I made some changes to the configuration of the braid that increases the stroke of the muscle. Coil length remains unchanged.

lol it has an... organic appearance. Using that finger trap cable management tubing was clever. The goal is to have a muscle-like appearance during contraction, if i understand correctly?

Many have been joking about the appearance lol That braided sleeve serves a purpose actually! It helps enable passive return of the actuator during relaxation, so you can drive underactuated joints with it. Also helps protect the assembly inside, and look cool 😎

Oh okay, neat. Gets pretty tough at these diameters. What's the final application you want to test this with? What kind of bot?

We're working on a humanoid right now and testing out several approaches to building artificial muscles for them. These electromagnetic ones are just one type that'll compete for the winning spot.

Internally, or is there a larger competition for humanoids? I haven't been involved in the space, just industrial.

Oh internally, just to see what we'll end up going with. If there actually was a competition for it I'd be interested!

There should be imo. The concept is certainly going to take off eventually, just not in every realm. I think the rednecks at the plant would laugh if you presented them an anthro cobot or something but they have smooth brains. For normal human interaction, as long as you stay out of the valley, I think people might prefer something more like themselves. The sexbot end of the spectrum could wreck that though, I see a lot of rejection of the idea out there.

Yeah, industrial robots are a thing for a reason (plus they're already rugged for their environment). Easy to maintain too vs the current state of soft robots.

Can you make it crawl around like a worm? I've seen some worms move exactly like this.

Sure I could! Might build a worm bot as a side project.

@Lazi_Learner Please build the fastest robot worm ever built.

@Lazi_Learner It will win races, I guarantee it.

Are you selling these

Not yet, but we plan to soon! Need to perform more characterization and testing first. I've got several other artificial muscle topologies in development as well, so there may be a few different types available.

Dunno what you're using for your outputs but I've been having good luck with these. If you don't mind the NPN outputs, they're 5A/channel. They're only half-enhanced at 3.3V though ,all of these modules lie about 3.3, use 5V.

Oh these could be useful! I'm using a Cytron MD20A driver right now for testing, but this could come in handy for a bunch of muscles that only need to fire in one direction. Passive return and all...

Very responsive. What's the stroke length? Can it be in series? Also curious about the power draw

I've got close to 50mm of stroke on a 120mm actuator. So far I've tested them in series and parallel, and they don't lose much performance with added muscles. The tests I've shown were all done at 12V, 3A (PSU setting). Average power draw to achieve full stroke was around 14W (per muscle).

Wonder what's the minimal form factor like if you are to pack 100 of them together (10 by 10) without losing performance.

Good question, these could be greatly scaled down in diameter to densely pack them. We could go from 22mm to 12-15mm without a significant loss in overall performance, if you pair some together to drive the same DOF. Similar to bundled organic muscles. Larger diameter muscles will always be more power-dense individually, but that could be matched by using several smaller diameter muscles in a smaller overall form factor. Definitely worth running some experiments on, I will share the results!

How much do these cross talk?

The two muscles side-by-side repelled one another a bit due to the magnet assembly inside. I thought they would get jammed or slow down during operation, but the opposite ended up being the case. Maybe I could leverage that to improve force output with muscles in parallel?

Cool

質問ですが、この人工筋肉の張力は電圧に依存しますか？　一定の長さを保つことはできますか？

Good question! The tension and force applied is dependent on the voltage, and the position is determined by the current. So you could make fine adjustments with a controller capable of both voltage and variable current control.

素晴らしいです！isometricな動作もできるのであれば、反応速度も十分ですし、人工筋肉として非常に有望ですね。

can it bend, can you regulate speed of contractions?

Stroke of the muscle, you say?

It that a transvestite appendage solenoid actuator

Bro what's this supposed to be 😭😭

I'm happy every time I see one of your posts As a kid I'd see science shows talking about how one day we'd see mechanical muscles, while they showed a bespoke prototype using exotic materials Now, I'm seeing updates of this while it's on it's way to real world use

if we replace human bicep with bunch of these motors, how much "strenght" it can produce relatively to the original one, in % ?

That’s getting dodgy :)