Robonaut To Escort On Space Shuttle Mission

An anonymous reader writes "The STS-133 crew will deliver robot Robonaut 2 (R2) to the International Space Station. Cocooned inside an aluminum frame and foam blocks cut out to its shape, R2 is heading to the station inside the Permanent Multipurpose Module in space shuttle Discovery's payload bay. R2, with its humanlike hands and arms and stereo vision, is expected to perform some of the repetitive or more mundane functions inside the orbiting laboratory to free astronauts for more complicated tasks and experiments."

Wouldn't it be better outside?

[Anonymous Coward]

... repetitive or more mundane functions inside the orbiting laboratory...

As a layman, I'd think the more troublesome tasks on the ISS involved the hassle of getting into an EVA suit for activities outside the space station. In which case the bulky pressure suit and heavy gloves are likely to give an astronaut no better manual dexterity than what could be provided by a robotic waldo. So wouldn't it be smarter to have the robot on an armature external to the station? Or is this just considered a testing phase to make sure that such robot is reliable and flexible enough before putting it on the more appropriate outside work duty?

And whatever happened to the Boba-Fett look of the previous prototype? I thought that was cool.

I love this stuff

[DomHawken]

As an oldish (40+) programmer who's only recently got into robotics (the simple stuff - arduino, sparkfun, xbee, khr3-hv), coding the control software to make a robot actually do stuff is way more challenging than the supposedly complex projects I work at on my day job. Programming a robot kit to walk or pull poses is simple enough, but coding 'any form of 'intelligent' decision making ends up with lines and lines of code and as many sensors you can sensibly add to the hardware.

I thought it would be pretty simple to build and code a robot cleaner - like a basic remote control car that just drives around the house with a duster underneath which heads back home when the batteries are running low and recharges. Clearly the challenge of climbing the stairs can move to the version 2 release, but if I stick it on the first floor, just stopping it falling down the stairs needs around five sensors and over 500 lines of code.

Two cameras for 3D spatial awareness? Try coding it to tell the difference between and apple and an orange. Built in GPS to get an absolute position reference? Even if you get a signal, 5 meter accuracy doesn't help much when you are driving it towards a lift shaft.

That's why I love this stuff.

Re:about fucking time

[khallow]

Agreed. Unfortunately, anything past that is going to be worthless. Benefit? Fully autonomous systems are going to get built extremely well.

Heh, and we get to the point about having humans in the mix. No need to develop fully autonomous systems (especially, if you didn't want the system to be fully autonomous), when we already have fully autonomous systems that we use every day and trust.