BiodivKI2 - Biodiversitätsfaktormessung mit Intelligenten Akustischen Sensoren (BioIntAkt-2)

Teilprojekt: Robustes, effizientes und intelligentes akustisches Sensorsystem

Project Period: 01.01.2025 - 31.12.2027

Biodiversity is considered an important indicator of the resilience of ecosystems. The decline of entomofauna in many ecosystems is currently the focus of public and scientific discourse. Biodiversity monitoring to identify insects of different functional groups in flowering areas and agroecosystems can therefore be seen as an integral part of sustainable land use and as a basis for agricultural strategies such as integrated pest management (IPM). Current solutions for the taxonomic recording of biodiversity are time-consuming, cost-intensive and require expert knowledge for many species groups. Depending on the method used, only certain groups are selectively captured, which means that only individual biodiversity factors can be assessed. Digital methods to support the determination of population densities and the diversity of species communities have enormous application potential, but have hardly been used to date. At the same time, advances in computer science and data science have led to the development of methods that allow the autonomous and self-learning taxonomic identification of species using artificial intelligence (AI) methods. However, a corresponding application in the field of entomofauna is not yet known, despite the enormous potential for identification. In this project, a digital, AI-supported solution for the monitoring of insects is being developed that works on the basis of acoustic signatures. In addition to the development and practical testing of a sensor system for collecting and processing acoustic data, this will be deployed, calibrated and evaluated in situ in practical studies. The combination with citizen science, i.e. a smartphone app for collecting acoustic measurements, will also complete the overall picture in a regional and national context and inherently raise public awareness of the topic of biodiversity.

Smart Sensors Group of Hamburg University of Technology (TUHH) is a cooperation partner within this joint project and is responsible in particular for the development of the intelligent, robust and reliable acoustic sensor system. Furthermore, the TUHH is working on methods for efficient data pre-processing on the already resource-limited sensor system, which includes the extraction of necessary parameters for local decision-making as well as general compression of measurement data. In this context, dedicated HW accelerators are being developed for the processing of embedded AI, which should enable efficient, local classifications.

Project Website

A joint project website, which provides information about the entire project, can be found here

People

Prof. Dr.-Ing. Ulf Kulau

Christopher Büchse

Davis Rakhshan

Publications

FGSN Workshop Potentials and Challenges of Ecoacoustics for Environmental Monitoring: A Small Field Study

Davis Rakhshan, Jan Lueken, Lucas Bublitz, Chingiz Seyidbayli, Andreas Reinhardt, Ulf KulauProceedings of the 21st GI/ITG KuVS Fachgespräch Drahtlose Sensornetze2024Accepted for publication.
[BibTex]