Hi Guys,

I´m in involveld in several airship and DIY related projects. For example we´re at the moment trying to fit an Ardupilot in this science vessel: https://www.youtube.com/watch?v=cNne5dP-eeM

Anyway, I wondered a long time why there are not much airship UAV´s out there and figured out that the most likely reason is the lack of easy accessible platforms. This is why we recently published the DIY high performance airship silent_runner http://www.silent-runner.net/index.php/Main_Page 

Now we´ve built quite a few of these ships and gathered some experience to work on the next generation. This might lead to a kit that contains some of the essential components like the envelope, structure parts and basic electronics.

So I believe it´s a good time to see if there are some key features a ship should have, to be useful for the UAV community. At the moment design criterias are:

- small (has to fit in a car)

- efficiency (run half an hour on a 500mAh/7.4V Lipo)

- maneuverability

- 30g of payload (arduino micro, Xbee + some sensors)

Is there interest in such a system and would you add some more critical abilities?

Some outdoor footages here:


Views: 8281

Reply to This

Replies to This Discussion

Hi Andreas,

Very, very interesting...
If it can fly by APM, adapting code of arduino micro, would be great!

I'm really interested about, more precisely, how APM could control vectorized system of engine axis.
Any idea?


I guess the APM might be oversized, as for an airship almost no stabilization routines are necessary. I guess for most outdoor application compass and GPS would be sufficient, connected to a super basic navigation system.

We´re not yet there to work on the autopilot, as we still optimize the platform and the building process itself. But once the platform is fine and can be produced without much efforts, we hope to proceed to autonomous flight.

Hi Andreas

I am working on outdoor autonomous airship.I bought APM for autonomous control and changed some codes but I am not achieve complete autonomous flight.

In my project problem is with yaw control, I am using reversible motor and ESCs but from ardupilot output goes to servos, so there is mismatch in time(Sine motor stop first and than start rotating in other direction) and it not giving me desire result.

can you please suggest me any idea for proper autonomous flight?

Hi Taruna,

sounds like a great project.
Actually outdoor airships should be controlled in yaw and nick by control surfaces such as rudders in the rear of the ship. In normal operation there is always wind flow that has to be overcome by forward thrust of the airship. That way you can even hover over a position using your rudders and control surfaces at low thrust.

See here what I mean with facing the wind: http://youtu.be/pzj_-sWY57o?t=58s

I would propose very normal cross fins that can be controlled from the standard APM setup. In fact using just motors and a cross rudder you can use the airplane code without any changes.

Good luck with the project!

Hello Andreas,

Thankyou for the fast reply. 

Problem is high drag and low speed because of big size, so control surfaces are not enough to provide that much force so I am using yaw motor for yaw control.

so now how to implement with ardupilot(as I specified before output of ardupilot goes to servos and I am using reversible motor so motor has to stop once before it start rotating in other direction and so there is mismatch of time and yaw motor is not responding to yaw control).

Hi Taruna,

I´m pretty sure that control surfaces are the better way to go, and low speed is by the way a huge problem when you go outside.

Anyway, there are two ways to hook up the motors in your specific configuration.

You could patch a second small arduino to the Yaw line of the ardupilot that takes care of providing the PWM signal for your two engines. That way you can make the arduino control all that stuff.

The other way would go deep into the APM code and track the line of the YAW control. I´ve not yet looked that deep into the code but it seems well written and documented.

Hi Andreas,

I hav tried with control surfaces but its not giving enough force to turn 8m long airship thats y we have choosen 3 motor configuration for throttle , lift and yaw respectively and also speed is approximately 2m per second.
We have used plus configuration fo fin I.e 2 vartical and 2 horizontal fin. I will paste the link of picture once I reach to my place as am travelling .

And I went through code throughly. .track yaw controller line.
Since the problem is that APM gives the output to yaw servos which changes the direction immediately but in yaw motor, motor has to stop first in order to rotate in other direction or give force in other direction so mismatching of output to yaw...

I tried to put intial output of yaw servo to zero bfr it gets the output from yaw controller (as motor has to be stoped so output zero will complete this task )

Result of this modification is that yaw servoes is not even responding to yaw controller output...

Can u plz suggest something which can be implemented to slove this problem...

Hi Monroe..

Bimpduino is suitable for indoor airship. It wont work with outdoor airship I think.

Very cool! Thanks for posting this and for the new webpage. I'll be looking for updates to the webpage. This looks very interesting.

This is what i was talking about



Sorry, have been on aerial survey yesterday and no time to reply.

But that´s an awesome one, congratulation to that nice airship platform. Actually it seems like your fins are big enough, I´ve flown airships with smaller fins and pretty good control. Remember that for using the fins you always need enough speed.

I think I understand your problem better now, you´re trying to use the motor in the backtail to be controlled by the Yaw channel? This might works at no/low speed but will stop working once you fly faster as drag effects on the hull start to be stronger than the motor force. You should rely on both, fin and motor, to maintain control.

Anyway, we had the same issue in a cuple of airships and solved it in different ways, here are the ideas:

1. Replace the brushless ESC and motor by a brushed bidirectional motor (RC car or ship). They are cheap but a bit weak and heavy. But these motors don´t have the problem with stopping as the thrust is reversed. We have used that as a yaw motor just wired to the normal servo line. It works great.

2. Put the Yaw motor to the moving part of the fin. That way the servo of the fin rotates fin and motor and creates a vectoriced thrust system. That is pretty elegant and will work on low and high speeds. But you need a reinforced fin.

3. Try to solve it from the programming side. Most good Brushless ESC have several brake options, try to activate them, this often leads to better results, but as the control movements of the APM are rather quick and fast, it will probably not work reliable. The other option is to smooth the yaw output in the APM, but don´t ask me how to exactly do it.

However good luck with your project, would be awesome to see some videos of the test flights. By the way who manufactured the envelope of the ship? Is it tight and reliable?

I was waiting for your reply but no problem at all.

Yes you understood well.We had field trial with this envelope and fins(yaw motor at tail) and yes it was not working good at high speed. So for new airship we are going to use both yaw motor and control surfaces(servos).

But problem is that how I implement this with autonomous . I need to control yaw motor because main objective of project is blimp for surveillance so it may has to hover, moving with low speed or high speed(less chances).

Anyway,I thought about break option available for reversible ESCs but am not sure how to implement it with programming. And what you mean by smoothing the yaw output in APM. I mean am asking about smoothing the output. Is it related to filtering the output of yaw control in APM?Am sorry you said "dont ask" but am still asking.

You can see videos of field trial with this envelope on the following links, but it is not autonomous.



This envelope is manufactured year ago, that time i was not working on this project. So exactly I don't remember but yes it is designed by some researchers and fabricated by mechanic.

Thankyou so much for giving your time. 


We have implemented Thrust vectoring system but it is not solving problem for yaw control.

Reply to Discussion


© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service