Wild Find

Main screen 3D Plot Signals Histogram

A software suite designed to track and map the locations of VHF transmitters which are typically used to locate wildlife in ecological studies.

It uses RTLSDR compatible USB dongles to receive the transmissions and any GPS unit which provides NMEA data.

Please note this software is at the Beta stage and currently undergoing testing.

Update: This project is no longer actively developed as I've been unable to make progress with the signal demodulation.

Installation

The following dependencies are required:

Optionally the following can be installed:

  • natgrid - Add the option to create heatmaps with nearest neighbour interpolation
  • SIP - Adds 3D plotting

Usage

Wild Find consists of two components: Harrier and Falconer.

Harrier is used to survey for transmissions and can be run on single board computers such as the Raspberry Pi 2.

Falconer is a desktop application which uses the data from Harrier to statistically map the locations of the transmitters.

These can be run independently of each other; the data collected by Harrier can be opened by Falconer at a later date, or Falconer can directly query and manage Harrier over a network link.

Harrier

Harrier is a command line application, most of the set up is contained in a configuration file (and example is included the the `conf` directory).  The configuration file should be updated and copied to your home directory unless you explicitly specify it's location.

Harrier can be started by specifying a centre frequency to scan around. For example to scan for collars near 150MHz:

harrier.py -f 150

This will write to a file called `harrier.wfh` which can be opened with the Falconer application.  If this file already exists survey data will be appended to it.

Pressing [CTRL][C] will exit the application.

Falconer

Falconer is used to map the data from Harrier.

From Falconer one or more surveys can be selected (a survey is generated each time Harrier is run), as well as individual scans (a scan is a sweep of frequencies at a particular time).  Finally the detected signals can be filtered (to include particular collars or exclude erroneous transmissions).

Start Falconer by using the command:

falconer.py

 

Categories: 

Comments

7

Click to view comments

Hi Al,

I am a wildlife biologist and I think I could use something like Wild Find. It sounds like it is fairly close to what I need. I have been thinking about using SDR with GPS for a while now to pinpoint VHF radio collars. It would be ideal, however to have a system that could use the results from Wild Find to autonomously navigate a 3DR Solo UAV, a drone with a Pixhawk autopilot, to the collar where it would hover for a period of time. I think this would require a onboard companion computer. I am somewhat knowledgeable of drone mission planning, I have made heat maps in ArcMaps, and I have used SDR# (SDR Sharp), but I am very new to the rest (i.e., companion computers, Python scripts, etc.). I just ordered my first Raspberry Pi yesterday - That should give you an idea of how new all this is for me. That said, I am very enthusiastic about making a system like this and would very much appreciated your input; especially on the drone navigation part. Is this something you might be able to help me with?
Thanks,
Kevin
----------

Al's picture

Hello,

Thanks very much for the offer.

At the moment I'm trying to improve the sensitivity of the system as signal noise can cause quite a problem.
I've taken some tentative steps towards implementing a basic machine learning approach, although I wonder if this may be too much for a single board computer attached to a drone (the ultimate goal).

I'm always looking for testers, can I contact when I've made a bit more progress?

Al

Hi Al,

I am a wildlife conservationist and have been looking for a system such as Wild Find for the past few months. I am just getting into coding using microprocessors (Arduino) and the Raspberry Pi but progress is slow due to a hectic schedule.

WOuld it be possible for you to send me the system images for Wild Find running on Raspberry Pi 3? It would enable me to simply image my Micro-SD card and run the system that way.

I have the following core system components,
2x Raspberry Pi3
1x RTL-SDR
1x Garmin USB GPS (Puck)

Happy to make a donation for your time and assistance to help me with this.

Best regards,

Kev M

Al's picture

Unfortunately I haven't really got anywhere with Wild Find for quite sometime. There are problems with detecting the signals from collars which I think will only be solved by using much more computational power, this is too much for a Raspberry Pi.

I'll update the readme as I don't think I can make any-more progress without getting more developers onboard.

Sorry to disappoint you.

Hi Al,

That is a shame, I do think that a system such as yours would be of great use. We do have some very basic and low-cost VHF beacons that act more like a pinger and can be received using any analog radio or scanner. I believe that I could still use your system for this so at least your efforts will have an impact on our conservation efforts. If you do have a Raspberry Pi image for the SD card that you could share then I'd still like to have a go.

Best regards,

Kev M

Al's picture

Sorry I don't have one, it run slow enough on a desktop computer and would be close to useless even on the latest Raspberry Pi.

Hi Al,

Thanks for the heads up, shame that it runs slow. Great concept though and something to keep in mind.

Keep up the good work!

Click to add a comment