Feasibility Check - GPS + Gyro + Servo's

I'd like to fund a project that brings a payload back to its targeted GPS coordinates. I'd first like to know if this is feasible using the Arduino platform.

I want to program a GPS coordinate into my 'payload' then drop it from an altitude and have it return to that spot.

Assumptions: The drop altitude is high enough for the payload to reach its destination The only 'acceleration' will be from gravity The Arduino will control the "Servo's" which will be connected to some sort of wing'd structure A side effect of this design will make the 'decent' stable ie. reducing spin of the payload (i think)

Project Outline: Everyone has seen the Weather Balloon camera's. I'd just like to design a payload that returns from the launch point rather than tracking it down across 150-200 miles. I realize there are other variables here, but lets just say for testing purposes it will be doped from an airplane at an altitude of 5k feet. (I've got a friend that will drop it for us)

Below is a good resource:

http://diydrones.com/profiles/blogs/ardupilot-main-page

Excellent resource! Thanks!

The space shuttle is/was a glider. So, with planning, it is possible to come from any altitude down to a fixed location. Balloon pilots seem to be able to plan their flights with a surprising degree of accuracy. If you have flexibility in where you drop your payload, in other words you take account of wind speed/direction at different altitudes then you should be able to reach a target (in my completely unqualified opinion). You need to specify an accuracy though, and you should talk with people who paraglide rather than (or at least as well as) computer geeks.

radman: The space shuttle is/was a glider. So, with planning, it is possible to come from any altitude down to a fixed location. Balloon pilots seem to be able to plan their flights with a surprising degree of accuracy. If you have flexibility in where you drop your payload, in other words you take account of wind speed/direction at different altitudes then you should be able to reach a target (in my completely unqualified opinion). You need to specify an accuracy though, and you should talk with people who paraglide rather than (or at least as well as) computer geeks.

While you're right about all of that, the space shuttle starts from 250miles above the surface. At 20-22miles my payload wont have nearly as much distance to find it's target.

Also while i am planning a glider, this will more of a ballistic trajectory.

If only i could find a ballistic forum :)

That sounds entirely feasible. But if you're envisaging this as a 'bomblet' steered by GPS+compass, you need to make sure the thing is inherently stable. If the thing's spinning, you'd never be able to work out which way to steer it. A classic bomb shape with some steerable fins at the back, GPS to work out where it is and where it wants to go to, and a solid state compass to work out when it's pointing in the direction it wants to go, would sound like a viable solution.

How much ground do you expect this thing to need to cover in the 'glide' phase? If you're going to taking a balloon through the upper atmosphere I suppose the wind speed is going to be very high.

I'd just like to design a payload that returns from the launch point rather than tracking it down across 150-200 miles.

Don't think that is going to happen. The typical hobby balloon will not be able to carry a gliding object heavy enough to make upwind penetration (balloons generally go down wind) over that distance. The rogallo wing was used in early guided cargo drops, which might be useful for testing in local overhead drops.

http://www.google.com/search?hl=en&num=100&lr=&ft=i&cr=&safe=images&q=rogallo%20wing&spell=1&sa=X

zoomkat: Don't think that is going to happen. The typical hobby balloon will not be able to carry a gliding object heavy enough to make upwind penetration (balloons generally go down wind) over that distance. The rogallo wing was used in early guided cargo drops, which might be useful for testing in local overhead drops.

Thanks for that link, not what i'm wanting to do.

PeterH: That sounds entirely feasible. But if you're envisaging this as a 'bomblet' steered by GPS+compass, you need to make sure the thing is inherently stable. If the thing's spinning, you'd never be able to work out which way to steer it. A classic bomb shape with some steerable fins at the back, GPS to work out where it is and where it wants to go to, and a solid state compass to work out when it's pointing in the direction it wants to go, would sound like a viable solution.

How much ground do you expect this thing to need to cover in the 'glide' phase? If you're going to taking a balloon through the upper atmosphere I suppose the wind speed is going to be very high.

Would the "XY and Z" optical sensors manage that? Spin would be controlled via the fins like on a rocket, so long as there is enough surface area, they should be able to compensate for spin, right? It looks like the AdruPilot mega has everything in it. Just need to retro fit it for 'bomblet' steering.

Threw research I've found that the rig will drift anywhere from 10 to 200 miles from it's launch point. There are websites that will predict your landing area based off wind speed measurements. Sure, if i were to hit the jet stream i'd never get it back. However if i launched only on days that my drift was within tolerances for getting it back I should be golden.

What i'm wanting to accomplish is launching the balloon, and making retrieval trivial. I'm also looking into options that will let me reuse the balloon. That will bring the cost to fly missions to around $20each.

I think you need to plot out the sort of trajectory you expect this to take on the ascent and descent, and what sort of glide slope you need to achieve. It the existing recovery brings you 200 miles down range from a peak altitude of 20 miles than naively you have to average a glide slope of 1:10 which is not easy. In practice you will probably need to achieve at least double to account for the thin air and give you a margin for control. That’s possible, but needs some pretty clever aerodynamics. Essentially, you’d end up with an autonomous model glider. When you look at the numbers you may find even that performance isn’t enough. It also needs to be big/heavy enough to have an air speed significantly in excess of the windspeed. This is starting to look like a pretty heavy payload. What’s your weight budget?

PeterH: I think you need to plot out the sort of trajectory you expect this to take on the ascent and descent, and what sort of glide slope you need to achieve. It the existing recovery brings you 200 miles down range from a peak altitude of 20 miles than naively you have to average a glide slope of 1:10 which is not easy. In practice you will probably need to achieve at least double to account for the thin air and give you a margin for control. That's possible, but needs some pretty clever aerodynamics. Essentially, you'd end up with an autonomous model glider. When you look at the numbers you may find even that performance isn't enough. It also needs to be big/heavy enough to have an air speed significantly in excess of the windspeed. This is starting to look like a pretty heavy payload. What's your weight budget?

To directly answer your question, my max payload is 4lb.

The rest of the answer The project is extremely flexible. If for example my glide slope is 1:1 i can simply wait to launch my balloon until the weather conditions meet that requirement. Although I think your right about all that, maybe its time to just go with AdruPilot and be done.

This sounds like a cool project. It probably wouldnt even be that hard to code it correctly for the steering system.

The problem I see is this glider will have to either be very stable, or some way is made to adjust the pitch to maintain a proper glide airspeed. Ask any pilot about the importance of maintaining a proper airspeed while gliding. Without pitch control, you simply will not be able to get the optimum glide distance over a range of altitude, weights, and turbulence.

To incorporate this, you will need some form of airspeed sensor. Im not sure what the best option for this project would be.

If the glide slope requirements are looking too challenging, maybe you could think of this more like a homing pigeon. If you know where you want it to come down, and launch it in roughly the right place so that the winds will bring it into the right general area, just fine-tuning the landing zone by a GPS-guided steering system would be much easier than trying to get the thing to glide 100 miles upwind.

PeterH: If the glide slope requirements are looking too challenging, maybe you could think of this more like a homing pigeon. If you know where you want it to come down, and launch it in roughly the right place so that the winds will bring it into the right general area, just fine-tuning the landing zone by a GPS-guided steering system would be much easier than trying to get the thing to glide 100 miles upwind.

Yes! This is exactly what i want to do... ie. choosing a launch day that would result in a predictable landing site.

I am a bit confused about whether this thing is a balloon, a glider, a bomblet or something dropped from a balloon. I think you are trying to launch a balloon to an altitude of 20miles then get it to return reasonably accurately near where it was launched. How does the balloon turn into a glider/bomblet ?

radman: I am a bit confused about whether this thing is a balloon, a glider, a bomblet or something dropped from a balloon. I think you are trying to launch a balloon to an altitude of 20miles then get it to return reasonably accurately near where it was launched. How does the balloon turn into a glider/bomblet ?

The Balloon is attached to a "payload" the "payload" is the "glider / bomblet or other shape which can be considered a steering mechanism"

You said you wanted to re-use the balloon as well. So how is the balloon deflated and then carried by the glider/bomblet ?

I've devised a way to deflate the balloon via a radio channel at any given altitude. It's light weight and very simple so it's not a huge weight gain.

I've also found the solution to the problem, AdruPilot. As well as possibly full control during the entire flight using Dragon Link & a 1.2ghz camera system. I'll start prototype testing this spring.

You thread is titled feasibility check. If the intention is just to deflate the balloon and leave it trailing I would have quite big concerns about how it would affect any glider it was attached to. On a bomblet maybe it would just act as drag and not really affect any guidance.