Jump in the discussion.

No email address required.

I don't understand why anyone is still trying to develop humanoid robots for in-factory manual labor. A 6 axis arm on a set of wheels is orders of magnitude more efficient.

It's existing tech, stable, uses less power, less overall movement, lower maintenance, and is a fraction of the cost.

But you know what's even more efficient?

A conveyer belt system.

All these automation showcases gloss over the fact that every scenario where a humanoid robot is used could be solved by modifying the process layout more efficiently, making the need for a humanoid shape completely irrelevant. Watch the conveyer video, count the boxes sorted in the single minute of footage. Then take that number and multiply it by the time it took this Atlas bot to move one item.

Jump in the discussion.

No email address required.

Atlas is more of a technology demonstrator but the benefits of humanoids is that existing factories are already design for use of bipedal humans, doing simple tasks like climbing stairs without the need of building a new ramp for non-bipedal robots. However I always had the impression that humanoids like Atlas are being tested for other purposes outside industrial controlled environments.

Time will tell if they're just hype, so far there's no market for humanoid robots, though Agility Robotics digit is entering mass production next year :marseyshrug:

Jump in the discussion.

No email address required.

I'm going to lean towards hype. I've been around manufacturing facilities across different sectors for 15 years now and seen probably a million square feet of production floor layouts. In my experience, stairs exist only for office mezzanines, not material movement since instances like that are perfect examples of needing a layout change. And if we're talking about older facilities, they're effectively built flat because that's generally how every facility got built.

I can tell you that my work has 500k+ square feet, we entertained the idea of humanoid bots for like an hour and then opted for changing the layout for a few thousand dollars and no additional complexity from having robocop here.

If they have a different addressable market for them besides fabrication/manufacturing areas, they need to start demonstrating that because I'm at a loss as to where someone would use these

Jump in the discussion.

No email address required.

Can't quadrupeds like Spot not already traverse most obstacles such as stairs just fine? I really don't see a reason for making it a biped when you can just give it as many limbs as you'd like :marseyhmmm:.

Jump in the discussion.

No email address required.

I have a hunch that the real reason DARPA-funded Boston Dynamics is developing 6 foot tall autonomous robots isn't to fill an Amazon warehouse with them…

:#marseysquint:

Jump in the discussion.

No email address required.

I'm still of the mind that DARPA is going to end their humanoid robot funding (that's already dropped what, 95% since 2013?) as soon as they get drone swarm development beyond LRIP.

I like the futuristic vibe of the robots but it's the same case where the human shape generally sucks compared to fit to purpose form. An autonomous drone with a c4 explosive and another drone with an auto aim pistol can clear a building without having to worry about whether you're going to lose a 6 foot tall $200k piece of DARPA tech in the Middle East.

Jump in the discussion.

No email address required.

the humanoid robots are intended (eventually) as drop-in replacement for human workers. over time a nursing home can be modified so that most worker don't have to be humanoid anymore, but they won't be conveyor belts either.

Jump in the discussion.

No email address required.

Link copied to clipboard
Action successful!
Error, please refresh the page and try again.