Iroh the Soft Skin Robot Sensor
HTML-код
- Опубликовано: 12 сен 2024
- This is a silicone magnetic skin sensor I invented that uses a flexible magnetic source and a hall effect sensor to measure touch pressure.
Please join my Patreon:
/ adrianperez720
Reference:
Soft force sensor made of magnetic powder blended with silicone rubber
www.sciencedir...
Please join my Patreon to support more videos like this: www.patreon.com/adrianperez720
Just as a quick pointer. I think this could work really well if you instead of going for the Hall Sensor approach, you tried a capacitive sensor. Basically, your silicone + iron shaving would work as a capacitor plate and pressing them gets the plates closer, thus increasing capacitance. This would require a way of measuring capacitance though....
Another solution is via resistance. Since the silicone is compressible, you can add metal shavings to it and pressing the silicone actually decreases the electrical resistance across it.
Your solution is actually pretty elegant I must say
@@yertzar775 I would love to try those other approaches at some point. Thank you. :)
I really like the idea of a conformal/molded pressure sensitive sensor membrane. Hall effect sensors are cheap enough, I could imagine putting them all over the place. neat idea!
@@natosaichek Thanks! :)
@@adrianperez3375⚠️ God has said in the Quran:
🔵 { O mankind, worship your Lord, who created you and those before you, that you may become righteous - ( 2:21 )
🔴 [He] who made for you the earth a bed [spread out] and the sky a ceiling and sent down from the sky, rain and brought forth thereby fruits as provision for you. So do not attribute to Allah equals while you know [that there is nothing similar to Him]. ( 2:22 )
🔵 And if you are in doubt about what We have sent down upon Our Servant [Muhammad], then produce a surah the like thereof and call upon your witnesses other than Allah, if you should be truthful. ( 2:23 )
🔴 But if you do not - and you will never be able to - then fear the Fire, whose fuel is men and stones, prepared for the disbelievers.( 2:24 )
🔵 And give good tidings to those who believe and do righteous deeds that they will have gardens [in Paradise] beneath which rivers flow. Whenever they are provided with a provision of fruit therefrom, they will say, "This is what we were provided with before." And it is given to them in likeness. And they will have therein purified spouses, and they will abide therein eternally. ( 2:25 )
⚠️ Quran
Wow that's actually really cool, i never thought of using a Hall Effect Sensor as a pressure sensor
Thanks. It makes for a really compact arrangement.
@@adrianperez3375I was thinking the exact same thing instantly. Absolutely brilliant. The some of the best things are just so simple and go under the radar.
@@MrChinkman37 I love simple solutions.
i love the way you smile at the camera when you get something working, really makes me wanna get back in the workshop
@@janthran Awww. Thanks. I really appreciate that. It's workshop time!
I don't think you need permanent magnets. Here's the idea: take the power wire for the hall sensor, and wrap it in a coil around the sensor coplanar with the front face. That sets up a magnetic field normal to the sensor, and you'll read a constant North (or South). Then when you press the iron sheet closer to the sensor, the magnetic field will couple better to it, and the field will get stronger. It's a little like a guitar pickup. I have no idea how many turns of the coil you might need, though, for it to be measurable.
That's a really cool idea! I love solutions that require even less components.
Please try that idea and record a cool vid like this one @@adrianperez3375
this is basically what i was considering doing thinking abt this topic a while ago ! i really think that would work
@@morgans.5190 I would love to see it get tried. It's a neat idea.
Chokebot safety had finally been achieved.
Finally we no longer have to deal with robots choking not hard enough due to safety concerns
I feel like this will eventually end up in a fleshlite lol
😂
gooner-pilled
A fleshlite that moans when you put it in. Yikes.
@@adrianperez3375 may i preorder?
imagine a array of magnetic sensors and the dots of iron. The skin can feel the streaching sense and pressing sense.
@@Ross_an_Artisan The future!
Fantastic job! I enjoyed seeing all the failures leading to success
@@Antichamberteam80110H Thank you. I really appreciate that. I like showing the failures to let people know that any adventure is going to have some along the way.
That's an impressive project ! I just found your channel for the first time, and I must say you're very likeable you seem nice like a few people do !
@@spacechips9626 Aww. I really appreciate that. Thank you.
Awesome stuff, now you just need to line them in a string to drape over a hand. Keep up the awesome work!
Thanks. It will be so cool once it's on a hand.
I know you are dealing with the silicone, but please I recommend using gloves that you can slip your fingers out of if you are using neodymium magnets. An example of some of the ppe used is in Brainiac75's video, "Monster magnet meets monster magnet..." It shows the gear used. Also to help avoid pinched fingers I recommend using a 3d printed device to slowly align the magnets together where you have one magnet attached to a fixed bottom where you then seat the silicon that of which is part of a insert stand. That insert stand then allows you to insert a magnet with a cap on top that can take a rod which you can use to slowly move the magnets together whcih is aligned to the insert stand.
@@pup4301 That is a really good idea. Next time I print a setup to slip them into place.
I've been working on making a prosthetic sensor bracelet (forearm) out of custom pressure sensors as well. This is great work, thanks for sharing!!
@@turfptax I just subscribed to you. Thanks, and good luck with your work as well. It's nice seeing there is a community of us pursuing these things.
@@adrianperez3375 Yes, it gives me hope for humanity! I also just subscribed to you! What you are doing is awesome!
0:04
2700 years later
the robot: "what is love?"
@@rkeewi hahahaha!
baby, don't hurt me. don't hurt me, no more.
@@woundedmonk1884 🖤
man this is so cool!!! and not out of reach, love it
@@TheChillieboo Thank you. That was my goal. Make a soft sensor that was in reach.
losing the touch feel was something pendient to fix for my next case, thx nee
I use the soft foam for integrated circuit packaging as pressure sensor material, its resistance lowers when it's compressed. It's quite sensitive, easy to use and very cheap. Just put it in between two sheets of copper foil.
@@Benadski Maybe I try that next!
This is perfect bro you should contact all the startups and big companies working on robotics so they can all use your idea and get in a race to give our soon to be robots amazing skin we could also use it to cool the PC with veins making it warm also ❤❤❤❤❤❤❤
Thanks. Your comment really makes me wonder what each company is doing for sensing. Force feedback? Resistance? Hall effect sensors? They can have my ideas for free since I'm putting them into the public domain through this channel anyway.
this is going the future humanoid/animal robot that will feel touch sense🤔
@@ferrio6773 Yeah. I want a robot hand to be able to pick up an egg without damaging it.
Rather than just reading the polarity from the hall effect sensor, wouldn't it work much better to read the raw values. Use the change in magnetic field as data output. Rather than having a -1, 0, or 1 output, you can have all the numbers in-between, read how much it's pushing / pulling.
@@ToastyStoemp I did both. The words get larger as the values get larger.
5:53 noOoOoOo, noshin, noshin. Theres deffinitely ai at play lol
The funny thing is that is in the original audio! My dejected pronunciation and facing away from the speaker made it sound bad. I almost cut it. The ai may have amplified it as it attempted to correct the levels.
This is awesome! So many cool applications!! Also if you ever want your silicone to feel and react more like skin you should look into using some deadener for plat silicone. Plat silicone is tricky though and can be a pain getting the levels right with multiple additives (iron filings and deadener) to where it will still cure, might want to try small test batches to get your ratios right. Just a thought! Great video btw! Can’t wait to see where you go with this, will def be following your channel 😊
@@emmamarx9284 Thanks for the suggestion on the deadener. Good idea.
Dude this is seriously awesome!
@@evanbarnes9984 Thanks. That makes my day.
Very neat!
@@legotechnic27 Thanks. Glad you enjoyed it.
All engineers who just done something really awesome have a moment of childish happiness 😅
@@alekseyvilchinskiy2913 Thank you. It's true! Such a great feeling.
Congrats on the progress!
So weird but I too was randomly asking chat gpt about embedding magnets in silicon and I got it to suggest hard ferrites and magnetites as alternatives to iron oxide.
My application doesn't need permanent magnetism but magnetite seems like it might be what you hoped iron oxide would be, if you want to look into it!
@@GseGodsSpeed Thanks. And thank you for the recommendation. I remember pulling magnetite out of the sand at the beach with a magnet. Good times. I will look into it.
subbed. I have been thinking of something like this but wanted to use in something else. Thanks
Thanks for the sub!
Man this was a long video to come up with just a totally different solution.
@@undersky596 I wish I could have conveyed the process better. I had two videos. The one that I was going to publish was the one where I failed with the iron. But then in the last hour before publishing I figured out the magnetic tape solution. And added that in at the last minute. Let me know how I could have merged the two better.
Why didn't you use the thin magnetic sheet material found on the back of flexible advertising refrigerator magnets.
@@jvebarnes At the end of the video I used magnetic tape, which is what is on the back of fridge magnets.
@@adrianperez3375 I saw, although I have a few such magnets and they appear to me to be a lot thinner than the tape you used. However speaking of magnetic tape, as someone who worked with computers when they had huge reels of tape similar to that used in music cassettes, what about 8 track tape or vhs video cassette tape it would be thinner and more flexible.
3:02 you talking without takling
@@Krazy0 I went to the timestamp and couldn't figure it out. Could you describe more?
This is dope. You could give people their feelings back using these sensors plus neuralink. Could put these on prosthetics. Arrange the magnetic film in a flexible pattern. Like for fingers, lay them down the fingers and then make a strip on top and align the strips longways and in a network. Make a cut out for the sensor to sit in on the prosthetic so it doesnt get damaged and then layer the skin like a fat layer and then the mag film or tape and then another layer to protect the strips. I like this. The stronger neo powder could be good to test and might be a better option even though it is expensive. Maybe the flexible tape is better for price and easier to make. Maybe wires could be more flexible, like the small strands of metal like strings. Thinking about it, the mag wires could allow different areas but still close to be picked up, like one sensor picking up multiple signals and with multiple sensors you could measure the readings like earth quakes and have the sensor share for multiple areas while picking up one group. Like two sensors with strands in between could be measured together and give multiple readouts for the whole area or something. Idk. This is sick though.
I'm picking up your vibe. This is what future prototypes will probably look like.
well this is a suggestion, you could mold the silicone with magnetic tape inside it rather than making it as separate sheets
Next time I make a skin, I'm totally going to do that. It will be nice to have one seamless structure.
@@adrianperez3375 Good luck ! Im subscribed to see how it would turn out :) hmu if you need any help with 3d modeling or cad stuff
i admire your perseverance & attitude to achieve your goal. What code do you employ to show "neutral" , "South" , "North" + screen color changes. ? PS (personal suggestion) loose the musici- it is too loud and quite distracting
Thanks. I used Arduino code to process the signals coming in from the sensor. And Processing to create the visual feedback. All the code was made using ChatGPT and some tweaking by me.
Hmmm I wonder what kinds of robots you could make with soft, sensitive synthetic skin... 🤔
Any ideas?
Nice theory, it working like in every magnet where the fields of all atoms line up and add up to a positive sum.
I believe the reason for it not working is that every single filing becomes a magnet by its own. This being a macroscopic scale, those small fields don’t add up they cancel out. Especially in such a suspended state in the silicone with comparably a lot of space between them. Combining magnets to increase their strength in one direction works, but only in exact patterns and when touching directly.
That’s what my intuition at least says 😄
But good solution in the end!
But if I’m wrong, or just the pattern has to be right. This would open up biological magnets (also flexible). 🤔
@@nextstopptop3963 Thanks! I like tackling projects where I have missing knowledge. That way I learn a lot.
@@nextstopptop3963 Biological magnets would be really cool.
With those magnetisers a blip is enough.
@@lachlanlau Good to know. I had never used one before.
What about to use 3 or 4 hall sensors and triangulate the pression point to know exactly where the point of pression is and I think with some neural network training, multitouch will be able also. Nice idea dude!
@@ivanfreedom Thanks. That's a really good idea. I was thinking even just two initially might be able to identify if something is slipping out of the fingers. Or measure friction.
@@adrianperez3375 good morning, I think you can also create a specific magnetic pattern in the magnetization process, and use two magnetometers like the qmc5883l giving also some 3d sensing resolution on the 2d plane of the finger tip.
this is awesome!!!
@@Alt.ZER0 Thank you. That makes my day.
have you looked into Flexible Piezoelectric Film Force Sensor PVDF ?
@@mcrotbot Yes. And I might try something like it in the future.
So would it work with nickel powder then I wonder... I would need to look it up but that sounds like a good start.
@@yzgrdyn-WiseGuardian- I think a nickel iron alloy.
Combine with SMD and flexible boards and you can make something...
@@youkofoxy great idea!
the next step is hall sensor matrix 0.0
I can't wait to get to that point! :)
Something feels a little weird about the sound on this one. Dope project Adrian!
@@mikegrace8814 Thanks. I may have accidentally left an audio clip on mono? Was that it?
Ok I'm not the only one. It almost sounds like ai, when he said part a and part b, it sounded like he said part gay and part b
@@gavitronv7246 I may have to rethink my audio setup.
Hey just putting in my 2 cents, it sounds a little warbled for me and sped up to the point where it's a little uncomfortable for me to listen to. Can I ask what processing you're doing with your sound and video?
@@guwuse Nothing with the video, but with the audio I use Adobe's new upscaling for podcasts.
Since the number of sensors will be discrete, why not just embed small neodynium magnets in the silicone at each sensor location?
@@rydude998 That could work. Suppose I hesitated to do that because I thought it would be too easy to feel as a hard spot through the skin. And neodymium is relatively expensive to magnetic tape or iron.
What about a capacitive solution? Like using foils between the layers, if u also add a resistive layer, you could also locate the pressure and that on a big surface.
There are resistive pressure sensors out there. They seemed only mildly flexible. But I bet it could be done.
I think it could be done, though you'd be getting into the weeds with details like dielectric constants and variable silicone weights/densities/volumes, etc.
This is very interesting!
@@_BL4CKB1RD_ Thank you. I really appreciate that.
there is an issue with the audio
Yes. It's something to do with the upsampling I do on my voice. My other videos also have it, but without these weird artifacts.
@@adrianperez3375 nice, it was kinda annoyish but the topic of the video was so interesting that i stuck to the end, hope to see an update in the future!
@@Gureenu Thanks for making it through. There will definitely be more robotics projects in the future. :)
This video is insanely underrated
Also are you using AI noise-cancellation? Your voice sounds very off
@@Kav_Games Thanks. I'm using AI upsampling on my real voice. I fixed it in the next video.
you could have used Magnetic Sheet Papers aka Fridge Magnets before printing paper thing which are magnatized
Fridge magnets could be a good source. Good idea.
Flexible fridge magnets are usually magnetised in alternating, rotating stripes to form halbach arrays. Halbach arrays have way more magnific field on one side than on there other, which can be seen if your magnet won't stick well to metal on the opposite face. You might be able to "erase" the existing magnetisation pattern and re-magnetise it to have a single North/South pattern through the thickness of the sheet, if that is what you want, or along The width or breadth of the sheet if that would work better.
@@MrRlnansel So that's why the fridge magnets don't stick on one side. They are Halbachs! Thanks for telling me. That's actually a good thing. The sensor skin won't stick to objects the robot holds.
@@adrianperez3375 That's true, but unless you have several Hall effect sensors spaced such that one of them is always aligned with a N or S stripe, you wouldn't be guaranteed the sensor would "see" any flux changes. The Seattle Robotics Society built a Pacific Science Center exhibit in the late '80s or early '90s with a couple of robots that homed on IR beacons toward the middle of the exhibit and with two stripes of flexible magnetic tape at either end of the display. The stripes were in grooves cut in the floor of the display. Each robot had a Hall effect sensor that was meant to let a robots know when it eas in the "goal zone". It didn't work as planned because the N & S stripes across the width of the tapes were essentially encountered randomly by reach robot.
cool vid :) also are you using ai for your voice though. or is that your natural voice?
Thanks. This is actually my natural voice. I do use AI to upscale the audio quality. Are you hearing artifacts?
@@adrianperez3375 yeah there is some artifacting that I noticed, watch the video through with your sound up and you might hear it. Not a big deal tho :) as I can recall there’s a part where you laugh and it just sounds very digital/robotic.
@@gabriel3437gfcxg I'll rewatch with sound up like you suggest. Hopefully the next video is artifact free. Or I'll just have to stop laughing. 😁
@@adrianperez3375 hahaha, never stop laughing man. prob just need to tweak a few settings
Those iron filings are hilarious, looks like it belongs in a kitchen
They do. I had to be careful not to put them in the spice cabinet 😂
do not show that to people with ai robot girlfriends
@@edwardlariviere9710 I won't. :)
Fix ur mic or audio in general like it’s not that expensive
@@emm4148 patreon.com/AdrianPerez720
@@adrianperez3375 no
Is it just me, or does he actually sound AI-generated?
@@Bednar121 That is my voice, but it is upscaled with AI. I used the upscaled version on my others videos just fine, but this time it produced a lot of artifacts. I wish I had caught the error.
Why, like i understand it would be incredible, but still, this has a bit of an… Aftertaste yk
I don't recommend eating silicone or magnetic tape. And definitely not a hall effect sensor. 🤣
@@adrianperez3375 not an actual aftertaste
@@thereal_wertzui Is there anything you noticed that I could have improved?
@@adrianperez3375 no, what i meant that many people seem to think of s*x robots when you put silicone on a robot
@@thereal_wertzui Haha. Sorry I didn't get it. I was temporarily dense.
Isn't "No, not love, just pressure" the asian parent mindset?
@@aksdoaskd lol
Please just google things and research shit chat gpt will always lead you astray
@@bigzed1331 Interestingly ChatGPT also lead me in the right direction after it led me astray. Google searches just kept leading to the classic kids experiments with iron filings and saying iron is magnetic without mentioning whether they are permanently magnetic. But ChatGPT later told me about the permanent magnetic properties. I should have put that in the video, but I ran out of time.
@@adrianperez3375 just search on Wikipedia or Google scholar. ChatGPT gets a lot of science stuff right but it is not reliable
🗣velostat
Good suggestion. I looked at velostat sensors. They are very cool. I just wanted to pursue the powder in silicone approach first.
Great video but the weird AI-esque voice is really distracting, you can probably get a cheap microphone if it's a quality issue. keep going otherwise.
@@DemsW thanks. It is my real voice, but unfortunately the AI upsampling I used really messed it up. It's fixed in the next video.
Cheeks when?
May I ask why you AI your voice? Why don't you use your real voice? It adds a strange quality to the videos, and I find it disorienting to watch with audio on.
@@D3moknight That is my real voice, and I upsample it to make up for my crappy microphone.
@@adrianperez3375 Okay, sorry it just sounded weird and I could hear what sounds like some AI artifacts. I went back to some of your older videos without the effect and it's very different.
@@D3moknight No apology needed. I appreciate you telling me something is funny with the audio. I do use an AI upsampling tool. I've used it on most of my vids. There is a chance it's creating small artifacts. And I've noticed some background music brings it out more than other music. So please keep up the advice. I need it to make the videos better.
@@adrianperez3375 to remedy your microphone issue, I recommend the “Samson Q2U” I believe it’s like $70? And it’s extremely impressive for the price. Studio grade for sure. It is a dynamic microphone as opposed to the standard condenser microphone. The dynamic ones are designed to cut out any background sound, and condensers(in the name) condense all nearby sound into a coherent sound. But often times that’s not what you want.
Producers have sort of flooded the market with these condenser options, without explaining the best use case for them.
Dynamic = good for vocals, instruments, commentary, etc.
Condenser = good for recording a meeting, perfect for phone cameras to capture localized audio.
I personally think it sounds fine, but that may be easier than a software or pricey mic!
@@adrianperez3375I mostly noticed it in the first scene of you sitting and showing the iron filings. To me it's the persistent attenuation-y kinda flutter that I pick up on the most. It reminds me of the discord noise canceling, but with a slightly different feel to it. Not super bad but I did notice it. (Would I have noticed it without seeing this comment before watching? Who knows, I'm guessing probably not.)
the AI voice is so annoying
@@pankordix5309 That is actually my voice. It is upscaled by AI, but it's me. There must be something out of my hearing range that makes it annoying. What's weird is that I have used this for most of my other videos and no one has complained. Maybe the algorithm changed.
Curios what the factors are on scale for this and how close the hall effect sensors can be. Perhaps it is someting based on thickness of the skin thinner skin will lean to closer sensors as when depresssed the skin averages less giving more percise reads on where pressure is
Exactly. I had to slice the silicone thickness in half to get it that bit closer.
imagine thinking you can turn iron filings into magnets. wow embarrassing.
@@MostConscious I was temporarily embarrassed. Luckily I found the solution at the end of the video.