BirdNET-Pi

An open-source listening tool that automatically recognises and shares live bird voices in the environment

The BirdNET-Pi is an open-source listening device that recognises bird voices in the environment. These observations can be shared live via communication channels such as Telegram. More than just recognition, this tool invites attentive presence and building a relationship with the feathered visitors in your environment.

De BirdNET-Pi bij Zoöp Creative Coding Utrecht. Beeld: Creative Coding Utrecht

On this page

  • Is it for us?
  • In the footsteps of
  • Step-by-step guide
  • More in-depth

Is it for us?

These questions will help you determine whether this tool is suitable for your organisation, objectives and situation.

What?
The BirdNET-Pi records bird voices in the environment and identifies the corresponding birds. Once a bird sound is recognised, the system records the species, time and location. These observations are stored and made available on BirdWeather, an international biodiversity monitoring platform. Through notifications, they can also be shared directly with groups to encourage awareness and connection.

What for?

  • Stimulate awareness and connection with bird neighbours in Zoop teams or the wider public via notifications
  • Monitoring of bird species present, including change over time
  • Support ecological interventions targeting birds
  • Contribute to citizen science by sharing data with BirdWeather

Where?
The BirdNET-Pi can be placed indoors (near an open window) or outdoors. Outside, it should be well protected from rain.

When and how often?
Once installed, the BirdNET-Pi listens continuously. Sightings are automatically recorded as long as the system is on.

What does it cost in terms of time, money, and materials?

  • Materials:
    • Raspberry Pi (model: 5, 4B, 400, 3B+, 0W2)
    • Laptop with SD card input
    • SD card (16 GB)
    • Microphone (a directional microphone is recommended for precise measurements)
    • A network connection (wired or via WiFi)
  • Costs: depending on which Raspberry Pi model and microphone is used, the hardware needed will cost less than €100. Also consider buying products second-hand.
  • Time for installation: depending on programming skills, several hours for build, installation and connection to BirdWeather and possibly Telegram or other communication platforms.
  • Note: please note that the BirdNET-Pi requires both power (e.g. via a USB charger) and an internet connection (wired or via Wi-Fi). These requirements may limit the choice of a suitable installation location. In practice, the microphone is often mounted outside, with the cable running in through a window, towards the Raspberry Pi.
This is what a bird's voice recording looks like using BirdNET-Pi. Image: https://www.birdweather.com/birdnetpi

In the footsteps of...

More traces

Step-by-step guide

Preparation
Creative Coding Utrecht has designed a comprehensive guide to building, installing and customising the BirdNET-Pi. Here, the steps are briefly summarised:

  1. Hardware delivery: Collect the necessary hardware (see 'Material and costs').
  2. Install: Follow the guide on the BirdNET-Pi website.
  3. Connecting to BirdWeather: After creating an account with BirdWeather, you will receive a BirdWeather ID and can start setting up your BirdNET-Pi station.
  4. Link to a communication platform (optional): Using Apprise, notifications for observed birds can be sent to most common communication platforms.
    • For Telegram, click here;
    • For other platforms such as Microsoft Teams, Slack, Signal or Zulip, here;
    • Tip: set to send only one notification per bird species per day.

Implementation

  1. View recordings: All recordings are documented on BirdWeather. Among other things, the website provides an overview of how often which bird was recognised with which probability and at what times it was recorded. All recordings can also be listened to.
  2. Adjustments: After installation, the BirdNET-Pi can always be moved. It is also possible to adjust the sensitivity or change the listening direction of the microphone.

Follow up
-

BirdNET-Pi analyzes the bird observations. Image: https://www.birdweather.com/birdnetpi

More in-depth

How does it affect the user?
The BirdNET-Pi provides a cognitive entry point to build connection with the living world. In doing so, it encourages attention to the life around us humans, which can have a motivating effect for work as a Zoöp. Setting up a group app where sighted birds are reported can stimulate enthusiasm and curiosity in the team.

Combinations and applications

  • BattyBirdNET-Pi: By using an ultrasonic microphone such as the AudioMoth, the system can also be used for bat monitoring. There is a modified variant of BirdNET-Pi for this purpose.
  • Education: Workshops on biodiversity, citizen science, installation and use of the BirdNET-Pi.
  • Awareness campaigns: QR codes at public installations or the Zoöp with live notifications.
  • Planning interventions: To measure which birds are in the area and potentially track changes.
  • Visualising and analysing data: By linking the BirdNET-Pi to a Prometheus server, for example, and analysis with Grafana OSS.
  • Living Datawalk: If the Living Datawalk is about birds, they can be identified in advance with the BirdNET-Pi to provide context.
  • Microseasons: Through the BirdNET-Pi, changes in the ecosystem can be tracked. As a result, it can provide valuable input for microseasons.
  • Enviro+: Install to collect data on air, temperature and sound environment as context for observations.

Challenges and limitations

  • Recognition is not always 100% accurate: BirdWeather provides scores per recording
  • Raw data requires interpretation to be usable in reports or public presentation
  • Basic knowledge of programming is required for installation and linking

Roles involved

  • Technical facilitator: can program; sets up BirdNET-Pi and connects to BirdWeather and reporting channels.
  • Administrator/researcher: monitors data and feeds insights back to team or audience

Inspiration
The BirdNET-Pi initiative stems from a research project by the Cornell Lab of Ornithology and Chemnitz University of Technology. Their goal is to support both experts and citizen scientists in monitoring and protecting birds. CCU applied it for the first time as a zoonomic tool.

Contact
For Zoöps: For questions about the development and implementation of the Living Datawalk, please contact Niels Janssen of Creative Coding Utrecht.

For other interested parties: contact the Zoönomic Institute via hello@zoop.earth and we will put you in touch with the right person.

Sources and links
This documentation is based on a conversation with Niels Janssen (CCU) and an internal documentation of CCU, developed by Martina Denegri.

 

Have a look at this

Back to all instruments and methods