An internal competition where I work is being held to come up with a Search and Rescue UAV by November of this year. There any many requirements, but I'll list the main ones below. The budget is $12.5k and we're starting from scratch. As an owner of an ArduCopter 3DR Quad, my mind jumped immediately to the conclusion that some form of an ArduCopter would be perfect for this competition. Am I biased since I own one, or are these requirements within the current or near term capabilities of the ArduCopter platform?
Requirements:
UAV must take off vertically, fly a pre-programmed route using GPS coordinates, and then return to the take off position and land, all autonomously.
UAV must stay between 50 and 1000 ft altitude during flight
Pre-programmed route will be a search grid of 1000ft x 1000ft.
UAV must stay within search grid and ground station should show search grid boundaries on map
Live video must be streamed back to ground station
Payload/camera can be controlled manually from ground station
Manequins will be placed within the search grid and be manually identified from the video stream, then tagged with GPS coordinates
Once flight is complete, UAV should hover over landing spot as long as possible
Weight must be under 5kg total, including any payload
UAV must fit assembled in a pickup truck bed
$12.5K total budget
Must be built and tested by November 2012
My initial thought is a 3DR hexa platform, APM2, a 2D camera mount, HD Hero2 camera, video transmitter, and the largest batteries it can carry. Assuming building 2 of them, the Radio, battery charger, etc, I think staying within budget is no problem. Thoughts?
Replies
Garrick, I live in Florence, less than an hour away. I have been working with this hardware for some time now. I don't think this is difficult. I have most of everything you would need to do that already. And it didn't cost no where near 12K :)
Where do you work and are they still hiring ? :-)
Seems reasonably within the current capabilities of arducopter, except the drawing of the boundaries on a map. But maybe you can come up with something if you take a close look at the geofencing code of arduplane and the sourcecode of the mission planner.
Looks doable currently, as long as there isn't too much wind for the landing portion. Currently the auto land command doesn't hold position and drifts with the wind until touchdown.
I don't know how to show the search grid boundaries on the map, but you could just manually draw it on a map and upload the custom map.