Hey
Me and a friend of mine are trying to figure out how to build a small multirotor aircraft that we can use for the robocup at our uni. We want to be the first to bring a flying robot.
The plan is to use the APM, we have used it in some projects, both quad and fix-wing, so we are familiar with the APM and the code.
Our problem is that the gates the robot has to fly through is only 45cm +/- 2cm (so 43cm worst case), it has to follow a line, we plan to use the Arduino "Video Experimenter shield" https://nootropicdesign.com/ve/ - having it running on a separate arduino (a mini), and have the two talk together.
As of now we need to find a frame that is small enough so it will fit through the gates (with some margin of error on the line tracker) and powerful enough to lift all the gear (camera a Sony 600TVL, APM, videoboard ect.).
The judges say that the props have to be shielded (so we need a box around the quad as wel), to protect people from the props if something goes wrong. This adds to the width of the airframe
Flight time needs to be around 5-7mins
more info can be found here (sorry for the google translate):
http://translate.google.com/translate?sl=da&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.robocup.dtu.dk%2Fkonkurrencen%2Fbanen_2014&act=url
we plan to "just" follow the right side of the line so we don't need to do other tasks then follow the line (mål is the finish line).
Any ideas?
Thanks
Regards
André
Replies
First off have you thought about something like this design it really is quite small it's designed for 5" props so maybe 12" overall envelope. Also the additional arduino and video shield seem unnecessary. How about using minimosd for FPV instead https://code.google.com/p/arducam-osd/wiki/minimosd
Thanks for the reply.
The copter you posted seems to be of the size we want, I'll look more in to that.
In regards to the ekstra vide shield, it is not for FPV or an FPV overlay, it is to do immage processing, so that the copter can fly by it selv and follow the line, I dont think the miniOSD can do image processing, I might be wrong.
If it can do it however it would be a lot smaller and light weight solution the the problem.
Ahh I see. Does the video processing have to happen onboard or could you run a base station from a laptop to examine the video stream and send corrections?
No, all calculations has to be done on the copter, sadly...
ahhh that's going to add additional complexity in terms of power consumption and weight. Assuming the nootropic can do that for a reasonable power + weight cost that may well be the best option. Or you could try to hack a Kinect to acheive this http://www.instructables.com/id/Kinect-Controlled-Quadcopter/ but you'd need to put some kind of lightweight Linux machine onboard that was both powerful enough and light enough. Maybe a parallela
Think the complexity will be too big for with a kinect for a small project like this.
We will make a Twincopter and try to modify the APM code to make it fly.
If that fails we will look in to a miniquad.
Thanks for the input.