ok here`s the deal :
since we`re flying we`re trying to film each other while doing some aerobatics or just flying in nice surroundings we could use a drone like UAV that could follow us in flight.
To me the flying part is the easiest but the programming part will be the hardest for me.
I`d like to learn the programming with some help from the OPENgarage geeksquad and then build an aerial vehicle to make the whole thing fly.
stuff i have in mind to use
You need to be signed in to comment
|
3 months ago |
ok , some good point some less, first things first : why a multicopter ? since we fly in instable conditions (themals, wind) the multicopter becomes the only option suitable for filming and making it doable to navigate. the whole thing would become way too dynamic to make it fly a somewhat accurate path. Landing would be a hell of a job with anything that`s not VTOL. |
|
3 months ago |
A showstopper with such a plane is the collision detection. It’s probably not doable. With only one paraglider in the air, with a radio beacon and software calculation future paraglider movement and taking that into account for the own flight path is probably doable. Yet there are other paragliders as well and there is the terrain. I think that doesn’t work, sorry… imagine soaring a ridge, the drome will fly into any paragliders lines or into the mountain (proximity sensors only work for very small distances). Only laser-guided systems such as Google is putting on its driverless cars can make that happen, and they’re probably too heavy… probably better to drop the idea of ‘completely autonomous’ alltogether and create a remote controlled cam-based drome. Like the ones you can already buy… but what’s the novelty then? |
|
3 months ago |
For the drone itself, a plane is a lot better energy useage wise and speed-wise than a helicopter/quadcopter. But for keeping and maintaining the distance, a helicopter is off course better and more acurate, the flight path calculation is also a lot more simple with a heli. However with heli’s the autonomy is an issue if you are looking for electrical powered ones… wikipedia told me the that rc helicopters have a maximum autonomy of about 12 minutes. Not useful. At all. So a RC plane then? This can potentially offer the autonomy we need, and the RC plane can also benefit from thermals as we do, increasing autonomy further… |
|
3 months ago |
For following paragliders, the speed range of the drone should be similar as what a paraglider is capable off… any pre-made vehicle in mind? Also, do you want it to ‘follow’ at a predefined minimum and maximum distance; doesn’t seem so easy? Will require some decent image analysis and object recognition, probably needs more processing power than the raspberry pi can offer. There are some remote controlled dromes already, but I think you probably seek it to be with complete autonomy? Good luck with that; unless you have at least 5 man-years of programming to offer… object recognition in real life video isn’t easy. A low cost 500m-range radio which emits something by the paraglider might be an option there. Something on which the drome could home in. Will avoid the excessive programming required for image processing and analysis. |