The objective of this study was to evaluate precision dairy farming technologies (PDFT) for estrous detection. Estrus was synchronized in 24 lactating Holstein dairy cows using a modified G7G-Ovsynch protocol (last GnRH injection withheld to permit expression of estrus) beginning 45–85 DIM. Resumption of ovarian cyclicity at enrollment, presence of a corpus luteum (CL) on the day of the final injection (designated experimental d 0), regression of the CL by d 5, and presence of a new CL on d 11 were verified by transrectal ultrasonography. Cows were observed for estrous behaviors for 30 min, 4× per day, on d 2 to 5. Blood samples were collected on d −2, −1, 0, 1, 2, 5, 7, 9, and 11 to quantify progesterone to verify luteal regression and ovulation. Potential periods of estrus (gold standard) were defined by the temporal pattern of progesterone (>1.0 ng/mL on days −2, −1 and 0, <1.0 ng/mL on d 2 and > 1.0 ng/ml on d 9 and 11). Eighteen cows followed this pattern. Cows that failed to follow the pattern served as negative controls (n = 6). Detection of estrus by PDFTs, an estrous behavioral scoring system, and by visual observation of standing estrus were compared with the gold standard (Table 1). Sensitivity and specificity for detection of estrus was similar among all PDFT. Only 56% of cows that ovulated were observed standing by visual estrous detection. All systems tested are capable of detecting estrus at least as effectively as visual observation. Table 1. Detection of estrus using alerts generated by PDFT¹ and visual observation (n=24)

Detection method	TP	FP	TN	FN	Sensitivity (%)	Specificity (%)
Afimilk AfiAct Pedometer Plus	16	0	6	2	89	100
GEA CowScout (leg version)	15	0	6	3	83	100
ENGS Track A Cow	14	0	6	4	78	100
Agis SensoOr	14	0	6	4	78	100
Estrus behavioral score2	12	1	5	6	67	83
Standing behavior	10	0	6	8	56	100

¹Sensitivity = TP/(TP + FN) and specificity = TN/(TN + FP); where TP = true positive, TN = true negative, FP = false positive, and FN = false negative. ²Scoring system as defined by Van Eerdenburg et al. (1996).