The application of Fuzzy Logic in programs for mastitis detection in dairy herds

Otto Kaufmann; Stefan Koehler

Authors

Otto Kaufmann Humboldt-Universität zu Berlin, Faculty of Agriculture and Horticulture, Institute of Animal Sciences, Philippstrasse 13 Haus 10, D-10115 Berlin.
Stefan Koehler Institute of Agricultural and Urban Ecological Projects at Humboldt-Universität zu Berlin (IASP), Invalidenstrasse 42, D-10115 Berlin

Keywords:

Fuzzy Logic, animal health, mastitis, modeling of parameters, Automatic Milking System

Abstract

An important factor influencing milk production is the health of the animals. In quality assurance programs for milk, the health of the cows is the most outstanding parameter. In the case of Automatic Milking Systems (AMS), the udder health becomes especially important due to the lack of visual control by the milking staff. On the other hand, AMS facilitate the opportunity to measure a lot of parameters of the milk and the milking process specifically for the single quarters of the udder. From this collected data, conclusions can be drawn towards possible udder diseases. The objective of the investigation was to elaborate an analysis system serving to supervise the udder health using the concatenation of selected milk- and milking process parameters. Fuzzy-Logic was selected as an appropriate method of dealing with vague data and diffuse relations. Based on the data of a thorough investigation in a dairy farm with two AMS, three different methods of modeling were selected, analyzed and compared with one another: models of a single parameter (threshold models), models of indices and Fuzzy logic. While the threshold models showed higher values of specificity, the best results of sensitivity were calculated for the models of indices. However, the optimal result was obtained with Fuzzy logic. The corresponding modeling featured the smallest statistical probability of wrong classifications (5.9 percent of incorrect diagnoses).
The results confirm the hypothesis that Fuzzy logic is a manageable and effective method to uncover the relations between parameters of milk and the milking process as well as the udder health.

References

De Mol, R.M. Woldt, W.E. (2001): Application of Fuzzy Logic in automated cow status monitoring. Journal of Dairy Sciences, 84(2) 400–410. https://doi.org/10.3168/jds.S0022-0302(01)74490-6

Frost, A. R., C. P. Schofield, S. A. Beaulah, T. T. Mottram, J. A. Line, C. M. Wathes (1997): A review of livestock monitoring and the need for integrated systems. Computers and Electronics in Agriculture, 17(2) 139–159. https://doi.org/10.1016/S0168-1699(96)01301-4

Köhler, S. (2003): Nutzung von Prozessparametern automatischer Melksysteme für die Erkennung von Euterkrankheiten unter Verwendung von Fuzzy Logic, Diss. Humboldt-Universität zu Berlin (2002), Shaker-Verlag(Agrarwissenschaft), Aachen.

Maltz, E., J. H. M. Metz (1994): An individual approach to manage the dairy cow: A challenge for research and practice. International Symposium, Tumba, Proc. 267–282.

NN (1994): Leitlinien zur Bekämpfung der Mastitis des Rindes als Bestandsproblem. Deutsche Veterinärmedizinische Gesellschaft. Gießen. 3-930511-03-7.

Schön, H., G. Wendl, M. Klindworth, J. Harms (2001): Precision Livestock Farming – Konzeption, Stand der Forschung, Zukunftsperspektiven, 5. Tagung Bau, Technik und Umwelt in der landwirtschaftlichen Nutztierhaltung, Hohenheim. 06.-07.03.2001, Tagungsband, 1-8, 3-9805559-5-X.

Wollrab, J. (1989): Eutergesundheitsstörungen. Hrg. Busch, W., K. Elze und W. Lange: Grundriss der Tiergesundheitslehre, Gustav Fischer Verlag Jena, 3-334-00291-8.