corvi (corivax) wrote,

That time of year again...

This year's proposal for zero-gravity flight for NASA due in a week. Whee and ARGH and everything in between. I am, of course, pulling my hair out and falling asleep on my keyboard (hopefully all paragraphs typed with my forehead will be deleted before NASA gets the documents - I don't think that's what they mean by stress-testing!) and forgetting to eat.

There are types of stress that make us want to hide under our beds and tremble, and types that make us just clench the knife tighter between our teeth, and climb a bit faster. I believe I'm happy, in a rather masochistic sort of way.

Livejournal brain-trust: NASA likes us to present an 'Outreach Plan' detailing how, exactly, we will be doing public relations with our experiment. They make it clear they want us to involve 'minority groups and institutions', yet they also claim we should be creative. We've done the rounds of Seattle Robotics Society meetings, demonstrations to classrooms of fifth-graders, and obligatory website with grainy video footage of team members floating upside-down in zero gravity. I am, however, stumped for anything either creative or involving minorities. Does anyone have any suggestions/useful connections?

(GYRE is a cubical robot, fourteen inches on a side, which uses three cameras to observe the world and twelve compressed-air thrusters to move around in zero gravity. It's not all that impressive on the ground - the thrusters aren't strong enough to make it hover or anything, but it can be hung from a frame and allowed to spin itself around for demonstrations.

This year's experiment actually aims to discover the best way for the robot to react to its motions once it figures out which direction it's drifting or spinning in. As an example, suppose you wanted the robot to just float in place somewhere out of the way, and it noticed it was floating up. It might turn on the thrusters on top of itself to move back down. If it left those thrusters on until it was back where it was supposed ot be, and then shut them off when it arrived, it would keep moving and overshoot its position. It needs to be able to guess when it's getting close and power down, and we're testing various computer programs to do that.

Or, in geek: we're trying to configure a decent PID (or PD, or PI) feedback loop based on the camera input for visual servoing.)

  • Post a new comment


    default userpic
    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.