Development of a fuzzy system for estrous detection

Author(s)

Leandro dos Anjos Brunassi

• Document type: Master's Dissertation
• Press: Campinas, SP.
• Institution: Universidade Estadual de Campinas (UNICAMP). Faculdade de Engenharia Agrícola
• Defense date: 2008-02-21

Examining board members

Daniella Jorge de Moura; Maria da Graça Pinheiro; Luiz Henrique Antunes Rodrigues

• Advisor: Daniella Jorge de Moura

Abstract

Production losses due to lack of precision in detecting estrus in dairy cows are well known and reported in milk production countries. As a result of technical progress in monitoring dairy cows using fuzzy logic automatic estrous detection has become possible by allowing the estimation of the moment of the estrous onset with minimum error. The objective of this work was to develop a fuzzy logic system combined with ROC which was able to efficiently detect estrous in dairy cows. For the input data, the system combines previous estrous cases information and prostaglandin application with activity measurements data. ROC curves will be used in order to make appropriate adjustments. The system outputs so-called alerts were categorized in three types: 'in estrous¿, 'maybe in estrous¿ and 'not in estrous¿. The system¿s validation was carried out in a commercial dairy farm located at São Paulo, Brazil, with a herd of 350 lactating cows. The performance of the test was measured by calculating its sensibility (estrous correctly detected) and specificity (non estrous detected as estrous). Within six month trial, over 25 thousands cases of estrus were analyzed from the herd database collected during the year of 2007. The sensibility found was 84.2% indicating that the developed system was able to detect estrous efficiently and it improves automatic estrous detection (AU)