Read the Progressive Dairy Canada digital edition

Exit alley blocking incidences and automatic milking systems

Janice Siegford & Jacquelyn Jacobs Published on 30 June 2011

Automatic milking systems, also called robotic milking machines, offer the potential to milk cows more frequently when compared with a conventional parlor system that must rely on hired or family labor. More frequent milking may in turn generate more milk when paired with consistent milking intervals.

Robotic milkers also are claimed to reduce the labor needed to milk a herd, which can free up time for other farm management tasks or family and recreational activities. This claim makes robotic milking attractive for farmers looking for relief from a demanding milking routine or hoping to spend less money on labor.



Of course, in order for a robotic milker to achieve more milkings (i.e. and potentially more milk) and reduce labor, the milker must function optimally and efficiently. All robotic milkers rely on cows acting as individuals, apart from their herdmates, and voluntarily entering the milking stall without encouragement from the farmer.

As a result, understanding the interaction among the cows and with their environment is critical to the success of the system.

To accommodate robotic milkers, barns must be built or retrofitted to encourage efficient cow traffic through the robot and to promote normal lying and feeding behavior. A large open area (6 to 7.3 metres ) at the entrance to the milking stall is recommended in order to give low-ranking cows the ability to avoid dominant cows when they need to be milked.

An exit alley with a one-way gate 1.8 metres or more from the milking stall is also recommended to encourage cows to move away from the robot once they have been milked.

When the new freestall barn was built at the Kellogg Biological Station Pasture Dairy Research and Education Center (PDREC), the design included the recommended exit alley and large open area near each robot. However, while collecting data for other projects at PDREC, we noticed that some cows seemed to block the end of the exit alley leading away from the milking stall.


Figure 1

These cows would stand with their heads facing the end of the exit alley and appeared to keep cows that had finished milking from exiting the alley (Figure 1). Occasionally, this blocking seemed to result in a cow traffic jam, sometimes forcing the milker to stop accepting new cows for milking and causing other cows to wait.

Figure 1: The white cow (white arrow) standing with her head at the gates of the exit alley. She is preventing the cow in the exit alley from leaving the alley. This action is what we called a “blocking or guarding event.”

Cows normally have access to the milking stall 22 hours a day, with two hours of down time dedicated to cleaning the robot. If cows blocking the exit cause the milker to be available for less than the expected 22 hours, cows in herds attempting to use the robots at their full capacity could be milked less frequently than would be ideal for their stage of lactation, production level or desire to be milked. Ultimately, this could mean herd milk production will not reach its full potential.

Figure 2

We wondered how often blocking might be occurring, how much of the robot’s time was wasted, whether cows that blocked regularly shared similar characteristics and whether cows that were blocked were similar to each other but different from blocking cows.


Figure 2: A cow (white arrow) blocking the exit to the robot. In this particular case, two cows are stuck in the exit alley, one cow (which has finished milking) is trying to exit the milking stall, and one cow is waiting to access the robot, and two cows are in the entrance alley. Thus, a single cow has effectively prevented three cows from milking and three cows from exiting and moving out of the robot area.

Prior to our work, little research had examined the behavior of cows in the exit area of robotic milkers. As mentioned earlier, an exit alley with a one-way gate 1.8 metres or more from the milking stall has been recommended in barns with robotic milkers to allow subordinate cows to move around easily near the exit area.

As part of studies examining several types of cow traffic in robotic milking systems, Stefanowska and colleagues found indications that exit speed was slowed when other cows were in the exit area. The authors also found that successful milking visits led to quicker exits from the milking stall compared with unsuccessful milking visits.

After an unsuccessful visit, 50 percent of the cows in their study revisited the milker within 30 minutes. Cows in that study also exited the milking stall more quickly if they were given a larger portion of feed concentrate.

In our study, the effects of blocking behavior in a passage (exit alley) between the robotic milker and the holding area were examined. Eighty-four Holstein cows of various parities, stages of lactation and days in milk (DIM) were divided between two pens.

Presence or absence of horns, horn positioning and bodyweight were noted. Cow locations and behaviors in the robotic milker entrance, exit areas and the adjacent holding area were recorded continuously for 14 days.

In our study, unsuccessful milking visits resulted in cows hesitating more often when leaving the milking stall (P<0.05), but unsuccessful cows were no more likely to hesitate leaving the exit alley than cows that had successful milking visits. In addition, cows receiving the “no-milking” decision were more likely to immediately circle and re-enter the robotic milker (P<0.001).

Cows exiting the robotic milker were significantly more likely to hesitate (p<0.05) when another cow was located directly at the exit alley gate. However, exiting cows were not more likely to hesitate in the exit alley if the other cow was located elsewhere in the area adjacent to the exit gates (holding area).

This suggests that intentional “blocking” of the exit alley may encourage hesitation of the exiting cow, but the presence of a cow in the holding area in general does not seem to slow exiting cows.

In addition, DIM and presence of horns were predictors for hesitation in the exit alley (P<0.05). More specifically, cows in early lactation with no horns (P<0.001), cows in mid-lactation with only a left horn (P<0.05), and all late-lactation cows (P<0.05) were more likely to hesitate compared with other combinations of horns and stage of lactation.

Cows late in lactation may have less motivation to exit than cows in other stages. Alternatively, cows with no horns or only a left horn may be more subordinate to other cows, and therefore easier to block.

There is enough evidence to suggest that a delicate balance of motivation to enter the robotic milker must be achieved; voluntary approach to the robotic milker is necessary to decrease farm labor, but unnecessary visits should be avoided to help promote an efficient system.

It should be noted that these findings are specific to the facility design and cow numbers at the PDREC and may not be able to be generalized to all barns with robotic milkers. Anecdotal reports of cows revisiting robots and blocking and hesitating in exit areas are common, but rarely have been published in scientific literature.

To determine how universal the problem is, more research should be done using different facility designs. In situations similar to those at PDREC, there may be several solutions to the problem.

The first might be to provide feed in external concentrate feeders located in the feed alley of each pen, to encourage the cows to move out of the exit alley and holding areas. Exit alley design also could be improved by using solid, curved sides rather than the straight gated V-shape currently in use at PDREC.

The results of our study can be viewed as a first step into the investigation of efficient cow traffic flow in and around the exit areas of robotic milkers. We are now analyzing additional data to determine if blocking events reduce the number of possible milkings a robot can achieve each day. If this is the case, the production potential of a robotic milking herd could be decreased, especially when robots are being used at full capacity.  PD

References omitted due to space but are available upon request to .

—Excerpts from Michigan Dairy Review, Vol. 16, No. 1, pages 18-20.