AI- and citizen science-supported biodiversity monitoring, reporting, and certification

Biodiversity loss erodes ecological resilience and threatens the long-term viability of agricultural systems, yet monitoring efforts remain fragmented, costly, and difficult to scale. The project, KICS-ZERT, addresses these limitations by integrating artificial intelligence (AI) with citizen science to establish a practical, smartphone-based framework for assessing biodiversity and ecosystem structures. Citizen science, understood as the contribution of lay people to scientific research, has become essential for filling spatial and temporal data gaps and for enabling scaling that conventional research teams cannot achieve. By leveraging widespread smartphone use and recent advances in AI, the project tests whether lightweight, device-agnostic data collection combined with open-source models can produce a more consistent evidence base for biodiversity certification and CSRD-aligned corporate reporting. The approach integrates species identification and structural habitat assessment into a single workflow, reducing reliance on specialized equipment and expert-only field campaigns that often fail to operate at landscape scales. A key output is a scientifically validated indicator list suitable for certification and reporting contexts, addressing the current methodological inconsistency of industry practices.

The project also investigates the social and operational dimensions of participatory monitoring, examining how citizen engagement, co-creation with nonacademic partners, and heterogeneous data inputs affect data quality, acceptance, and feasibility. While citizen science can generate large volumes of observations, upscaling introduces risks related to uneven coverage, variable expertise, and potential overconfidence in AI-generated classifications. KICS-ZERT therefore scrutinizes these constraints rather than assuming that participatory data or AI ensures reliability. Embedding the resulting tool into an online marketplace for nature-positive projects enables companies to use these assessments to estimate ecological impacts in supply chains, albeit with critical attention to uncertainty and interpretability. Partnerships with research institutions, NGOs, and citizen-science networks provide diverse datasets and testing grounds for evaluating the limits of real-world deployment.

Through the combined development of AI models, ecological validation, and participatory methods, KICS-ZERT aims to produce a realistic and technically grounded contribution to biodiversity assessment without inflating the capabilities of current AI technologies or overlooking the complexities of citizen-driven data collection.