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

Join diydrones


  • My pleasure!

    But if I were you, I'd consider using the analogue video out on the Raspberry to feed a 1.2 Ghz link. That would 10x your range for less than $100. And still you have the pi there to do all the other work. 

    Or do both! Redundant image transfer...

    Keep it up! :)

  • Hi Magnus,

    Thanks for the Info! 

    Although FPV is a great solution, it's application specific.

    We do prefer a general purpose platform, with embedded Linux, that is flexible and it can run whatever code we want (despite it may underperform application-specific platforms). Anyway, very good hints!

  • Hi Domenico,

    1. in case you have missed it, there is a whole wonderful new world called "FPV". True, different countries have different frequences, but manufacturers have equipment to suit your needs. I'm not sure, but I believe 900Mhz and 2.4 as well as 5.8 is OK in the US. You'll find that on google very fast. But you'll probably need a HAM license to operate the better transmitters.

    With for instance a $80 transmitter you'll get pretty good quality video and audio transmission of 500-1000 meters with stock antennas. Get better antennas, or even better several ones with a diversity reciever and you'll get several kilometers of range. Get an antenna tracker and you'll find examples of people who have flown about 10km with video link. is one place to start to look for equipment, but there are many others too.

    Your idea could also be used in conjunction with this, ie using a quad as a relay station in the air for even better reception, and to get above the trees for best possible reception.

    Very cool project though!

  • Thanks guys!

    @Magnus. Your comments are great.

    1. Using Wi-Fi 802.11n we have distance of up to 300 meters with two hops. However, tests were done with embedded antennas, and we are now testing several types of external antennas. On the contrary of WiFi radio, 1.2 Ghz video may not be used worldwide (e.g. in the US). However, I need to admit I'm not very familiar with the products in this band. Do you have any product you would suggest to have a look and any material for the specifications?

    2. your comment makes perfectly sense. However, think about a scenario in the mountain or wherever there is no cellular coverage (there are always those spots, also because of disasters). having a hybrid network built around the target area will allow to find the missing person, even if he is in a forest and image/video processing may not detect him. Here, a smartphone App would be very helpful.

    3. it is always needed some rough idea of where to start to look at. for instance last cell tower that the person visited.

    true, the concept can be improved, it's a research project, that's why comments are more than welcome :)

    the work is conducted at ETH zurich, in the group of Prof. Plattner.

    you can find here a short description of the setup, and more information in the related work of that paper.

  • oh ya 

    are working with Raffaello D'andrea

  • congratulations !!!

    would you be able to post details of the setup/software


  • That's very cool! I hardly dare ask since I don't want sound negative, but some basic questions:

    1. What kind of distance are we talking about for each quad/hop? In the video it seemed to be less than what a standard 1.2Ghz video link would deliver? With good antennas and a tracker you would get 10's of kilometers instead of a few hundred meters as shown in the vid?

    2. This might be beside the scope of your concept, but the person to be found has to have an Android phone with an app running? What are the odds of that? And if they do have a Smartphone with them, can't they just call 911 and start google maps or something and tell their coordinates? :) Any plans on alternative ways of finding people? Thermal imaging etc? That would be awesome.

    3. What would be the practical range of finding lost people? I know that quads that size with some equipment on them typically fly about 15-20 minutes on a charge. That's round trip. I would think planes would be a better idea, to be able to cover larger areas? 

    Again, sorry for sounding like I'm raining on your parade, I just wanted to check if there was more to this story. Very cool stuff though, very useable parts but I feel the concept as a whole can be improved. 

    Kind Regards

    Magnus, Sweden

  • Interesting work. I have had a similar idea with the intermediary step being a ground based rover which the quad would launch from once the ground station reached its near maximum range.
This reply was deleted.