Friday, August 3, 2018

Origins and Status of RAM

[Dan and I] are back in Maryland for the final push; RAM successfully integrated and is cleared for flight, but we haven't had the time for our entire software team (read: Dan) to actually, you know, write and test the software for flight. We have 16 days before 'pencils-down' and we ship to New Mexico. This is a rigid due date, so we're in a stringent hardware lock even though we ended up pulling a lot of our camera system out in order to duck some power-brownout issues that surfaced recently. We'll (I'll) have to clamp down on any (my) temptations to try and sneak features or components back in.

RAM is working again, resurrected and reassembled. We burned out one of our critical servos due to some bad commanding, but swapped components to get us limping along again (though with a dead gripper, as shown in the video below.) By this evening we'll have a fresh one swapped in and will be back to full functionality.

Here's a click-through link in case the embedding doesn't work.
One of the big hopes we had for flight was to run completely autonomously, letting cameras and computers and reference tags (the blocky black-and-white bits in the picture below) do all the driving. For some of the reasons mentioned above as well as some other schedule slips -- most of them mine -- we haven't been able to develop that control feedback system to the level we'd want for flight. 

Because everyone loves a picture of a picture, this is the BusyBox from inside the vacuum chamber, as captured by a high-resolution CSBF camera and output to a laptop over radio link. Thanks to Chris Field for letting us piggyback on his beautiful imagery system.
However, we're still on track for an exciting flight! We're shifting to doing most of our driving by 'dead reckoning', i.e. with absolutely-referenced instructions sent from on board. Rather than reacting in real-time, the arm will be moving -- blind, but moving -- to manipulate all the components on our 'Busy Box' using explicit commands we plan out for it. We'll see if we can get some lighter test elements working and tested well enough for flight, but right now we need to keep focused on the minimum deliverable of a working arm.

Thanks again to Peter Sooy for inviting us up to Goddard way back in December of last year; the information we learned during that visit helped inform the shape of the project that we're mere weeks away from launching into near-space. Thanks also to Zakiya Tomlinson for fielding all of our questions and showing us the ropes. And, finally, thanks to the patient HASP ground crews from LSU and CSBF for hosting us last week: Dr. T. G. Guzik, Doug Granger, Josh Collins, and Anthony Ficklin were all more than accommodating and gave us everything we needed to succeed. And, last but far from least, thanks to NC Space Grant for their continuous support and funding; we couldn't do this without y'all!

No comments:

Post a Comment