Go-2-market of a predictive agri-tool for identification of milking cows in need for treatment with sensor data The aim of the first project phase is the adjustment of known and, where applicable, extended sensor and analysis combination on retrospective data records of 5 German dairy herds belonging to research stations that are connected as full members to the web-based database system 'KuhDaM” by the company TiDa GmbH. The focus of the second project phase is then the development of a tool for the identification of cows requiring treatment. This is performed on the basis of insights gained from algorithm applications in the KuhDaM dairy herds and is primarily undertaken by the commercial partners of this project due to their expertise in the area of web-based applications. The first step is the inventory and evaluation of the algorithms for the identification of dairy cows requiring treatment. The second step is the planned development of an automated evaluation and decision tool to improve identification of dairy cows requiring treatment. Decision criteria here include in particular the sensitivity, precision, positive and negative predictive values as well as the coefficient of determination. Supported by the scientific expertise of the university, the integration of predictive forecasts is then undertaken using SAS Analytic components within the 365 FarmNet management platform (Dairy Management powered by GEA). The decision criteria for this are the provision of correct, complete and plausible data, cross-system standardisation of the data room and the formats, availability of a consolidated historical data pool as a basis for analytics, as well as a provision of data rooms according to the client’s requirements (data protection/sovereignty). The final phase is the 'Go 2 market” and/or the go-live in the 365FarmNet-shop, which comprises a user-friendly feel and look, as well as the completion of an App for mobile use, including the successfully completed test phase with existing regular clients of 365FarmNet/GEA.