From Particle Physics to Birdsong: CERN’s AI Sensors Advance Eco-Innovation in Biodiversity Monitoring

Birds – both local and migratory – seen on CERN sites. (Image: CERN)

(IN BRIEF) CERN has expanded its AI-based environmental monitoring initiative, “Ambient,” which uses solar-powered audio detectors to identify bird species across its vast campus. Developed in collaboration with the University of Côte d’Azur’s LEAT laboratory, the system employs low-energy EdgeAI to analyze birdsong locally, sending only detection data via long-range networks. After promising results from 11 detectors deployed in May, 40 more will be installed before the end of 2025. The project, which focuses on 14 regional bird species, highlights CERN’s commitment to sustainable innovation and biodiversity preservation through advanced AI and energy-efficient technologies.

(PRESS RELEASE) GENEVA, 4-Nov-2025 — /EuropaWire/ — CERN’s campus, often associated with cutting-edge particle physics, is now also becoming a hub for environmental innovation. The “Ambient” initiative, an artificial intelligence (AI)-driven biodiversity monitoring project, is using smart, solar-powered detectors to identify birds by their song. After successful results from the first 11 detectors deployed in May, the team plans to install 40 additional units by the end of the year to enhance biodiversity mapping across CERN’s 625-hectare site.

“Far from being a concrete jungle, CERN actually has huge areas of greenery,” said Felix Schachtschneider, a member of the Ambient team from CERN’s Science and Civil Engineering Department. Of the site’s total area, 415 hectares remain as natural environments, providing habitats for a wide range of animal and plant species, including both resident and migratory birds.

Members of the Ambient project install detectors, stored in wooden “birdhouses” around the CERN site. (Image: CERN)

Developed through a collaboration between CERN’s Site and Civil Engineering Department, its Occupational Health, Safety and Environmental Protection (HSE) Unit, and the University of Côte d’Azur’s Laboratory for Electronics, Antennae and Telecommunications (LEAT), the Ambient project aims to create an efficient, autonomous system for monitoring bird populations.

Each Ambient detector is discreetly housed in a wooden structure resembling a birdhouse — though designed to keep birds out. These small, solar-powered stations contain circuit boards equipped with low-energy AI models that process sound locally. Unlike traditional systems that transmit large audio files over Wi-Fi to data centres, these devices analyze the birdsong on-site and send only the detection results. This EdgeAI-based approach significantly reduces energy consumption and extends the detectors’ range via Long Range Wide Area Network (LoRaWAN) connectivity.

“We chose to work with a low-energy AI that processes the audio directly on the sensor,” Schachtschneider explained. “Each time it detects a bird, it sends only the species name rather than the entire recording. This allows us to operate sustainably while collecting valuable biodiversity data.”

Ambient’s AI programme is trained to detect the song of different bird species typically found around the CERN campus. This graph shows the total number of each species detected from May to June 2025 – the same bird may be recorded multiple times. (Image: CERN/ Bird Image: The New York Public Library)

The AI model was originally based on BirdNet, an open-source neural network capable of identifying over 3,000 bird species. For greater efficiency, the Ambient system has been adapted to recognize 14 local species common to CERN’s region. Data collected from the first phase of detectors has already been used to retrain and refine the model, which will be further optimized as more detectors are deployed.

By combining AI innovation, energy efficiency, and ecological awareness, CERN’s Ambient project is contributing to a broader understanding of how technology can support sustainability. The initiative is also a step toward identifying and preserving biodiversity across scientific campuses, reinforcing that even in a place dedicated to physics, nature has a vital role to play.

Media Contact:

Emma Hattersley
Tel: +41 (0) 22 767 34 32 or +41 (0) 22 767 21 41
Email: press.office@cern.ch

SOURCE: CERN