AI-Supported Monitoring of Peatland Agriculture (KIMoPa)

Background:

The KIMoPa project is developing innovative AI technologies for the detection and classification of bog vegetation and investigating their practical application in agricultural value chains. The goal is to make the management of paludiculture areas—that is, the agricultural use of wet and rewetted peatlands—more efficient and economically viable.

To date, the heterogeneous vegetation composition of peatlands, consisting of sedges, rushes, cattails, reeds, and other plants typical, has posed a challenge for agricultural practice and material utilization.

KIMoPa addresses this gap: Using drones and artificial intelligence, we generate detailed vegetation maps that enable an objective assessment of the areas—ranging from the occurrence and distribution of individual plant species to the estimation of biomass quality and greenhouse gas potential. Agricultural businesses use this information to adapt their processes and management practices or to utilize the biomass specifically for various applications.

Project Objectives:

• AI-supported vegetation recognition: Development of machine learning models for the automatic identification and classification of bog and paludified crops using drone data
• Digital monitoring platform: Development of an end-to-end solution for use in the field, including interactive visualization of vegetation heterogeneity and management information
• Pilot study on bale labeling: Field testing of the assignment of biomass composition to harvested bales for quality-based marketing
• Practical applicability: Involvement of stakeholders in practical development, investigation of integration potential, and economic feasibility assessment