You can see in the first shot that moving the robot over one area on the table did provide some varied readings -- this was also a bad table for optical mice. Our rig to keep the height of the sensors consistently within the range that they'll accept worked reasonably well in an empirical sense (we were able to roughly calculate position and rotation), but we didn't perform any measurements of their accuracy, nor did we attempt to utilize the sensors as part of a closed loop feedback system.
Thanks for the props. Especially since UCSB is on a quarters system, there's no way we could have done this in a single quarter. We had about 10 weeks to do the design, schematics, and board layout. In the next 10 weeks we waited for boards to be spun and populated and started some of the initial software work. In the final 10 weeks we did the rest of the software and did all the systems integration.
were all of these people computer engineer undergrad
roaringriku 2 years ago
yes
kekiefer 2 years ago
You can see in the first shot that moving the robot over one area on the table did provide some varied readings -- this was also a bad table for optical mice. Our rig to keep the height of the sensors consistently within the range that they'll accept worked reasonably well in an empirical sense (we were able to roughly calculate position and rotation), but we didn't perform any measurements of their accuracy, nor did we attempt to utilize the sensors as part of a closed loop feedback system.
kekiefer 3 years ago
Pretty awesome project. Was curious though, were the Agilent sensors consistent in their measurements? Was the robot where they reported it to be?
robologist 3 years ago
whats with the random music? kinda distracted me
wolfenthegreat 3 years ago
I think the music enhances the video -- dramatizes what the robot is doing.
Pigfarmerfromdakota 3 years ago
Thanks for the props. Especially since UCSB is on a quarters system, there's no way we could have done this in a single quarter. We had about 10 weeks to do the design, schematics, and board layout. In the next 10 weeks we waited for boards to be spun and populated and started some of the initial software work. In the final 10 weeks we did the rest of the software and did all the systems integration.
kekiefer 3 years ago
Kickass...I can't imagine it all got done in a single term, was it a year-long project course?
aelfdane1 3 years ago