Need Help! Development of 3 drones for flight formation and acrobatics performance

[Glenn876]

As the title stated for my new project(formation flight), how do I begin with?

1st - What drones should I buy? I have a budget of <$400 for each drone. (Must be programmable for them to stay in formation and do stunts)

2nd - How do I program it for them to stay in formation?(eg. staying in a triangular shape position)
what do I need? any extra parts?

3rd - How do I control all 3 of the drones in 1 controller?

This is for my school project and I have little knowledge about drones so any help would be appreciated .
Thank you!

 

[theGreenOrange]

As the title stated for my new project(formation flight), how do I begin with?

1st - What drones should I buy? I have a budget of <$400 for each drone. (Must be programmable for them to stay in formation and do stunts)

2nd - How do I program it for them to stay in formation?(eg. staying in a triangular shape position)
what do I need? any extra parts?

3rd - How do I control all 3 of the drones in 1 controller?

This is for my school project and I have little knowledge about drones so any help would be appreciated .
Thank you!

When you set up your kickstarter page, you better set a goal of about a million bucks. That should give you a nice start.

 

[jamesb72]

OK I'll bite, I'm not aware of any of the shelf quads which can fly in formation, you have to do it the hard way.

You would probably do best with something like QX90s, they are <$70 each, and have a decent FC, to get them to fly in formation you will need a 3d camera rig and some custom software to track and control each quadcopter from an external computer - this is a pretty big challenge, so unless you have a team of developers to help, I suspect will be the sticking point.

To simply fly 3 quads in a roughly similar shape for a few seconds, you can find some cheap quads which will bind multiple quads to a single transmitter, I've seen this done pretty effectively with GPS quads (the early 350QX had a demo video of two flying in sync on a single transmitter), but for cheap non-GPS quads, you are only going to get maybe 30 seconds of flight before they will drift out of sync and it will all get messy, but this would be much cheaper as potentially you could get 3 quads for under $100 total, and you could do a quick flight with some flips etc with 3 in sync with each other.

Just found the 350QX flight James Haley flew this - this is no longer a current model, but you could theoretically do this with 3/4/5 350QX v1s, as they used DSM2/22ms which would allow multiple receivers to bind to a single transmitter - anything newer than this (DSMX/DSM2-11ms capable) won't as they have an acknowledged bind sequence I believe, so only one receiver can be bound at once, the 2nd/3rd etc will fail to bind. I bet you could fine old 350QX v1s on ebay, so this wouldn't be impossible but is again not trivial as a school project.

 

[theGreenOrange]

OK I'll bite, I'm not aware of any of the shelf quads which can fly in formation, you have to do it the hard way.

You would probably do best with something like QX90s, they are <$70 each, and have a decent FC, to get them to fly in formation you will need a 3d camera rig and some custom software to track and control each quadcopter from an external computer - this is a pretty big challenge, so unless you have a team of developers to help, I suspect will be the sticking point.

To simply fly 3 quads in a roughly similar shape for a few seconds, you can find some cheap quads which will bind multiple quads to a single transmitter, I've seen this done pretty effectively with GPS quads (the early 350QX had a demo video of two flying in sync on a single transmitter), but for cheap non-GPS quads, you are only going to get maybe 30 seconds of flight before they will drift out of sync and it will all get messy, but this would be much cheaper as potentially you could get 3 quads for under $100 total, and you could do a quick flight with some flips etc with 3 in sync with each other.

Just found the 350QX flight James Haley flew this - this is no longer a current model, but you could theoretically do this with 3/4/5 350QX v1s, as they used DSM2/22ms which would allow multiple receivers to bind to a single transmitter - anything newer than this (DSMX/DSM2-11ms capable) won't as they have an acknowledged bind sequence I believe, so only one receiver can be bound at once, the 2nd/3rd etc will fail to bind. I bet you could fine old 350QX v1s on ebay, so this wouldn't be impossible but is again not trivial as a school project.

Some good tips if he is serious. However, a cyscan system (with a target reflector on the lead plane) would likely be easier than a 3d camera. I am guessing the drones would need to be be pretty big.

 

[jamesb72]

Pretty sure the Cyscan would break the budget of $400 per quadcopter 3d camera rig is very cheap, you just need 2 or 3 camera's, one on each side of flying area, I'm pretty sure you could get the parts for a system together under $500 all in, but would need a lot of software development to get it all working together, which is the difficult part.

 

[theGreenOrange]

Fair enough, I suppose you are right. The cheapest cyscan might be more than $400.