Individual differences in training predict how well dairy cows adjust to automatic milking

Treatment was randomised at the cow level in one pen and assigned at the pen level in the remaining three. During PRT, cows were given one opportunity daily to approach and enter the AMS, with pelleted concentrate as an incentive. After each session, cows were scored on the furthest stage of engagement reached: Step 1 (no interaction), Step 2 (approached but did not enter), or Step 3 (entered or walked through).

Cow behaviour was monitored for four weeks following the transition, with the daily maximum interval between AMS visits as the primary outcome variable. We tested the effect of PRT in the randomised pen, then examined whether training progression predicted subsequent AMS use across all PRT-trained cows. PRT had no detectable effect on maximum visit interval in the randomised pen (difference between control and treatment: 8.3 ± 39.2 min; 95% CI: -70.7 to 87.3 min).

Among PRT-trained cows, however, those that engaged the least had the longest maximum intervals between visits: Step 1 (n = 6; 877.0 ± 50.4 min), Step 2 (n = 50; 720.0 ± 20.5 min), and Step 3 (n = 54; 593.9 ± 17.7 min). Individual responses to PRT may provide an early indicator of how well a cow will behaviourally adjust to AMS milking, allowing those at greater risk of a difficult transition to be identified before it begins.

More broadly, these findings suggest that understanding AMS adaptation in dairy cows may require looking beyond the overall effects of training interventions or broad behavioural traits, and instead focusing on the behaviours cows express towards the technology itself as a more direct indicator of adaptation.