we don't have the biggest simplest button in the world on that device they simply can't figure out how it works so a lot of the kinetics just don't quite have that uh nice response to the human interest where if you have a human controller who's doing something that doesn't seem obvious it can really add to the magic or the uh implicit I don't know how to talk about it the implicit performance of what's happening so people can see this robot but they can't tell what is quite firing it off and that sense that you don't quite know what's causing it makes it more interesting as you spend time with it so uh a little bit more about future work what would we do in future with this project so I think that the biggest thing is that we need to do a big write up for the repository for the edited scripts and code code and maybe make it possible to work on a b rate that is not 9600 which uh might or might not be programmed into the Mayo itself properly uh now that we have this device and the control system in place uh one of the big weaknesses is just it takes almost a full second for the code to run through and to fire the new pose detection so we could speed it up we might be able to hook this to a larger scale pneumatics project which would be interesting to do as an installation in a physical space that would be cool uh and the big thing we fou
nd out is that the reason the my armband isn't popular for this type of work is because it's not very good for this type of work so it's good for some kinds of work but not really for this because although I don't love buttons for doing this type of installation buttons are underrated because they are very consistent if you toggle a switch and that switch does something uh people have a lot of confidence that the switch is working where the armband takes away a lot of that confidence and it could add magic but it's just not smooth enough to really have the the same feeling of gesture control that people are now used to from things like capacitive touch screens or from actual click buttons so thank you very much for coming to our talk uh and we look forward to seeing you in the jitsy chat for it later on if you're interested in more things that we're working on you can find us in the chat and we will give you some contact information uh we also are both at the University of Maryland College Park so you can find us there if you are interested in talking to us more thank you very much have a great day so welcome back uh first of all a bit of a dumb question but why condoms seriously why condoms well thank you so much for asking uh it turns out that if you're prototyping pneumatics uh you can do a lot worse than a condom usually pneumatics are prototyped with uh latex balloons but latex balloons people people don't get angry when they break uh whereas people get very angry really quickly if a condom breaks so condoms are simply a lot tougher than latex balloons which is useful if you're going to inflate them and deflate them latex balloons also need to be stretched before they will inflate and they're not intended to deflate and then inflate and then deflate again whereas condoms are intended to remain stretchy for the entire duration of their life uh so they have good shot control they have a much tighter quality control uh they're about the right size to build a small
prototype with and uh they're very affordable if you have a university health center nearby which would like students to not get Knocked Up um so you can go get them pretty easily at like 8:30 p.m on the night is due what what what kind of challenges do you do you encounter when you use condoms to to build something like this honestly not many um so the trick is that if you have ones that are pre-lubricated they are slippery uh and that's annoying because they make things greasy and greasy things get dirty faster and so you want to make sure that you get the non-lubricated or less lubricated ones which are usually done with cornstarch and water instead of with oils uh which is better if you're going to be working with 3D printed parts or wood because the cornstarch and water will dry really beautifully and the condoms will still remain not sticky they but they won't be oily which you would prefer so why why did you decide to use a old device that relies on using abandonware uh to to build this thing isn't there anything else to to do the job apart from Mayo uh there are buttons and you the sculpture piece we could control very easily with any kind of in or out button device but the Mayo was the piece that made us interested in doing this work uh it's got some nice things about it uh one it's a Bluetooth device to as you saw in our video it grants a letter of performance when you're controlling the robot uh so you can do things like physical jokes you make a fist and the robot gets hard you relax your hand and the robot goes all to pieces and collapses in a really beautiful way um we also wanted to see if it was possible because the myo armband has this weird position in Prosthetics research where it is the standard for or getting uh EMG information out of uh people's musculature that is perhaps no longer in the state that one might prefer their musculature to be in um so we wanted to see what it would take to work with AO if we didn't have the backing of uh meta or
control Labs or one of those things what would it take to actually work with one of these armbands and we thought that this was a good project opportunity to figure out not only how to work with one but how to make one into a small desk top sized package so the actual robot is about this volume that I'm holding in my hands it's about five inches on a side or I guess that's 100 millimeters on a side maybe 120 um it's small it sits on a desk really nicely so if we were able to get a Mayo through a Raspberry Pi it means that we could build quite a number of systems and have a distributed collection of robots around a space which is good for interactions and for prop design which could then be controlled by one person prer in with with their gestures it it makes for an interesting experience that does indeed sound like a really cool concept um so the the Mayo um it still relies on on abandoned Windows software do you intend on maybe working working on that and uh reverse engineering further than the communication protocol we would like to uh it's a qu it's an open question whether it's worthwhile to do that or whe whether it would be more worthwhile to use a different EMG sensor that is already open uh and which does not require that situation so the two ways that could go would be to build a better encapsulation in a more Reliable Software environment for more open EMG sensors as versus reversing the myo itself which would probably invol involve buying a lot of mayo and then cracking them open and then sticking them on a logic analyzer um just because the myo does require that you use a Windows uh computer if you want to change the default configuration for what are called poses so the myo armband has six pre-programmed poses uh including fist fingers spread wave out wave in and double tap I think there's one more that I've forgotten um those are all predefined on board on the band so if I make a fist the Mayo can tell that I've done something that is like a fist the
reason that we would would want to change that is because the Maya was trained on some beefy dudes from Ontario and honestly if you're not a beefy dude from Ontario it has a lot of trouble telling what you're doing with your hands um so my colleague is not a beefy Ontario man uh and so we had some problems when we were presenting where like our nervous systems just run fast and so the Mayo will record gestures in a kind of unreliable way if we don't custom calibrate it so it would be advantageous to reverse it and that might be fun um but it's for this particular project we're mostly happy to identify that that would be a necessary step because that wasn't immediately clear when we opened it up one one last question so you you currently currently using just normal white condoms did you consider using colored ones we did we were W able to find ones that weren't lubricated like the lubricant grease is actually a real problem because it means that you can't quite snap things into place but I think for future work it would be fun to put together a larger peristaltic collection of these and maybe a a physically larger one as well as more of them uh and to make them very colorful indeed because people responded so well to the condoms we really weren't expecting that we were expecting something else I think I mean having having just like a a room full full of them in like different colors and different color combinations and responding to to one person does sound really cool I just think it would be a fun art project and sometimes what I want to do is a fun art project so in terms of fun art projects uh what's up next do you do you have any plans for for any other cool uh art projects Maybe involving condoms maybe involving condoms now that we've done this we haven't quite discussed it yet but there are a couple of paths we could go with it uh one of them is we have actually built this maette so we could pitch this to a couple of the art parties that happened international
ly and be like let us build giant human scale condoms with vacuum cleaners behind them uh and that might be an interesting idea uh I do like building comple complex static sculptures so we might do some more work with that and with the 3D printing a compilation of bamboo skewers sort of pieces of it um but at the moment I think what we are mostly focusing on is both of us our PhD students and we are almost done our coursework so I think we might shove things for like three months and then look into either reversing more abandoned wear or coming up with uh new 3D printed shapes to force the pneumatics through to do soft gripping and to do other robotics any are any any other Gadgets in your in your drawer that you're really itching to play around with but don't have the time to Yes uh so there's also something called The Leap Motion which was the precursor to valve's index system for their VR controls and uh we think that that might be very promising for more pneumatic control work I really want to shout out Tinker mind's programmable air system here um he only made a hundred of those he handbuilt every one of them but uh the the programmable air pneumatic system is just so easy to work with relatively speaking it required one line of code change to its main library to work and so it really opened up the possibility of making more of these gadgets and so hooking a leap motion to it might be interesting um because that does hand tracking so you can mount it to your forehead or you can mount it to your your chest piece and you could do some fun things with like LEDs or with further pneumatic stuff on costuming and so that's a thing that we're thinking of so you mount it to to your forehead or your chest and it has like a camera and films your hands and then tries to detect yeah it's an IR sensor yes it's like a little miniature lar sensor so the leap motion was a miniature lar sensor that can detect once it's been calibrated where your hand position is so it works like
a connect basically like a a camera yeah but you could mount it here or you could mount it here so early versions of the Oculus Quest system worked with the leap motions technology and they would just glue one of these things to the outside of a VR headset so that the hands could be detected and now of course um all that technology is built into those headsets like the cameras are built in so you can get um proc detection but the leap motion project still exists uh which means that just the hand detection without the VR system is is probably possible which I think is interesting I have some colleagues working with lar as well but I don't because lar is so useful I don't find it as interesting so for back to the Mayo um people are asking me where you've played around with the with the calibration process like what does it include how much work is it is it even worth it to try and calibrate something or have you not tried to do that we did two or three different custom calibrations which doesn't sound like a lot but I think that if somebody wants to hack on the Mayo I honestly think that it is not it should not stop you that it is supposed to be on a Windows machine or that you don't have Windows access um all of your work was done in Python on the Raspberry Pi directly from libraries that we found on the internet that aren't maintained but also do not have they have to be updated we have that those updates and we have to make the proper repository and publish it but uh if you are competent at reading python the updates are not a huge time block they're difficult but they're not possible I'm not even going to say they're difficult it's more like they're a pain in the ass pardon my language realistic yeah um and so if you get one of these devices I think and you are comfortable with python and the Raspberry Pi environment as long as you don't lose the Bluetooth dongle this thing wants to pair to which comes with it you probably are fine to just use the poses that are
on board because a more interesting thing is that the Mayo does shoot off raw data for your muscles at all times so if you were interested in doing custom calibration and you didn't want to use the thalmic lab software it would probably make sense to take the 20 minutes to two days to write the calibration on yourself using the raw inputs from the Mayo rather than worrying about using the Windows device simply because the calibration process for thalmic labs is so weird already and it's honestly just a timer that reads raw data while you're doing a pose so like you would hold your fist like this and the myo spits a bunch of raw data data back uh and then the software averages that out so that it's like on average you're probably doing a fist and this electrical signal means that you're doing a fist I don't think that it would be a problem to write that the problem is getting it to upload to the device and so that piece I don't know how to do but I also don't know if it's important because once you've read that information if you write the calibration it could live on your laptop and take the raw information read and then output the pose from inside your laptop instead of from inside the armband so it's convenient that the Mayo has six pre five or six pre-programmed poses if I were to do this on my own computer I would probably write the calibration and keep it local so that I could back it up to GitHub and know that it was going to work because that same raw data is very likely to be similar across EMG sensors so then the work wouldn't necessarily go to away if for example your irritable 14-year-old armband decided to stop charging one day yeah that's a lot of talking you just mentioned that it's important to keep the dongle if I lose the dongle does the device just refused to pair with any other Bluetooth dongle or or onboard Bluetooth technically it shouldn't but at the moment ours is finicky so the Bluetooth don't Le is preferred because we believe that it uses p
rivate codes we're not 100% on this but we're pretty sure it uses keyed codes toar to its own dongle and so it is a Bluetooth device it is using a Bluetooth communication pattern but as I'm sure everyone here knows Bluetooth isn't really a great standard because it's got like 17 to 20 different implementations so we have not hunted around to make sure that it could pair directly to the raspberry P but in theory it can just be paired to the Raspberry Pi we didn't focus on that this this year and so actually if we were to come back that's a more likely thing for us to work on than reversing the hardware itself you mentioned that oh sorry for interrupting you I interrupted you um you mentioned that the Deo uh is also used in biotech can you can you elaborate on that a bit have you have you done research into that no we just know this because you are from a neuroscience background at renell and you worked with this in their many terrifying Labs um and so we'd heard of it being used in labs and there's footage out there of it being used in Labs but neither of us have worked in Labs where this arm band is is being used we just know that it is because there's recent footage of it happening and we recognized the device so I believe that that is a good reason I mean I mean it does just spit out raw muscle data like you can just read it once it's paired as long as you don't lose the dongle it does just spit out raw muscle data so and the muscle data is I mean in as much as anything is accurate TM it is accurate so my best answer is no sorry um are there any op Source or open Hardware Alternatives that you're aware of or yeah there are a couple of them out of Ukraine and they're available on um one of A's uh purchasing sub forums there are a couple of ones that are available and they are really interesting um of course Adafruit sells sticky pads that you can attach to uh as EMG sensors that feed through an open board uh the reason we went for this one was because it's such a n
ice package we wanted to double check that the hardware wasn't going to waste but I think that there are other ways you could do this and you could get the a similar magic effect I don't like using EMG pads they're itchy yeah yeah anywhere they Stucky it's fine um I don't yeah but yes you could find open answers I I'd imagine that and EMG P also like leaves the both a problem of cost and like viability cuz you you can't wear the pad for too long I'd imagine and reusing them seems difficult or questionable at best yeah so it's really um the the thing that makes the myofunctional and there is an open project it costs about 50 bucks uh I I saw it on hackaday the other day because we it was proposed that we made would maybe use that the other project um we wanted to use this because we wanted to see what the blockers were really it was a curiosity but the big thing the Mayo offers is that it has like I would have to look at a picture of it but every one of those blocks on the Mayo is itself an EMG sensor so it just has has a lot of sensors and they appear to have done the work to make sure that each of those sensors is reading data in a clear way that outputs a number that is actually reflective of what the muscle is doing where the sensor packages we see online that are the open sensor packages cost less money because they only have one EMG sensor in them and it's held on place by an elastic band so the risk there is that if it's placed differently on your arm and it's not glued to your muscle which is how the ones that you buy that are peel and stick stickers that have the wires on them um the risk is that it'll move around on your arm and it will take different readings for each position on your arm that it is where the Mayo appears to a of all it holds itself in place like stink it's really really really really good at that um and like it doesn't shift or move around but be of all all of those sensors appear to work in concert to get you the muscle data out and so t
hat piece I think would be expensive to reproduce I don't believe I could do it for less money than the cost of the actual armband as it was produced so it kind of fuses the data to and like to get rid of that that rotational variability yeah I think it just gives you really good clean output data and if the labs that I've seen it attached to are like the media Labs weird fancy new legs lab and so I feel like if people who have the funding are using it that it might be the thing to use but that's just me being sneaky like I don't know if that's a real thing maybe I'm wrong but I think that that's I think that's probably the actual advantage of using this band is that they did the they did the noise leveling yeah um I'm just going to check really quickly whether we have any other questions going on um uh um I can't see any right now but uh thank you very much for your talk uh it was uh amazing it was enjoyable something something really cool uh thank you for joining us today so on this channel we will continue with bringing the hack back into the chaos by Daniel maslovski at 11:00 p.m. um you can as usual send us requests questions or memes on Mom by using the hash ch23 you can send us questions in the IRC channel on ic. hack.org with the name sh23 um we have a YouTube and a Twitch account uh at fire shs each which also each have a chat that you can post your questions into we will read and uh respond to them um we have a social viewing experience in the aquarium in the jitsy uh the link to that can be found in our info beer shortly um our mainstreaming platform is streaming. media.ccc.de uh we also have a real life setup there so you can rewatch past talks that you've missed as this one um for everyone at 37c fre we're a bit hard to find us the info desk if you need help
