Implementation of unmanned cow estrus detection system for improving impregnation rate

In the paper, we reduce non-pregnant conditions and improve impregnation rate by unmanned estrus detection and decide proper time for artificial insemination. It is too hard to detect estrus only by using activities, we develop unmanned estrus detection system that consist of RF activity sensors, cow management program and estrus detection algorithm that uses information of activities and breeding. We verify performance by experiments in four similar scale stockbreeding farmhouse. Each stockbreeding farmhouse breeds 87, 81, 93, 82 cows and expected estrus cows are 14, 19, 15, 17. In expected estrus cow, we fail in weak estrus detection - 3, 2, 1, 3 cows, but detect successfully normal estrus - 11, 17, 14, 14 cows. After artificial insemination, 10, 17, 13, 14 cows became pregnant successfully confirming that proposed unmanned estrus detection system is effective for deciding proper time for artificial insemination in normal estrus.