Making aerial heat map - What can possibly go wrong?

This week-end I had a wonderful idea! Why not taking some aerial thermal pictures of my house? I could possibly detect heat lost from the roof of my house or garage and I could see my land in a different perspective!

So I started this project. The bebop drone I is supposed to be able to lift 200g. My Smartphone weights 145g and the FLIR One weights 110g so I am slightly overweighed, but not too much, so I decided to give it a try anyway on a very calm day so that it doesn’t have to fight against the wind.

Because the bebop has a few sensors under it, I had to hook the equipment on its top. This reduced a little bit its stability but prevented interferences with its sensors. Although GPS was a little bit hide by the Smartphone…

The take-off was very smooth and the video from the FLIR One was relatively stable. It doesn’t seem to be subject to motion blur from the vibrations or to rolling shutter effect.

When I reached about 100 ft altitude, the fairly tale suddenly turned into a nightmare! I lost the reception with the drone and I was unable to reconnect to it. I stayed stable for a few minutes in air until the batteries depleted and then it was the crash on my roof followed by another crash on the concrete!

Result: a moderate crack on my Smartphone screen but still working great, the FLIR one casing was broken but easily glue back together (Camera is fine) but the drone camera was damaged pretty bad! (Hopefully the video is still usable when looking everywhere except on top, but I will check if I can replace the lens eventually using a door viewer.)

The black box crash analysis concluded that the lost of reception was caused by my Smartphone attached to the drone. I left the WIFI at ON and I guess it attempted to connected to the drone, preventing my controlling Smartphone from the ground to reach correctly the drone.

Extracting the most revealing video frames allowed me to create a 5mp heat map picture of my house which I am pretty happy of.

I think that the experiment was not that risky it wasn’t that I forgot the WIFI at ON, the landing would probably been successful, however, seeing one grant crashing like this on the concrete might not be recommended for people with hearth-risk diseases!