Feedback sought re; plans for weather balloon + UAV paraglider + higher-capacity multirotor

I'm looking for advice on a few things, it's a long story so I'll start at the beginning:  The Engineering Club at my college wants to launch a high-altitude (AKA weather, sounding) balloon with a payload of cameras, etc. to go up to hopefully ~100,000 ft. and take some video.  A lot of people have done this, so we wanted to try something a bit different.  Usually it's done just with a sealed foam box attached to a parachute, with the top of the parachute attached to a line running to the balloon.  When the balloon bursts from the inevitable air pressure change, the round parachute carries the payload to the ground - to god knows where.  Lately people put a GPS device with radio of some sort - ie, cell phone - to make retrieval easier.  But if the payload ends up landing in an inconvenient place, such as a large body of water, the top of a tree, or if the parachute fails - you've lost your stuff.  Often the payloads do get lost, so people tend to send up cheap cameras and accept the risk.

Since we're both on a budget and want to do something different, we'd like to have a second parachute open at somewhere under 40,000 ft. - a ram-air parachute, with a controller and servos inside the payload able to guide it towards GPS coordinates.  Since some payloads end up in the ocean, etc., any distance closer to a decided location would be helpful.  The guided parachute could potentially be a powered paraglider, if adding a prop doesn't complicate things too much.

We're also planning on building a quadrotor which can be used, among other things, to test the parachute / powered paraglider.  I have those plans pretty much laid out but if you have some suggestions, links I could use them.

Any advice about the UAV paraglider + balloon would be appreciated, thanks

Views: 1909

Reply to This

Replies to This Discussion

It's no secret that plenty are working on return-to-launch HAB payloads, myself included.  The ram-air parachute is an interesting idea and could be used to steer the payload to a more appropriate location, but I suspect it would be difficult to do much more with it.  The glide-ratio of a parachute/inflatable wing of any description is always going to be a little compromised and when you only get to deploy it at lower altitudes, the distance you can travel will be somewhat limited.

If you deploy too far up, you run into the problem of limited penetration speed as you pass through the jetstream.  IMO, you need to be able to put on a good turn of speed through this altitude band to have any chance of making it back home, since headwinds could be 200mph or more.  You don't want to spend long there, and when you are there, you want to minimise the headwind effect as much as possible.  That equals high speed on both counts.

As for powered return-to-launch, given that you have to carry the energy requried throughout most of your mission, it might be difficult to carry more than 30-40mins of powered flight endurance.  At low, model-aircraft-type airspeeds, that doesn't represent much distance.

I'd encourage anyone who's considering anything like this to get the pencil and calculator out and do a few simple fag-packet sums.  A little engineering goes a long way in this area, although many appear to be reluctant to do it.  I'd also recommend X-plane (or similar) simulations, since you can get a very good idea of what logisitcs are needed for very little outlay.

Good luck with your attempt and keep us all posted!



Good points, thank you.. yeah it sounds like powered flight is probably not much of an option.  I have found some other similar projects, one pretty much identical, but just plans.  Otherwise there's been guided payload recovery research / applications by NASA, Lockheed, etc. that are good for inspiration.  The flexible wing hang glider type ones look interesting too.  

I should add; since we're in Seattle the polar jet streams definitely a consideration, but I think that at least makes it predictable (they seem to always blow due east).  The plan so far is to launch from a time and place that we'll have a good guess what region the wind could send it to, then have the guidance system try and steer it towards an acceptable landing area.  One idea is to launch it from a mountain in the Cascades and hope that we can get it to land within one of the national parks east of there.  Just getting it to an accessible place is the first goal.  Probably we'd give it a selection of acceptable regions and it steer itself towards the closest / most reachable.

I wonder what the ideal type of UAV glider would be for dropping from 100,000 feet.. it seems like when you have that kind of altitude, you should do something with it - other than just fall.

I think the FAA might step forward shortly if this becomes a really popular activity.

Remember it's a UAS if it can be controlled in three axis, a balloon, or free falling parachute falls outside of the UAS remit.

I think it's a popular activity mostly in theory at this stage Gary.  If the FAA starts to regulate onerously, most of this type of development will happen elsewhere in the world.  I'm sure the FAA recognizes that they need to define how things are to be done, rather than forbid that things be done. 

The sky wasn't made to be colonised by commercial interest groups whose interests are defended by taxpayers...

Hi there, just reading the post and found it very interesting.

In regards to your comment relating to the FAA ( or what here in Australia is referred to as Civil Aviation Authority ), we are in a difficult situation as far as FPV goes here in Australia.

From what I understand, FPV systems ( by law ) are to have a MAXIMUM output of only 100mw.

and its now against the law to fly out-of-sight when using any model aircraft regardless of type.

If I have this wrong ( an  hope I do ) could someone  that does FPV here in Australia ( especially NSW ) please correct me. - it seems that even "hobbies" are fast becoming the next target for big brother. - Some times I wish paranoia would be blended with just a little common sense.

Cheers all.

Interesting.  Well, I'm not in Australia and it's not going to be powered, so I'm not concerned about that.  As for the FAA and what not - this is being done through the college, so we'll be sure to follow all the laws, which should't be difficult.  The balloon will be covered in high-vis fabric for the small chance of an aircraft encounter.  

We'll start testing the parachute soon, that should be interesting.  We'll be using a bunch of small helium balloons tethered at a few hundred feet, with a line & pulley to hoist the parachute up to.  I'll put some videos up of that then.

Well, the budget has been approved and the non-powered UAV paraglider will be the project.  What I'm trying to figure out right now is exactly what we'll need.  It will need to be able to do the following:

- Trigger servos in different situations according to input from: 6DOF data, magnetometer, GPS, barometer, airspeed sensor?  -- 2 servos to pull brake toggles, 1 to shift it's weight R/L, 1 to cut away 1st chute and deploy second.
- Make decisions on how to use the above according to a map of good vs bad landing zones

So the parts list I'm working on includes so far:

- Ardupilot 2.0 (overkill? it seems like a good value)
- Four ~5V 180 deg. servos
-  an external airspeed sensor?  (I haven't been able to locate one but I assume you must need something that actually measures the relative wind.)
- a reliable ~5V lithium battery & charger

Besides all that, the payload will include a Contour HD camera, a phone with GPS tracker for retrieval, two parachutes and the necessary hardware to connect to the servos and cut away the 1st / deploy the 2nd at a specific altitude.  Its body will be made of foamcore wrapped in insulating foil.  Then of course above it will be a balloon.  The line will have a round parachute on it, and packed into the top of the body will be the steerable ram-air chute that deploys when the round chute is cut away.

The plan is to begin by throwing it off a building and seeing if it can find a target, then using a tethered helium balloon with pulleys to repeatedly hoist it up and drop it to test it's parachute deployment and navigation skills.  Then we will use larger balloons to release it from progressively higher altitudes.  When it repeatedly navigates well from ~400 ft. drops, we'll see how it handles thousands, tens of thousands of feet.

For the software, we will give it polygons indicating safe and unsafe landing zones on the ground.  The plan is to predict the general area it might land in, and map as much as possible using satellite imagery & maps showing where good (parks, fields, etc) and bad (rivers, freeways, airports, etc.) landing zones are.  Hopefully we can make it avoid the bad places and steer towards the good using weight-shift and brakes, like your typical parachutist.

Any advice would be appreciated!  especially about the parts.  I'm a bit confused.


Yeah I thought of that too, but I think the potential problems with fixed-wing flight through the jet stream make it a bad idea.  The chances of it going into a spin it can't recover from seem a bit high, plus the risk of high wind forces breaking it to pieces.  

In the proposal they're having me write for the project I'm having to focus on the maneuvering abilities of the parachute-UAV, so I don't think I'm actually going to be sending it up to high altitudes any time soon.  The first tests will be from ~200', then 400', 600', and progressively higher.  I'll use a smaller sounding balloon tethered to the ground, repeatedly lift the UAV up with a pulley and drop it.

I ordered the Ardupilot 2.0, but while I'm waiting for it to arrive I'm going to throw together a simpler UAV using the Arduino + accelerometers from a Nintendo Wii nunchuk, just to experiment with.

Does anyone know a good resource for using the Ardupilot for non-RTF Aircraft?  All I've found are instructions for setting it up for use in the Easystar, Quadrotors, etc.  I basically just need to know how to work with data from the sensors to control servos.  

have you worked out what paraglider design you wish to use. I am a paragliding pilot and the biggest problem we have is cloud suck. and for your deployment idea I would strongly recommend a D bag system. the largest problem is line overs and twist. once the wing is up and going pitch control will be a problem and getting the software to work that out. also as a point our glide ratio on my mistral 6. is only 1:9. best paraglider I have seen is 1:12. I know hang gliders go a lot faster only need weight shift and have better glide ratio. sorry for the rant just my 2 cents.

Reply to Discussion


© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service