House bill seeks to gut NASA’s Artemis plan, resurrect Journey to Mars

Lawmakers also appear to like cost-plus contracts.

Read the whole story

 

[Jaimie Chao]

It's sad more than funny.

2004: We're going back to the Moon!

2015: No, fuck that, we're going to Mars!

2017: No, shit, the Moon we meant we're going to the Moon!

2018: Sorry, that was supposed to be Mars.

2019: Moon, we meant to say the Moon.

2020: MARS, GODDAMMIT!! ON A BILLION-PLUS-DOLLAR ROCKET THAT CAN'T LAUNCH MORE THAN ONCE A YEAR!!! HAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHAHA!!!!!!!!!!!!!!!!!!!!!

 

[wagnerrp]

I also am suspicious of reliance on robotics. The asteroids are generally too far away for teleoperation. NASA has achieved a lot with rovers on Mars, but they've also demonstrated how limited that approach can be. A problem with drilling a hole that would take less than half an hour for a human to fix can set back a robotic mission by six months.

A human would fix it using their significant mass, and several hundred W worth of physical exertion. The drill package on InSight weighed 3kg, and was allowed 2W. How much better would it have fared if it were the primary duty of that lander?

SpaceX plans for Mars seem plausible to me because they use humans to set up the propellant plant. I suspect asteroid mining will be done by humans before it can be done by robots.

Curiosity was by far the largest piece of equipment we ever put on Mars, and it wasn't even 1t. Starship intends to deliver more than two orders of magnitude more to the surface, in additional to delivering itself and all the support capabilities it itself offers. You can't compare deficient operation of mass constrained missions to what could be possible with multiple 100t payloads of robotic equipment.

 

[BrangdonJ]

I also am suspicious of reliance on robotics. The asteroids are generally too far away for teleoperation. NASA has achieved a lot with rovers on Mars, but they've also demonstrated how limited that approach can be. A problem with drilling a hole that would take less than half an hour for a human to fix can set back a robotic mission by six months.

A human would fix it using their significant mass, and several hundred W worth of physical exertion. The drill package on InSight weighed 3kg, and was allowed 2W. How much better would it have fared if it were the primary duty of that lander?

SpaceX plans for Mars seem plausible to me because they use humans to set up the propellant plant. I suspect asteroid mining will be done by humans before it can be done by robots.

Curiosity was by far the largest piece of equipment we ever put on Mars, and it wasn't even 1t. Starship intends to deliver more than two orders of magnitude more to the surface, in additional to delivering itself and all the support capabilities it itself offers. You can't compare deficient operation of mass constrained missions to what could be possible with multiple 100t payloads of robotic equipment.

Nevertheless. Have you seen the Boston Dynamics robots that are so impressive? "Impressive" means one of them can open a door by turning a handle. That's where the state of the art is. And this is without limitations on mass or power. Without having to deal with Martian conditions. And the first robot couldn't even attempt it and needed the other to help. Do you think, if one of them broke down, the other could repair it?

https://www.youtube.com/watch?v=wXxrmussq4E

 

[wagnerrp]

Nevertheless. Have you seen the Boston Dynamics robots that are so impressive? "Impressive" means one of them can open a door by turning a handle. That's where the state of the art is.

It's Mars. Where are the door handles? Something that can open a door by turning a handle is only state of the art in a situation where you have to interface with the human world. If you interface with a world where the equipment and robot are designed in parallel to be compatible, things work a hell of a lot better. Consider industrial automation like a fully automated mill/lathe, that can take a piece of stock material from start to finish with manipulators and swapable tooling.

And this is without limitations on mass or power.

Of course there are. They have to be functional within a human environment, along side humans, and have to match human speeds and endurance. That places significant requirements for power, and limitations on size and mass.

And the first robot couldn't even attempt it and needed the other to help. Do you think, if one of them broke down, the other could repair it?

Who cares? Why do you need to perform repairs? Bring a dozen, and just push damaged equipment out of the way. Bring enough that any failed equipment can sit and wait for repair when the crew eventually gets there. You don't need a system that gets you 100% of the way to fully autonomous self-sufficiency. You just need something that can do most of the prep work, such that an initial crew only has to fill in the gaps.

 

[wyrmhole]

Nevertheless. Have you seen the Boston Dynamics robots that are so impressive? "Impressive" means one of them can open a door by turning a handle. That's where the state of the art is.

It's Mars. Where are the door handles? Something that can open a door by turning a handle is only state of the art in a situation where you have to interface with the human world. If you interface with a world where the equipment and robot are designed in parallel to be compatible, things work a hell of a lot better. Consider industrial automation like a fully automated mill/lathe, that can take a piece of stock material from start to finish with manipulators and swapable tooling.

And this is without limitations on mass or power.

Of course there are. They have to be functional within a human environment, along side humans, and have to match human speeds and endurance. That places significant requirements for power, and limitations on size and mass.

And the first robot couldn't even attempt it and needed the other to help. Do you think, if one of them broke down, the other could repair it?

Who cares? Why do you need to perform repairs? Bring a dozen, and just push damaged equipment out of the way. Bring enough that any failed equipment can sit and wait for repair when the crew eventually gets there. You don't need a system that gets you 100% of the way to fully autonomous self-sufficiency. You just need something that can do most of the prep work, such that an initial crew only has to fill in the gaps.

"Turning a handle is only the state of the art when in a well-controlled human environment. If you highly constrain everything about the environment and ensure absolutely everything is precisely controlled and designed for robots, then robots can do amazing things! And when I think 'environment entirely designed from the ground-up around ease of robotic automation with no surprises or variance, I think Mars."

 

[Faanchou]

Who cares? Why do you need to perform repairs? Bring a dozen, and just push damaged equipment out of the way. Bring enough that any failed equipment can sit and wait for repair when the crew eventually gets there. You don't need a system that gets you 100% of the way to fully autonomous self-sufficiency. You just need something that can do most of the prep work, such that an initial crew only has to fill in the gaps.

Your ISRU propellant plant just sprung a leak. Do you
a) repurpose some of the semiautonomous hardware on site to fix it
or
b) design and send a repair kit

 

[Stuart Frasier]

"Turning a handle is only the state of the art when in a well-controlled human environment. If you highly constrain everything about the environment and ensure absolutely everything is precisely controlled and designed for robots, then robots can do amazing things! And when I think 'environment entirely designed from the ground-up around ease of robotic automation with no surprises or variance, I think Mars."

I'd imagine a quadruped robot would be better suited to varied Martian terrain than the wheeled robots we've used so far.

 

[O/Siris]

"Turning a handle is only the state of the art when in a well-controlled human environment. If you highly constrain everything about the environment and ensure absolutely everything is precisely controlled and designed for robots, then robots can do amazing things! And when I think 'environment entirely designed from the ground-up around ease of robotic automation with no surprises or variance, I think Mars."

I'd imagine a quadruped robot would be better suited to varied Martian terrain than the wheeled robots we've used so far.

Maybe so, until power gets factored in. Quadrupedal motion seems to require a hell of a lot if (compared to wheeled, at least).

 

[Faanchou]

"Turning a handle is only the state of the art when in a well-controlled human environment. If you highly constrain everything about the environment and ensure absolutely everything is precisely controlled and designed for robots, then robots can do amazing things! And when I think 'environment entirely designed from the ground-up around ease of robotic automation with no surprises or variance, I think Mars."

I'd imagine a quadruped robot would be better suited to varied Martian terrain than the wheeled robots we've used so far.

Maybe so, until power gets factored in. Quadrupedal motion seems to require a hell of a lot if (compared to wheeled, at least).

It doesn't have to, but four is an awkward number and you tend to end up fighting yourself when you use locally linear models.

Hexapedal motion doesn't have to use any more power than a wheeled one, if momentum recovery is applicable. That is, if the drives are hydraulic or electric^†. It also allows more of the terrain constraints to be mostly about normal forces instead of friction. That's not really a problem on Earth, Moon or Mars but it could become an issue on other moons, dwarf planets and asteroids.

^†Instead of chemical, pneumatic or some such. What I mean here is that the power you supplied to the limb is not fully committed but can be recovered by a counteraction.