Biodiversity Monitoring of Grasshoppers with an AI-Supported Mobile App (ChirpScan)
Auch auf Deutsch verfügbar
Content of this page
Profile
- Project leadership
- Tobias Peter (ZHAW), Adrian Busin (ZHAW), Matthias Riesen (ZHAW), Matthias Nyfeler (ZHAW)
- Project team
- Oliver Hawlitschek (UZH), Niklaus Heeb (ZHdK), Judith Lanka (PHZH)
- Organisational unit
- F-FD (Fachdidaktische Forschung)
- Project phase
- Completed (1.7.2025 – 30.6.2026)
- Project type
- Research & development project
- Funding
- Digitalization Initiative
- Participating institutions
- Zürcher Hochschule Angewandte Wissenschaften, Universität Zürich, Zürcher Hochschule der Künste, Pädagogische Hochschule Zürich
Description
Grasshoppers are important bioindicators of the ecological quality of habitats, as they are sensitive to environmental changes and serve as a crucial food source for many animal species. Their presence and diversity play a key role in assessing biodiversity. Given the drastic decline in insect diversity in recent decades, efficient and comprehensive monitoring of grasshoppers – and thus biodiversity – is more urgent than ever.
The project is developing the freely available ChirpScan app, which uses artificial intelligence to automatically classify grasshoppers based on their songs. The app relies on a specially developed deep learning model that enables reliable species identification for both experts and laypeople. It significantly simplifies biodiversity monitoring for environmental agencies, organisations, and research groups.
Since grasshoppers are familiar to many people, relatively easy to observe, and appeal to both sight and hearing, they are particularly well-suited — compared to other invertebrates – for educational projects in the field of biodiversity. As part of the project, the app will be evaluated for its suitability for students in science education at schools and as well as universities students. The app also promotes citizen science by allowing observations to be shared on platforms such as observation.org. This enables comprehensive and resource-efficient recording of grasshopper diversity as an indicator of overall biodiversity.
Publications
Peter, T., and M. Nyfeler (2024) Abschlussbericht zur Vorstudie «Heuschrecken Klassifizierung mithilfe von Deep Learning»