Oestrus detection of sows is an important activity within pig husbandry. It takes about 30% of the overall labor input, and financial losses related to non-productive days may vary considerably. Therefore, the principle of using an infra-red sensor has been evaluated to detect automatically oestrus of individually housed sows. The infra-red sensor used is a commercially available device, primarily intended for security alarms and which consequently had to be adapted for scientific use. The movements of the sow are converted to an analog signal by the infra-red sensor, and a mean value is calculated. The frequency range of the sensor is 0.01 Hz to 10 Hz and the maximal sensitivity is approximately 0.1 Hz. The bandwidth is 0.72-1.6 Hz and determines the range of detectable movements of 0.2 m/s to 0.5 m/s. An ultra-sound system is equally installed above the sow to get more information about the movements and an indication about the position of the sow (standing up or lying down). In the near future, the results of these measurements will be evaluated.

The infra-red sensor was mounted 50 cm above the front of the sow's body. Fifty-eight multiparous individually housed sows were monito-red from the day after the piglets were weaned. Four parameters of body movement as quantified by the sensor's output voltage were investigated: mean daily activity, stand-ard deviation of mean daily activity, minimal and peak value. The reference method was the standing behavior before the boar and the inseminator. 80% of the sows could be classified correctly when using mean daily activity as the selection parameter. Up to 86% were classified correctly when daily peak activity was also inclu-ded. Positive and false positive sows could be distinguished at the 95% level by using a canonical discriminant analysis. Increase of mean daily activity, increase of standard deviati-on and increase of daily peak activity were statistically significant explanatory variables of a logit funtion predic-ting the onset of oestrus.