The hypothesis that sensors can detect changes in behaviours and physiologiesassociated with cows becoming clinically lame was tested by comparing trends of sensor data from lame cows with non-lame cows. Sensor data included data from live weighing scales, pedometers (average steps per hour) and milk meters collected between November 2010 and June 2012 on five Waikato dairy farms. Farmers were trained in detecting and diagnosing lame cows. For each lameness event (n = 318 events affecting 292 cows) a time period of 14 days prior to the day ofdetection was randomly matched by farm and date to 10 non-lame cows. In this period, lame cows decreased in weight, steps taken per hour, milk yield in the first two minutes of milking, total milk yield, and milking duration. Lame cows also entered the milking platform later. In comparison, non-lame cows had no change in sensor data trends. These differences (P <0.05) in sensor data trends imply potential value of sensor data in detecting lameness automatically. Large variations in sensor data values between and within lame and non-lame cows indicated that future research should focus on combinations of variables that show the best potential to detect lameness automatically.

C, Kamphuis, JK Burke, and JG Jago

Proceedings of the New Zealand Society of Animal Production, Volume 73, Hamilton, 5-10, 2013
Download Full PDF BibTEX Citation Endnote Citation Search the Proceedings

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.