Automated Bird Surveys, From the Sky

Fly a drone over the colony. AI-BIRD finds every bird, tells you its species, sex and behaviour, checks whether it is alive, and hands you the map and the report.

10
seabird species identified
4
traits per bird: species, sex, behaviour, status
3
steps from flight to report

AI-BIRD (Automated Intelligence for Bird Inventory and Reproduction using Drones) is a web platform that turns drone imagery into bird surveys. Upload the photos from a flight over a colony and trained vision models detect every bird, classify each one, and produce georeferenced inventory maps and exportable reports, ready to compare from one survey to the next.

How It Works

A survey is one flight and a few clicks. The drone does the walking, the models do the counting.

How AI-BIRD works: a drone flies the colony, every bird in the imagery is detected, each one is classified by species, sex, behaviour and status, and the results become a georeferenced map and an exportable report A drone flies the colony, every bird in the imagery is detected, each one is classified, and the results become a georeferenced map with an exportable report.

1. Fly the drone

Capture aerial imagery of your survey area with an off-the-shelf consumer drone. No specialist hardware, and nobody has to walk through the colony.

2. AI detects and classifies

Upload the flight. Trained vision models find every bird in every frame, identify the species, and record sex, behaviour and whether the bird is alive.

3. Get maps and reports

Explore the georeferenced inventory map in the browser, then export detections and summaries in standard formats for your own analysis.

One Bird, Four Answers

Counting is only the start. For every bird it finds, AI-BIRD records four things, so a single survey tells you not just how many birds there are but who they are and how they are doing.

One detected bird with four answers: species, one of ten seabird species; sex, male or female; behaviour, what the bird is doing; status, alive or dead Every detection carries a species (one of ten seabirds), a sex, a behaviour and a status.

What AI-BIRD Does

Automated detection

Find every bird in your imagery without manual counting, frame after frame, survey after survey.

Species classification

Identify ten seabird species automatically with trained vision models, plus sex, behaviour and whether each bird is alive.

Geospatial maps

Turn detections into georeferenced inventory maps, so every bird has a place as well as a label.

Reproducible surveys

Re-run the same pipeline on the next flight for consistent, comparable results across the season and across years.

Exportable reports

Download detections and summaries in standard formats and carry them into your own GIS or statistics workflow.

Avian influenza monitoring

The flagship use case: survey colonies for outbreak and mortality signals without setting foot among the birds.

Flagship Use Case: Avian Influenza

Highly pathogenic bird flu can tear through a dense seabird colony in days. Because a drone survey is fast and disturbs nothing, it can be repeated as often as an outbreak demands, and the alive-or-dead status on every detection turns each flight into a mortality count.

Survey

Count the colony without entering it

Walking a colony to count by hand disturbs nesting birds, risks trampling eggs and rarely covers everything. A drone flight covers the whole colony in minutes and can be flown again next week.

  • Whole-colony coverage in a single flight
  • No disturbance to nesting adults or chicks
  • Safe to repeat through the breeding season
Signal

Read the outbreak in the numbers

Every bird is flagged alive or dead and placed on the map, so a survey shows where mortality is clustering and how it moves. Comparing flights over time reveals survival and reproduction, not just a headcount.

  • Mortality mapped, not just totaled
  • Survey-to-survey comparison on one pipeline
  • Early warning for conservation response

Read about the bird flu monitoring project

AI-BIRD grew out of our work with Lumax AI on monitoring avian influenza in seabird colonies. The project page covers the threat, the field context, and how the survey pipeline came together.

Bird Flu Monitoring project

Our Partners

AI-BIRD is built in close collaboration with Lumax AI and Sovon, the Dutch centre for field ornithology.

Ready to survey from the sky?

Tell us about your colony and your survey needs, and we will set you up with a demo of the platform.