Automatic observation of cow leg health using load sensors

The milking robot offers a unique possibility for dynamic measurements of the leg health of dairy cows. Four strain gauge scales were installed into a milking robot. The sensors were connected to an amplifier and the data were collected into a PC using dedicated computer programs. The measurement was automatically started and ended based on the cow ID acquired from the milking robot. MATLAB was used for automated data manipulation of over 10,000 data files and criteria for detecting leg injuries were analyzed. The leg weight monitoring was developed into a real-time system with software that alerts the user of possible hoof diseases and other leg problems. The system makes it possible to detect an injured leg by separately measuring the load on each leg. It is also possible to analyze the step and kick behaviour of the cow during milking by measuring the leg weights and analyzing the number of kicks of the cow. (c) 2007 Elsevier B.V. All rights reserved.