Fixed wing drone for thermal mapping of a large area

Hi there

I am a researcher within the field of Arctic science. I am in need of a fixed-wing drone solution for a project in the high Arctic involving surveys of surface temperature over sea ice. I am hoping people within this forum will be able to give me some pointers and leads on how to achieve the right setup.

The drone should be able to follow waypoints and automatically take bird's eye view pictures with a thermal camera. The end result should be a stitched map of surface temperatures. The area in question is somewhere along the lines of 4x4 - 6x6 miles in size. The size of the area and the low temperatures at the site (often -30 degC / -22 degF) is the reason I am looking for a fixed wing solution although I am open minded to suggestions.

Seeing as I have little experience with drone operations I would prefer a setup that is both user friendly and fairly straight forward to assemble. I have actually looked at the Aero-M (on sale through this site) which looks like it meets all requirements, except that it does not come with a thermal camera. Would it be possible to custom build a thermal camera such as the FLIR Tau2 onto the air-plane do you think? I am no stranger to coding, and have even touched base with some simple arduino coding in the past. I imagine that might come in handy in this context.

Thank you all in advance!

Cheers

Jakob 

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –

Replies

  • Do they make carb and pitot heat for gas RC?
  • LIPO batteries don't like to be that cold. You might consider a gas powered aircraft.
    • If you were interested in the Aero, Event38 makes the same plane but improved in many ways. They will customize for you. Or you can hire one of the community members to build it for you. It wouldn't be hard to determine one's skill level based on their post history here, on RCGroups, Github, and the APM Forums.
      • Event38 looks like a very promising option. I have written them to see if they will be able to help. Thank you very much for the suggestion. Concerning the option to hire a community member to build it: do I just write specific need/budget/concept etc here in the forum and hope that someone wants to get involved or how do people usually do this? 

        Thanks!

  • Jakob,

    How spatially accurate do you need your surface temp readings to be? The reason I ask is that I think you will have a hard time stitching together thermal images from a drone especially over water. Stitching is usually done visually using matching points between two images. The images are pre-aligned using GPS but the final placement is visual, and I think you may not get enough distinct features in thermal images over open water to align images together.

    A plane like the Aero can probably handle that type of mapping mission, assuming you're willing to do a few flights or more depending on GSD requirements. The Aero and similar aircraft fly relatively slowly though, so if there's going to be strong consistent winds then you may need a heavier, faster drone to make any progress. Are you launching from ground or from a ship? Ships add a whole extra dimension of difficulty unless you're at rest in dead calm waters.

    • Hi there

      Thanks for your response. 
      I think you bring up a valid point. There has to be some distinct features in the thermal images to produce a proper stiching. I learned that the hard way trying to stich together a series of pictures which had only faint clouds in them.

      I will be flying over sea ice and thus expect fairly uniform temperature readings unless traversing a crack in the ice in which case it should be much warmer than the surrounding area. The key to my study is (1) stitching together a surface temperature map if the temperature field is heterogeneous enough to allow this, and (2) detecting the location and temperature deviation of any local features such a crack or hole.

      I guess one way to approach this task would be to assume a uniform temperature field based on an average temperature reading for cases where no features are detected and only attempt stitching pictures together when a certain level of heteorogeneity is present in the temperature field. Secondly, if a single feature is detected I can probably just match any such picture with a GPS reading and get an approximate location while assuming an average temperature distribution every where else in the field of interest. 

      I talked with the developers of the Aero-M and they said that their system did not support thermal cameras and that they did not do custom builds like the one I am looking for. Actually they suggested I write in this forum and see what comes up so can you (or anyone else) recommend another type of drone that I might use? I have been talking with a Danish distributor about the use of an DJI Inspire 1 with a thermal camera but I am a little worried about the limited flighttime given the large area of interest and the cold temperatures. Even with the battery warmer and isolating patches that DJI has produced. I still think a fixed wing solution might be best. I am also very open to gas powered aircrafts.

      Thanks in advance!

      Jakob

This reply was deleted.

Activity