What? Dancing robots going to mars?

[PB666]

http://www.bbc.com/news/technology-34845483

This cannot be for real. Whats the point of a robot. Anyway the link is glitchy.

[Findthepin1]

December Fools

:D

[fredinno]

No, it's a robot being tested by dancing, which is somewhat complex. Don't mistake the two just to get clicks, please.

[PB666]

Dancing tests what essential function for scientific experimentation on mars? i seemed to have missed the non-elective course in dancing while in graduate school. 

How about testing robot by bend knees while maintaing balance, then bend over and pick up rock and place in basket?

How about take shovel and dig down, then sweep off remaining dirt, then holster broom, unholster coring drill, drill core, reholster coring drill, remove chilsilng hammer chisle out section of rock along side core, chisle out core.

Take probe insert into loose soils, analyze water percent, follow grid. 

[micr0wave]

but they can do it with style

[RainDreamer]

52 minutes ago, PB666 said:

Dancing tests what essential function for scientific experimentation on mars? i seemed to have missed the non-elective course in dancing while in graduate school. 

How about testing robot by bend knees while maintaing balance, then bend over and pick up rock and place in basket?

How about take shovel and dig down, then sweep off remaining dirt, then holster broom, unholster coring drill, drill core, reholster coring drill, remove chilsilng hammer chisle out section of rock along side core, chisle out core.

Take probe insert into loose soils, analyze water percent, follow grid. 

I still don't see why it has to be humanoid to do all those things.

 

Anyway, actual news link with details here: http://www.ibtimes.co.uk/nasa-gives-r5-valkyrie-humanoid-robot-mit-northerwestern-develop-software-mars-1529363

Edited December 7, 2015 by RainDreamer

[Nibb31]

It's because it's NASA, and everything NASA does these days has to be tied to Mars some way or another.

A humanoid robot makes sense if it's designed to use tools and vehicles that were designed to be used by humans. Other than that, anthropomorphic robots are a silly idea. The human body isn't some sort of universally perfect form. It's more often a compromise that performs suboptimally for most specialized tasks.

[fredinno]

1 hour ago, Nibb31 said:

It's because it's NASA, and everything NASA does these days has to be tied to Mars some way or another.

A humanoid robot makes sense if it's designed to use tools and vehicles that were designed to be used by humans. Other than that, anthropomorphic robots are a silly idea. The human body isn't some sort of universally perfect form. It's more often a compromise that performs suboptimally for most specialized tasks.

It's probably just for research, to see how far they can take the bot.

[Nibb31]

Well, the same was true for "Robonaut 2". Sending a humanoid robot to the ISS was a cool PR stunt, but it's been pretty much collecting dust and taking up space ever since.

[Rakaydos]

18 hours ago, Nibb31 said:

It's because it's NASA, and everything NASA does these days has to be tied to Mars some way or another.

A humanoid robot makes sense if it's designed to use tools and vehicles that were designed to be used by humans. Other than that, anthropomorphic robots are a silly idea. The human body isn't some sort of universally perfect form. It's more often a compromise that performs suboptimally for most specialized tasks.

For specialized tasks, sure, but what about as a general purpose design? Can it do everything "well enough" to accept the compromises?

[Nibb31]

1 hour ago, Rakaydos said:

For specialized tasks, sure, but what about as a general purpose design? Can it do everything "well enough" to accept the compromises?

I don't think so. Look at locomotion for example: a bipedal robot has more single failure points than a 6-wheeled chassis. It's also less stable, meaning that if it does fail, it can fall down and damage its payload, whereas a wheeled platform will remain stable and can still work as a fixed station.

And the same is true for arms. Why go with only two arms. Why not four or five ? Why use an imitation of the human hand when you can design a claw that fits much better onto those EVA rails, and a purpose-built electrical/mechanical interface for tools.

Edited December 8, 2015 by Nibb31

[Rakaydos]

46 minutes ago, Nibb31 said:

I don't think so. Look at locomotion for example: a bipedal robot has more single failure points than a 6-wheeled chassis. It's also less stable, meaning that if it does fail, it can fall down and damage its payload, whereas a wheeled platform will remain stable and can still work as a fixed station.

And the same is true for arms. Why go with only two arms. Why not four or five ? Why use an imitation of the human hand when you can design a claw that fits much better onto those EVA rails, and a purpose-built electrical/mechanical interface for tools.

A 6 wheel chassis is specialized for "relatively" level ground. After a certian point, enough articulation to get over increasingly rough terrain results in a 6-legged walker with wheels on it;s feet anyway.

A Bipedal walker embraces instability, like a bicycle. It requires much more on-site processing as it requires a minimum "reaction speed" to compensate for it;s own instability, but if a bipedal  sence of balance becomes a solved computational problem, combining bipedal tool use with quadrapedal vertical mobility ("climbing") results in a flexible exploratin platform.

[RainDreamer]

Why bipedal when you can do spider bots?

But anyway, this bot is for testing human environment on mars, I assumed, so it has to be human in form.