Live data from Hacker News

Artificial muscles robotic arm with full range of motion [video]

youtube.com

11–20 of 75 posts

Re: Artificial muscles robotic arm with full range of motion [video]

#11
For comparison, a decently trained human male can hold approximately 80kg (175lb) per hand (350lb total) fairly easily without any assists such as straps, and without using hook grip. With hook grip or straps, the strength of your back is the limiting factor.

That said, as a first step this is very impressive and very lifelike, too.

The main issue with this is going to be proprioception more than anything - it doesn't seem like you'd be able to do what robots do: rely on position encoders. To make this do something for real it's going to need a way to know its own position, and feel pressure.

Re: Artificial muscles robotic arm with full range of motion [video]

#12

A few comparisons: > forearm with hand [weighs] only 1 kg That is slightly below a human forearm with a hand. > This artificial muscles robotic arm is operated by water and consumes 200W at peak The ideal consumption necessary to raise this 7 kg dumbbell by 1 meter in 1 second is 7 watts. Human muscle is about 20% efficient, so it probably consumes about 35 watts. Almost a tenth the power. A human arm can also achiev…

> Human muscle is about 20% efficient, so it probably consumes about 35 watts. Almost a tenth the power.

Being just an order of magnitude away from biology that took millions of years to develop and evolve is really impressive in my eyes!

> That feels like a stretch, given the complex needs of a brain (not just in glucose fed through veins, but also immunity protection and so much more). Brain transplant into an organic body seems easier, and even that does not address the fact that brain cells themselves age.

That said, this is probably the more reasonable argument. Ideally, eventually we'll just find ways to prevent much of the cell damage with aging - initially it probably won't lead anywhere near "biological immortality" (so it's pretty likely that everyone reading this will die and will cease to exist), but at the very least it'll lead to better quality of life at the more advanced ages and less human suffering overall!

Definitely an area that should have anywhere from 10x-100x the resources that have been allotted to it currently. Of course, one can also dream about staving off the eventual and inescapable clutches of entropy, if that allows them to live without existential crises and dread sneaking up on them, as it does with me.

Regardless, even advancements in robotics are immensely useful, paving way to everything from improving the quality of life for those who've suffered accidents, are differently abled in various ways, or have other factors that otherwise make their lives more challenging.

As for robot transplants or even living transplants, unless you can connect back all of the nerve endings that map roughly to what the host body needs, it's likely that it'd be like being fully paralyzed. In that regard, cryonics would be a better bet, to preserve the human being until the medicine has advanced to the point where any ailments and such can be addressed. Of course, seeing how new and still fledgling (and niche) the industry is, it's unlikely that many if any of the people will ever actually be successfully revived, much like we can't unscramble the brains of mummies in tombs.

But hey, if nothing else, it's an interesting pursuit and the trans-humanism movement is always curious to behold. Who doesn't enjoy racking their brains, imagining engineering solutions to problems that no one would have even considered a century ago?

Re: Artificial muscles robotic arm with full range of motion [video]

#14
post #11

For comparison, a decently trained human male can hold approximately 80kg (175lb) per hand (350lb total) fairly easily without any assists such as straps, and without using hook grip. With hook grip or straps, the strength of your back is the limiting factor. That said, as a first step this is very impressive and very lifelike, too. The main issue with this is going to be proprioception more than anything - it doesn'…

good points, if the system had tension connectors to other points on the body than it would be able to increase output. an arm using more than what is available in the arm, other parts of the body help the strength of an arm. think jiu jitsu and types of hits, and their impact

Re: Artificial muscles robotic arm with full range of motion [video]

#16
post #11

For comparison, a decently trained human male can hold approximately 80kg (175lb) per hand (350lb total) fairly easily without any assists such as straps, and without using hook grip. With hook grip or straps, the strength of your back is the limiting factor. That said, as a first step this is very impressive and very lifelike, too. The main issue with this is going to be proprioception more than anything - it doesn'…

I have a really bad back and as a result I can't really exercise. So let's say I can carry 8kg max per hand (1/10th).

And yet I'm a fully functional adult human being and can get by perfectly fine in life.

So for me from a robotics point of view a much smaller starting range is perfect, once you've built it into a platform you can start optimizing and increasing potential.

Re: Artificial muscles robotic arm with full range of motion [video]

#17
post #10

This guy is a great engineer, I’ve been following his stuff for a while now. But there are unfortunately several fundamental limits which mean this type of system will likely never reach human level performance in a real world system. I work in robotics, specifically compliant and modular actuators, similar to the ones in the video. As I see it, (and many others in the field) One of the biggest issues in robotics is…

This and op is all wholesome content.

On the criticism, do you think more limited / lab-bound systems will help develop control technology in advance of better muscles?

Re: Artificial muscles robotic arm with full range of motion [video]

#18

A few comparisons: > forearm with hand [weighs] only 1 kg That is slightly below a human forearm with a hand. > This artificial muscles robotic arm is operated by water and consumes 200W at peak The ideal consumption necessary to raise this 7 kg dumbbell by 1 meter in 1 second is 7 watts. Human muscle is about 20% efficient, so it probably consumes about 35 watts. Almost a tenth the power. A human arm can also achiev…

> achieve human immortallity [sic] by transplanting the brain into the machine

the intuition is that it's probably easier to figure out how to keep a brain alive than an entire body alive

Re: Artificial muscles robotic arm with full range of motion [video]

#20
I love how this project looks. Feels closer to Metal Gear Solid 4 sci-fi bio-tech, than to regular mechanical arms.

Maybe falls a little bit under the uncanny valley but with clothes on, I could see this achieving a pretty natural look to a prosthetic, plus this type of design seems more analog to an actual arm in terms of control.

I know that there is a long way for real-life use, and maybe I'm naive because I don't know all the caveats that need to solved to make it feasible, but this type of tech gives always puts a smile on my face.

Post reply on HN