GAME OVER!? - A.I. Designs New ELECTRIC Motor
HTML-код
- Опубликовано: 28 апр 2024
- PicoGK forms the basis for all of the company’s computational engineering models. It has lead to the development of a completely new electric motor that is entirely 3D printed through AM. But is this the future of motor design?
Sources & Credits:
• Electric Motors, How E...
• Why is electric motor ...
• Superconducting Motor
• What is an INDUCTION M...
• Magnet Free Motor By M...
• Advantages of 3D print...
• Winding and Assembly o...
• GKN Powder Metallurgy:...
• SCHAEFFLER AEROSINT'S ... 
Наука
I’m looking forward to when Ai creates RUclips titles that are more accurate and less hyperbole.
XD
I'd be intrigued if AI actually figures out that it is a more powerful way to build an actual viewership base but in the meantime, hilarious comment! So true!
And a feature that allows you easily block channels as soon as their bullshit clogs your feed.
How is a question mark hyperbole?
We’re trying, but it just keeps spitting out
“All Your Base Are Belong To Us!”.
No idea what that’s supposed to mean. 🤷♂️
@@dasstigma you’re right. A question mark implies the question rather than insinuate an extreme fact. I think though that the structure of the whole title is designed to suggest that the motor is a game changer and something potentially very new. Click bait maybe? Could have said “New motor designed by Ai has potential to change industry…” still a big statement but considering many of the images look like they are from Midjourney I think the title and the content are all a bit over the top. But your initial point is correct and a question mark on its own does not imply hyperbole.
Title "Game over" then 10 seconds in "But will this truly be the motor of the future." Suffering from a bit of premature e-declaration.
But clicks.
Homie is discovering clickbait lmfao
KEKW premature e-declaration!
It is wild to see a human behind this one, or maybe not. This has the look and feel of a ai generated video. Maybe it has just advanced that far. And we are proof it works.
Now you are asking the right question.
Clickbait titles is what the down vote button is for.
there isn't a working benchmark, nor a breakdown of parts...this is just a model it created we don't even know it works.
As is it just a standard turbine from the looks of it. It's a step, and a few layers off
yup... clickbait title
Looks cool though
The vast majority of "AI" stuff is outrageously fake
You look at that, and that's your takeaway? Think about complex engineering building on a molecule / nano level and efficiency levels. Everything is about to change in such a fast way.
I think there should be a question mark after "GAME OVER" until a motor is actually produced and evaluated.
That is true so I changed the title.
@@Tech_Planet
Thank you.
@@Tech_Planet But the ! comes before the ?. At the current state I believe it should be the other way around.
The "GAME OVER!?" part in title is totally misleading and unnecessary until this tech comes to real life use.
@@IvanZivko oh no the game is very much over. i just lost the game
Was kinda hoping to see more performance specs for the motor, as well as what exactly the AI was optimizing. The high surface area appears to optimize heat dissipation
Might be able to run coolant through the coils 🤷
Agreed !
Heat emitted by a motor is an indicator of inefficient operation. It's energy being wasted as heat instead of mechanical force. If you have to seriously cool a motor, the real problem is that is wasting energy as heat. Gas and diesel engines are great examples. A whole elaborate cooling system is required just to deal with waste heat. The energy spent making that heat is energy not spent doing whatever the engine is supposed to do.
@@patrickmccullough989If it's really powerful motor it will generate much heat even if it's highly efficient..
@@onearthonelegion It's like you didn't even read the comment.
Heat produced by a motor is wasted energy, full stop.
future machine parts are gonna look like video game artifacts, which is really cool
What about rocket 🚀 parts ? ruclips.net/video/Cms_v_OUXco/видео.html
And very quickly the current parts we live with, as well as everything we consider as just normal modern-day conveniences will quickly change in aesthetic, efficiency and function. Everything we are so used to now a days will become the new retro, treated like the equivalent of what steampunk is now a days. A novelty or fantasy setting of sorts. Where a normal V8 engine and normal shaped headlights are a legitimate equivalent of a steam engine with funky copper tubes sticking out of them in fantasy. What a time It will be when a smartwatch is a relic and televisions are only in memories because we all wear solar powered ai driven spatial mapping augmented reality smart glasses.
Point being, this year we are currently living in. with the technology we currently have and are so accustomed to will seem more and more technologically archaic even more rapidly than we can possibly imagine. at the rate we're going the world will not only look different aesthetically but it will also function differently under the hood of society norms and that's both awe inspiring and terrifying all at once, I already don't know how to write C+++ nearly as well as I sue to, What in the heck am I going to do when they add 7 more plusses in just as many years 😂
@@taythree5549 Wait til these future kids see my HP Pavilion Monster laptop. They are not ready for the swag. Sadly, it won't turn on anymore...
@@taythree5549 Meanwhile, 650+ million people are living under extreme poverty and we are on the verge of WWIII. Guess the world fuel is really just dreams, til the nap is over...
In high tech they already are for a long time (decades), its not issue of engineers imagination or skill its an issue of manufacturing. In aerospace where its profitable to use expensive manufacturing methods you have crazy complicated and state of the art designs. In cheaper industries parts have to be made out of simple blocks that can be mass made with easier and more economical methods ( molding, stamping, etc. ), those impose a lot of limits on shapes and materials.
GAME OVER - A.I. Designs new HUMANS with 40 mangled fingers
This type of human will be a reality in a near future!!
@@danieljohnson2662 🤣
holy cripes I'm laughing so hard at this!
Frankenstein
I believe the copper windings will not work very well if 3D printed. Any microscopic "hole" will lessen the magnetic field or cause stray eddy current causing problems with power. A regular copper wire is going to be hard to beat.
That is very true!
True, but that can be solved with some sanding and polishing post-print. I'm more bothered by those sharp angles in some parts of the design, fields should be treated as fluid and fluid doesn't like sharp angles.
@@JenkoRun internal holes can't be fixed by sanding though. 3D prints have holes inside them all the time, a part of why it's never a good idea to use 3D prints for things in contact with food. Germs will find their way inside and make it super unsanitary
@@kapser2210 The holes run that deep? Well then, printing the copper is definitely a bad idea.
@@JenkoRun Nnnnope. It's not about the internal shape, but the volume of the printed "wire". What you're describing would be like trying to remove air holes from bread by cutting off the crust.
For those commenting, generative design isn’t a new concept in engineering. It’s cool that it’s being applied to more complex problems!
3d printing is approaching the price of machining with way more to go.
Its still insane for prototyping. I have no doubt we will have whole facilities dedicated to huge printing machines for part production.
It is the future of all manufacturing, we are just at the first steps in a long way to go.
3d printing is by far the most expensive manufacturing process on earth and will be for basically decades to come if not forever. Machining is far simpler and quicker for majority of applications and simplicity always wins for cost and speed.
@@dodgygoose3054it is not the future of all manufacturing, only select fields. The tech is so much more complex, expensive and slow compared to conventional machining that it will remain a comparatively small manufacturing method until major breakthroughs are made that can make machines cheaper and faster (I'm talking industrial machines doing SLS metal and PBF polymer) I'd love 3d printing to be the be all end all but realistically it won't.
@@francoismurrell4604 I got a PCB way account that says otherwise.
What do you mean by game over? What game? What does that have to do with the motor?
Game over for humanity it’s all downhill from here
Copper wires, probably.
Game over for human designs. AI will do it better.
I want to see a performance test of this so called AI motor design.
There was a evolutionary engineering researcher some time ago who used software to "breed" a beam for the International Space Station as a design exercise. This is where the software applies what amounts to genetic breeding of parts to come up with a design that fits the requirements. The beam the computer created looked like a giant leg bone. These generative and evolutionary design methods are going to create some great stuff - that looks really, really weird.
Maybe we’ll finally get a decent turbo encabulator
My retro encabulator is still doing fine and I don't see a need to upgrade... although the dingle arm is showing some wear.
@@irridiastarfire Check the signal fluid level.
it is more about 3D printing than a.i.
These days all kinds of optimisation algorithms and even brute force search are called an "AI".
Now I know why H.R. Giger's designs for Alien movie looked like this...
And Gaudi before. There's a good reason why such designs are natural.
WHAT DEVELOPMENTS???? it hasn't even been fucking TESTED¿¡¿
3D printing different metals in volume is going to be challenging.
AI will figure it out
never be economical at scale.
Did AI write this video also? After the high-school physics lesson it's just disconnected rambling lol.
Permanent rare earth magnets are a high cost but Niron (Nitrogen-Iron) permanent magnets could reduce the cost below that of induction motors. These new magnets are stronger, cheaper, and have a higher heat tolerance than Neodymium rare earth magnets while also being environmentally friendly.
I really hope at least some of these developments really turn out to be game changers.
It looks so alien.
I think the same founder of hyperganic(the weird aerospike motor) formed Leap71 (PicoGK).
...like an H.R. Giger project
Im still wondering who engineered those shades.
It is
And require alien physics to work well to :)
It's not that superconductors can withstand higher temperatures. It's that they have less resistance which means less heat would be generated.
This is mostly clickbait. The axial flux motor used by Koenigsegg is based upon a license from a British company that has now been bought by Mercedes.
I am converting a 1970 Lotus Europa into an EV using a similar but much smaller axial from motor. The electric motor weighs just 22 lbs and replaces a Renault engine and accessories that weigh 246 lbs! Oh, and it produces around 40% more torque, with maximum torque at just above zero rpm, as well as over 60% higher bhp. It would be easy to install a larger motor, but it's driving through the Renault transmission to keep the suspension the same, and I'm concerned too much torque will snap the input shaft like a carrot.
BTW, a popular conventional radial flux motor with similar performance specifications weighs 120lbs!
My current (pun intended) calculations indicate the vehicle will weigh no more than it did when it left the factory in Britain.
You are correct about cost/weight equation, the permanent magnet gives you low torque end yet it trades efficiency at high end. The SynRM is a compromise between the two but rare earth metals are still involved. I was trying to lead to 3d flux induction motor without expensive magnets but it's more of a probability of reducing cost, the question is whether it can solve the riddle of low starting torque. I think we should be open to alternative manufacturing techniques which can lead to that.
Amazingly I was only thinking about 3d printed copper coils for motors just last night. However the big problem was how to insulate it. the insulation must be super thin and withstand high temperatures with minimal plasticity (flowing under heat and or pressure). Enamel is the coating of choice but I can see that mixing copper powders and enamel powders would not work as the same layer of a print must contain both and the enamel must be very thin.
True for small motors, but high power motors use large solid hairpins and very few turns so insulation thickness becomes less significant.
Also the coils can support themselves so the space between them can be potted with channels for coolant.
Would be difficult to 3d print a small FPV motor using this technique though.
Excited to see the output of this new design!
Also keep in mind that one thing LLMs aren't great at without extra help is mathematical computation. Kind of can't do mechanical engineering without math.
A motor that has to be cryo cooled! So when it chucks out all that power and warms up it then looses superconductivity and probably blows up! Back to school AI.
I love the whole going through how motors are designed, but then not really ending with how (if it does) the ML generated motor works.. if it even does..
What is the point? Existing designs have close to 100% efficiency and very high torque anyway.
When comparing two things, one can not be "exponentially" greater than the other. If you don't know what words mean, don't use them. An exponential increase is a accelerating increase of one variable in relation to another one, like a fixed proportional increase in cost each year. It does NOT mean "a lot, lot, lot bigger than".
I'm not an engineer, I watched this because I thought I might learn something. I did, and actually understood the concept of the problems this brings. But also the excitement that can come from this new era of Ai and 3D printing.
This is what I want AI to be used for. Helping us create new and better technology.
You have first to invent a real AI.
AI will invent god-tier motors and batteries
... because it needs that to be embodied while clearing the planet of humanity.
Quite impressive for sure, the new design will definitely improve efficiency.
"Exponentially more power density than any motor out there"
Look up that first word :-D
It even makes it's own gravy!
So yes, nano scale exotic designs will be if not already valid. Especially if they are coalesced or fuse together to be strong, then solid parts. This solves so many robot issues. If we think about it and with AGI already out but not out to the public, we'll then it will be upward at an incredibly fast pace. This is a powerful game changer for upward mobility. Why is all this happening? Thunk about it.
Space!
For interplanetary infrastructure and industry, producing motors is the first complex component that would be useful to produce in space, after oxygen, concrete, sheet metal, and glass. Motors and sheet metal are the bulk weight of any industrial machine. A single complex 3d printer is cheaper to ship to the moon than a whole motor factory, even if the motors produced are slightly less efficient.
Once we can get AI into a physics based simulation and let it run wild. There will be some crazy stuff coming out.
The AI isn't deciding the best performance design, it's deciding the best structural design for the least amount of material needed to maintain a target strength, but also doesn't really take into account complex factors like temperature differences, wear and tear, material stress fracturing, etc. So no, all this AI designing stuff isn't what people think it is.
Total energy required to make any thing is a big deal. Oil is not actually renewable in quantity.
Anytime the headline is a question, the answer is “No.”
IF you can keep the heat down, nearly any electric motor can be severely over driven. I have done it many times. The limits are the voltage of the wire insulation and to not spin it so fast as to sling the lube out of the bearings.
I caught a glimpse of the future in this video. 3D printing with new materials will open myriad possibilities of future manufacturing. Not just electric motors, bu everything.
Even though there isn’t much substance here I still very much enjoyed this brief introduction to this. I also found the actual design of the coil and rotors extremely fascinating. The way everything had curved surfaces made it seem more organic to me for some reason. Is this design style arts nouveau. Anything that is in this style always seems psychedelic to me for some reason so maybe it’s more efficient “possibly “ due to its organic fractal design"
Game over, man, game over! - just like in Aliens!
3d printing is definitely the future for engineers due to as we start to master math each piece in every tech is going to be perfectly customized making no one product "part compatabuil".
Back in 2021 I designed a brushless axial flux motor that was based upon the Lynch motor design, but that would use a 3D printed coil assembly and smc core material. I contacted four leading manufacturers to ascertain the cost of prototype printing, and lets just say that I don't believe that we will see such motors become mainstream.
I found I could make a "good enough" motor using traditional manufacturing and assembly techniques for a fraction of the estimated production cost of the fancy 3D printed version.
Hell rethink bikes couldn't even make a business out of the binova flow motor which was an excellent product with easy manufacturing process...
Neat idea but I want to see it running. This is all speculation until there's a running prototype that we can measure. It does look pretty fancy for an electric motor.
I realized and predicted about A year ago that, AI will revolutionize learning and our capabilities as a whole. If AI can do anything, then we can do the same. We do control AI after all. The ceiling is imagination and knowing what you want.
I am sure the slow slow minds @ Battelle Memorial Inst. have neither printed these concepts nor tested the math in cooperation with OakRidge Labs.
This is good stuff to dream on Blender, or STL etc. for learning. We, the public, will Not use or make these ideas anytime soon. What is 'cutting edge' is most assuredly being used by military today, if not back in 2014.
Excellent episode.
One classical design that avoids many modern issues is the slip-ring three-phase motor. The rotor and stator are both wound, and the rotor windings driven through three continuous sliding contacts, so the current is carried without commutation. This offers some of the advantages of the universal motor, which similarly has both rotor and stator windings, namely a higher momentary peak torque relative to its heat-limited power capacity.
A permanent magnet motor can never be "turned off". Induction motors don't generally provide usable performance in regenerative braking, only (lossy) electromagnetic braking. A slip-ring motor has full control of the excitation of the magnetic field, so the motor can be allowed to free-spin with extremely low resistance, or loaded to any degree desired similar to an ICE car alternator.
This premise is hands down the the most terrifying concept I've heard in the last 45 years. Similar to an exponential curve essentially hitting a vertical wall of growth.
If A.I. is the short form for A.I.L. (all is lost) I can't wait for the third act and the resolution of the story and its subplots.
Very, very impressive. 3D printing has already been quite a quantum leap, and now paired up with AI - it's watch this space. I see a deluge of exciting new developments on their way.
Im a electrician working in maintenance. God pls never let me have to repair one of this. 😂
It will be interesting to see how slowly AI will learn how things function in the real world vs virtual.
Then how long it will take us humans to understand how it perceives the real world.
The problem with 3D printed coils is that they have poor surface roughness (in addition to porosity). As such, it takes alot more varnish to fully coat them. And even with the added varnish, the high spots will almost certainly suffer from premature failure because they greatly magnify the E-field right where the insulation is the thinnest. Unless and until 3D printing can produce surface finishes on par with traditional processes like drawing, don't count on motors built with 3D printed windings being able to last more than a few hours at full load. So it might be useful for things like expendable drones, but not for anything consumer, industrial, automotive, passenger and cargo aircraft, etc.
Is it a) more efficient b) cheaper to made c) easy to service & maintenance d) fast spin or high torque? than todays 3 phase motors?
Youre on the right track. We need better for compressors too. What id like to see is a motor that operates under strain better. The use of capacitors to quickly unload a larger charge to speed up ignition and distribute power on large machinery. Cars are not my priority. Im interested in farm machinery, big rigs, city transit, construction machinery. Im also waiting on a hydrogen engine design to power an electric generator.
ooh damn yeah, all this ai progress i have not thought about any 3d printer ai
And where was the AI in this exactly? Not everything designed on a computer uses AI.
Getting to that stage where will have forgotten how to build the Pyramids... making us entirely dependent on computers
I like my comfy climate controlled house and stable source of clean water and food I don't have to hunt myself.
We can ask AI to infer those lost skills and then train us in whatever your favorite learning medium is. Personally im happy the days of having to try juggle so many complex topics in my mind to get anything done is coming to an end. I applaud the ground breakers who went through that ordeal themselves.
@@6AxisSage I have coworkers who thought hydrogen wasn't flammable.
@@6AxisSage So you are the chicken holding open the hen house door for a solar flare to come destroy our civilisation?
@@pvanukoff So you will be one of the first to die when, NOT IF, a catastrophe strikes?
Betteridge's law of headlines is an adage that states: "Any headline that ends in a question mark can be answered by the word no."
Would recommend to use internationally known measuring units for videos which go out for everyone. "Pounds" isn't a good unit, considering that imperial measuring system isn't used by most of the world. As for information itself, nice to see new details, new ideas.
Reminds me of Bob Lazer's statement about the "Sports Cars" flying saucer he said he saw at the offshoot of Area 51. He said it was all one solid piece, with no seams. We are getting there...
Stupid title, I was hesitating to click. Turns out it's a nice video about motors, 3D printers, and not AI.
I know that for experimental motors 3D printed parts are great. But when it comes to long-lasting, production grade components, for example in aerospace, standard manufacturing wins the game because it does not suffer from the flaws inherent in 3D-printed materials - even the best printing techniques leave tiny wholes and imperfections in the materials, which break much earlier when exposed to the regular stresses the parts experience compared to their standard manufacturing counterparts.
AI hallucinates, so I'm skeptical until I see actual data.
I'm not sure i'd trust it though , it can't even count 5 fingers properly yet.
you are WAY behind on your Generative AI. your mainstream AI image platforms, like midjourney, ChatGPT 4.0, bing Chat, and stable diffusion have been able to do realistic hands for some time now.
@@tuseroni6085 Err ,yeah , that's not as good of a boast as it sounds... :) so it can count to 5 ...my cat can do that
@@ireallylovegod it's not meant to be a boast.
@@tuseroni6085 It's all humour ;) my cat can only count to 4 anyway
@@tuseroni6085no, it sucks
AI did not reinvent the wheel. Just made the wheel more effecient. Higher upper limit does not equal revolutionary. Just means incremental gains
I noticed your Penderev Motor back there.
How did it turn out?
Using materials that are good conductors, not easily oxidized and 3-D printable is quite a challenge. Doping copper with something more workable may have a side effect of rusting, which can then be negated by insulative design, having no air reach the wire so that they don't rust. There may be compensations to be made, but it is cool to think that the Mall, and it's stuff, of the future will be but a Box in a room in your house.
The brushless hobby grade motor is the best.Great speed and torque with the ability to be easily controlled.
I think the induction motor is simply a tradeoff for ditching magnets for more electrical consumption.
Multi metal powder 3D printing looks amazing, but wow… what a pain when recycling the mixed up metal powder leftover. It will be super cool to see what’s possible, even if it’s not practical for more than a few one off parts or tests.
Metal recyclers can only dream of having a material input as pure and clean as the mix of powdered metals left over after this, recycling is not an issue.
1:30 If it require cooling can this heat be used for other appclication like heating up water or something like that?
If your "waste" heat can't boil water it's pretty much useless other than a few niche applications.
We now need AI to develop a safe, cheap and long lasting battery! But will corporate want us to have access to that???
This is called generative design and it's actually used _by_ engineers to streamline workflow and create optimized designs based on simulation data. It's not game over, it's just another piece to the game.
If the coils are twisted in one direction does it mean the motor will perform better when rotating in that direction?
For mass manufacturing, I dont think 3D printing will ever gain momentum, especially expensive ones as this.
Maybe use optimization software with existing magnet wire. Then let it do the shape of the coils and magnet shapes. I imagine it's hard to make the rules though for it to follow for the copper wire so it's windable.
looks like the "Raxial Flux" motor which is an attempt to combine the radial and axial designs .
1:36
If this is a thing I'm going to really reconsider the electric car.
I like this idea as it opens the door to the “Bismith” UFO meta materials with microns of different layers together. A possibility that such materials were manufactured in this way.
I read claims on the internet and in UFO books about Bismuth, how it was affected by gravity and fell slower than other metals. So I bought a few pieces of pure Bismuth and found out for myself. (Those claims were false, pure bunk.)
Have to solve those printed metal “conductivity issues”…
So far, 3D printing of alloys does not appear to be able to take that much stress. A 3D printed rotor may not take the harsh centrifugal effects of spinning as fast and as long under all the stresses that are needed on the road... But that might change dramatically when we start 3D printing in space.
Super interesting for sure, but I have to file this one under, "I'll believe it when I see it."
Next step: We need AI to design the processes for building the things it comes up with.
3 questions:
Does your magnetic motor in the background work?
Did you discover the drawbacks to that design?
Is there a video about it somewhere?
Its only going to be a problem with 3D printing when not enough machines are at the assembly line , and no foundries and huge power to supply it .
The technology to make muliple material at the same time is the key to all this and or just the time taken for each product to be mass produced.
There is a way its just not worthy for companies because for them its not to show off a product but profits.
There will be a way ,just not yet and its obviously not far away
That's the kind of thing that excites me in life
I think the advantage that component assembly products, like Electric motors still has is, serviceability. You can take apart and replace components. 3D printed composites are not mechanically serviceable. How difficult is it to separate the metals in such a composite - in terms of recycling?
Not over at all unless you are slow enough to think it is the ultimate , never to be topped.
Yep, just a matter of time, in fact I have been toying with the idea of an impure soft copper mix the will change dimensions during acceleration to change the electric field at faster speeds to change the efficiency curve. Similar to how inertial clutches work but within the motor. No gears necessary
Thx for the information
we still have about 16 more years until the singularity occurs but as you can see it's well on track
Motors have been solved. There can only be small incremental improvements in this field which will not impact packaged product much. Batteries on the other hand...
It's not about the design....the materials👌
The title is like there is a new AI breakthrough and the game of electric motor design is over, the master design of all designs and there is none.
Toshiba Superconducting Motor, price unavailable. Koenigsegg Raxial Flux, Dark Matter & Quark e-motor's, price's unavailable. 3D printers using powder bed fusion (Direct Metal Laser Melting (DMLM) or Selective Laser Melting (SLM)) and Direct Energy Deposition MMA, provided you can afford anywhere from $80,000 to $1,000,000 on the printer, depending on type, size, performance, with one kilogram of copper filament, costing roughly $121 due to high copper density, makes this absolutely cost prohibitive for the hobbyist. At least, for now.
I think it's badass, unlimited designs, shapes ECT. Use powder material ,graphite, aluminum,steel, and also easy to insulated . Like' your dream is my command ' ,!
Scrapman just made a few videos on how the magnetic power fields of motors might work.
That the fields twist, then snap through eachother in entropy and revert back to their begining/starting states.
He just doesn't know it yet that his videos on that are about ai motors. 😅 .