Last week we started testing our release system. The aim of this first manned flight was to check the behaviour of the mechanism and ensure that the basic software functionalities worked as expected.

Launching a stratospheric balloon is rather expensive and requires extensive logistics. Then how to get experimental feedback on our drone’s behaviour? Although that took a leap of faith at first, we decided to launch our paraglider drone… from a real paraglider! Our crazy-but-experienced pilot took on this challenge to drop the drone from above the drone’s pilot, and it proved effective! We’re now able to iterate quickly over design enhancements. Here’s what it looks like on video:

As a key takeaway, deploying a wing properly is harder than it sounds. There is still much to do in order to make this mechanism reliable. In the test above for instance, a cravatte occurred and degraded the wing’s behaviour. Software issues also caused problematic wing stalls, but no crashes to report: the wing opened up and behaved perfectly from 20 meters above the ground!


<
Previous Post
Designing and building a release system to launch our drone from the stratosphere!
>
Next Post
Improving robustness with sensor calibration on our Navio-based architecture