Artificial intelligence (AI) – innovation for ecological research

We often collect very large amounts of data during ecological studies in the field. This applies in particular to surveys using digital techniques, such as high-resolution digital images and videos or audio recordings. 

To evaluate this data efficiently, we use advanced AI methods, in particular deep learning. This allows recognising complex patterns in large amounts of data and automatically detect and identify object classes such as birds or marine mammals.

The use of AI promises more accurate, faster and more reproducible results compared to purely manual processing. Despite the automation, we continue to apply standardised quality control by experienced experts to ensure that our results meet the highest scientific standards.

Our interdisciplinary team develops AI applications in-house or in collaboration with partners. Individual solutions are optimised to meet the requirements of ecological investigations.

The team includes IT specialists, computer vision engineers, specialists in AI-based image recognition, machine learning developers, biologists and environmental scientists.

The combination of remote sensing methods and AI creates a variety of synergies that significantly enhance the potential for analysing and using large, complex data sets more efficiently. BioConsult SH makes use of this potential today. The portfolio includes

• Digital aerial surveys (HiDef method) for the large-scale monitoring of birds, marine mammals and other marine megafauna,
• the use of drones to map species, vegetation, habitats and beach litter,
• the SPACEWHALE service (www.spacewhales.de), which records large whales on high-resolution satellite images,
• the analysis of automatically recorded bat and bird calls,
• AVES, the camera-based system for recognising and identifying bird species, which was developed to prevent collisions with wind turbines and effectively deter birds, for example at airports.