THE ROLE OF THE COW IN AUTOMATIC TEST CUP ATTACHMENT

A system for automatically locating the teats of a cow and attaching teat cups was developed and used to milk nine cows for 10 consecutive d. For the first 5 d, the cows were milked three times a day at fixed intervals of 12, 6, and 6 h. For the subsequent 5 d, cows were intercepted for milking on their way from a bedded area to a forage feeding area; the mean numbers of attendances per cow were 3.2/d (range 2 to 4). Of 279 cow visits for milking, 72% were successful. Of the 77 occasions on which all four teat cups were not attached, 13 were attributable to the response of the cow (for example, kicking the robot); 46 were due to a difference between the estimated teat position and the actual teat position; and 14 were due to operational failures of the equipment. Eighty-five percent of attempts to attach individual teat cups were successful. Of the 162 failed attempts to attach teat cups, 15% were due to cow response, 54% to positional error, and 21% to engineering malfunctions. In treatment 2, cows that stayed in the stall for more than 5 min after milking were prompted to leave, which occurred during 13 (9%) visits. Cow behavior did not appear to be a major obstacle to the unsupervised use of automatic milking.