Smart Cities - BirdNET-Pi Project

Overview

• BirdNET-Pi uses a USB microphone connected to a Raspberry Pi to record and identify bird calls.
• The system runs 24/7 and can upload data in real time to https://app.birdweather.com
• Data can also be downloaded via an API so that you can do some additional data processing using Node-RED.

Learning Objectives

• Learn about the important role of vegetation in supporting small birds
• Learn how to install BirdNET-Pi on a Raspberry Pi 4
• Learn how to interpret data from a local installation of BirdNET-Pi
• Learn how to use an API to retrieve data from the BirdNET-Pi installed in Bundoora

Noisy Minors and Red Wattlebirds

• Noisy Minors are native honey eaters.
• They have done very well in urban areas because of the abundance of eucalyptus trees and mown lawns.
• Noisy Minors are also territorial and work cooperatively to drive our smaller birds.
• In many parks small birds are locally extinct and only survive close to waterways.

Small Birds

• Small birds need protection from thick scrubby vegetation so they can build nests and also hide from larger birds.
• Examples of small birds include:
  • White-browed Scrubwrens
  • Brown Thornbills
  • Spotted Pardalotes
  • Willy Wagtails
  • Yellow Robins
  • Golden Whistlers
  • Eastern Spinebills
  • Grey Fantails
• Government agencies, conservation groups and schools are working together to plant more shrubs and bushy vegetation to encourage small birds to come back to our local parks.

• You can read more in this article.

https://www.portplaces.com/noisy-miners-and-the-plight-of-small-birds/

• There is also a documentary you can watch on ABC iView called The Secret Lives of our Urban Birds. Start at the 41 minute mark to see the section on Noisy Minors.

https://www.abc.net.au/catalyst/the-secret-lives-of-our-urban-birds/13734884

• Friends of Darebin Creek - Small Woodland Birds

http://friendsofdarebincreek.org.au/wp-content/uploads/Small-Woodland-Birds-V2.pdf

Plants for Small Birds

• There are some parks in Melbourne that have high number of small birds and small numbers of Noisy Minors.
• Trin Warren Tam-boore in Royal Park is home to 153 bird species.
• Local waterbodies are surrounded by an impenetrable layer of Tangled lignum (Muehlenbeckia florulenta) which is a place of safety for small birds.
• Jawbone Reserve in Williamstown is also home to many small birds.
• The park is full of Sheoaks and Tangled Lignum.
• Other plants that are recommended are Silver Wattle (Acacia dealbata) and Black Wattle (Acacia mearnsii)

Hardware requirements

• Raspberry Pi 4.
• USB Microphone - (e.g. Gyvazla USB Microphone Lavalier Clip-on Omnidirectional Condenser Microphone for Computer)

https://www.amazon.com.au/Microphone-Omnidirectional-Condenser-Interviews-Recording/dp/B072Q2GH99/ref=sr_1_5?keywords=Gyvazla+USB+Microphone&qid=1655199739&sr=8-5

• Heat sink and fan for Raspberry Pi. Optional, but highly recommended because it will keep the processor temperature less than 40degC and prolong the life of your Raspberry Pi.

Full Installation instructions

• Full installation instructions can be found on the Core Electronics web site - with GUI interface.

https://core-electronics.com.au/projects/bird-calls-raspberry-pi/

• Installation instructions are also available from the creator of BirdNET Pi (Patrick McGuire) - Lite version (no GUI interface)

https://github.com/mcguirepr89/BirdNET-Pi/wiki/Installation-Guide

Bird NET Pi - Adam's system

• Most Bird NET Pi systems are available through Bird Weather
• Adam has also made his Bird NET Pi installation directly available on the internet Online example of Raspberry Pi Bird NET Pi

Imaging Raspberry Pi Operating System

• Install and open the Raspberry Pi imaging software on your computer.
• Connect a 16GB Micro SD card to you computer

• Select Operating System.

• Select the Raspberry Pi OS (64-bit) version.
• Select the full 64 bit version, not Lite.

• Select Choose Storage.

• Select Mass Storage Device 16GB.
• You can install the OS on larger memory cards. 16GB is the smallest you can use.

• Click on Write.

• Select Yes - to write the image to the micro SD card.

• The installation takes approximately 15 minutes.
• The writing process is followed by a verification.

Initial Operating System setup

• Once the installation of the of the 64-bit OS is complete insert the card in a Raspberry Pi and complete the OS initialisation.
• During the initial setup you will be asked to:
  • create a username and password. Create user pi and generate a difficult password. You will be cautioned about using the username pi but use the pi username to avoid software update issues.
  • set up localisation and language options
  • provide local WiFi connection details
  • upgrade the Operating system to the latest version
  • restart your computer

Adding a temperature sensor to the top panel

• It is good practise to monitor the temperature of the Raspberry Pi CPU.
• Right Mouse button click in the Top Menu Bar.
• Select Add / Removal Panel items

• Click on the Add button on the right of the active window.

• Select CPU Temperature Monitor and click on Add
• Then Close the Panel

• In the top menu you should no see the process temperature being displayed.
• A temperature less than 40degC is normal.
• Above 50degC is too hot, but will not harm the Raspberry Pi in the short term.

Microphone installation for testing

• Plug the microphone into the USB 3.0 port.
• The USB 3.0 plug is coloured, either blue or yellow.

Microphone final installation

• The microphone needs to be installed on the exterior of the house to receive bird calls.
• Open a window and place the microphone in a sheltered location under and eave so that the microphone doesn't get wet.
• Use sticky tape to attach the microphone to the house.
• Close the window carefully so that the cable connected to the microphone is not damaged.

Installing the BirdNET-Pi app

• Open the Terminal
• Enter the following code curl -s https://raw.githubusercontent.com/mcguirepr89/BirdNET-Pi/main/newinstaller.sh | bash
• The code was obtained from this site https://birdnetpi.com
• If prompted, answer Y (yes) to any questions.
• The installation will take approximately 15 minutes.
• The Raspberry Pi will reboot once the installation is complete.

Demonstrating BirdNET-Pi

• Open up a Web Browser on the Raspberry Pi
• Enter the address http://birdnetpi.local in the search bar.

• This will open up the BirdNET-Pi Home Page.

Checking microphone settings

• In the main menu select Tools
• You will be asked to enter a username and password.
  • The username is birdnet
  • The password is empty. You can change this later.

• Click on Web Terminal

• Enter alsamixer -m to enter the ALSA sound mixer.
• The -m option permits "mouse" mode.

Swap file expansion

[Expanding the swap file on the Raspberry Pi]

BirdNET-Pi API

• BirdWeather has an API in place and the BirdWeather (web) App makes use of it in producing the results seen via the web page portal.
• Documentation is here - [1] https://app.birdweather.com/api/index.html
• There is also an online query tool for testing queries here - https://app.birdweather.com/graphiql

Example Python code

• This is a simple python program that will return some BirdWeather results from station 273 (my station in Bundoora, Victoria Australia)

• The output is shown in the Python Shell.
• The data is in JSON format

• Copy this output and paste it in JSONviewer to make the data human readable https://jsonviewer.com
• The data should be similar to the screen shot below.

BirdWeather online GraphiQL tool

• An online query tool for testing queries at this URL - https://app.birdweather.com/graphiql
• An example output is presented for station 273.

• Use the online documentation to modify your query.
• This will help you understand how the API work.
• In this example below the wikipediaSummary has been added to the query.