Fuzzy Logic-Based Decision Support for Dairy Cattle Welfare Integrating Different Benchmarks

Nowadays, one of the key areas of sustainable agriculture is increasing animal welfare. However, in the absence of generally accepted measurement criteria and systems, measuring animal welfare can be considered a subjective area that makes measuring animal welfare complex. As a result, both increasing welfare and making intervention decisions are not clear for farm management. In our research, we develop a fuzzy logic-based decision support system that is able to handle the subjectivity arising from determining animal welfare. During focus group interviews, experts pointed out that animal welfare assessment systems do not provide adequate support in decision-making. However, the integration of different benchmarks (past, best values and competitors) and the triangular membership functions assigned to them in the assessment significantly supports decision-making. The models were tested with data collected with the Welfare Quality Assessment System of three dairy farms (Austrian, Hungarian, and Slovak). In our result the models show different assessment results; therefore, an aggregate assessment model was created by aggregating the results of the models. The aggregate model incorporates the value judgments and importance of the different models by applying the Choquet integral, thereby providing a more accurate assessment according to the criteria that meet the expectations of decision-makers. Our research shows that animal welfare assessment systems should be based on fuzzy logic and the application of multi-criteria benchmarks until standards reduce the uncertainty in measuring animal welfare levels.