Funding tool: Applied research projects
Project-ID: FTI25-A-017
Project start: November 25^th, 2026
Project end: October 25^th, 2029
Runtime: 36 months / not yet started
Funding amount: € 359.929,00

Brief summary:
European agriculture, particularly small and fragmented vineyards, faces increasing pressure from climate change, volatile microclimates, and plant diseases such as Oidium (powdery mildew). Although disease risk can be forecast using microclimate parameters before visible symptoms occur, existing digital farming solutions rely largely on permanent sensor infrastructures or centralized cloud processing. These approaches are costly, inflexible, and often unsuitable for rural regions with intermittent connectivity, limiting adoption by SMEs. The Disease Forecast as a Service (DiFaaS) project addresses this gap through a scalable, service-oriented, and Edge-AI-enabled platform. DiFaaS enables temporary, on-demand deployment of monitoring stations during critical disease-risk periods. These stations capture relevant microclimate and canopy parameters and execute tailored Edge-AI models locally. Data retrieval is performed flexibly via drones acting as mobile data mules or smartphones, coordinated through a unified service architecture and supported by an edge-native MLOps framework for systematic model deployment and updates. The project pursues three objectives: (1) defining and validating a DiFaaS service model aligned with agricultural and business requirements; (2) designing a scalable architecture integrating sensing, drones, edge nodes, and orchestration; and (3) developing and experimentally validating Edge-AI and MLOps concepts through an Oidium forecasting demonstrator. While vineyards serve as the representative use case, the solution is designed for transferability across crops and diseases. By combining temporary sensing, multi-modal edge data fusion, and lifecycle-managed AI deployment, DiFaaS advances beyond the current state of the art and strengthens regional competencies in Edge-AI and digital agrifood systems in line with the FTI Strategy of Lower Austria.

Keywords:
Artificial intelligence, Internet of Things, Edge-AI, MLOps, Plants Disease Forecast