• robotic milking program canadian dairy trouw nutrition shur-gain

A Comprehensive Guide to Cow Traffic Systems in Robotic Milking

In this article we are looking to give you a generic review of the different cow traffic systems available that can be used with robot milking. The most popular are free-flow cow traffic and forced cow traffic, of which milk-first and feed-first cow traffic systems are the most common. Critically evaluating the different cow traffic systems is difficult as each has its ardent advocates. All three cow traffic systems work well, so it is important for producers to select the system that best suits them, promotes cow well-being, and meets their goals.

In this article we are looking to give you a generic review of the different cow traffic systems available that can be used with robot milking. The most popular are free-flow cow traffic and forced cow traffic, of which milk-first and feed-first cow traffic systems are the most common. Critically evaluating the different cow traffic systems is difficult as each has its ardent advocates. All three cow traffic systems work well, so it is important for producers to select the system that best suits them, promotes cow well-being, and meets their goals.

The basic parameters by which the different systems can be evaluated include the feeding strategy, system capacity, labour efficiency, initial investment, and typical advantages and disadvantages generally accepted in the industry (Rodriguez, 2013). These differentiating criteria reported in the literature and promoted by industry help to set the expectations for each cow traffic system.

The free-flow system

In the free-flow cow traffic system, the cows have unrestricted access to the feed bunk, stalls, robot(s), and water. The cows are fed a partial mixed ration (PMR) at the bunk. The minimum and average amount of grain ration must be greater than that fed in the milk-first system. The main motivation for this system is providing a palatable and high quality pellet. In addition the amount of robot feed delivered is another key factor in motivating cows to visit the robot. This system accommodates 55-60 cows per robot. The initial set-up cost is the lowest of all the systems (Rodriguez, 2013).

The key benefit is that cows have free access to feed, stalls, the robot, and water all the time (Rodriguez, 2013). Early lactation cows are milked more frequently with shorter inter-milking intervals compared to the forced traffic systems (Gygax et al., 2007). There is less risk of cows blocking gates, passageways or walkways (Lely 2013). Some reports suggest that there are more fetch cows compared to the forced traffic systems (Bach et al., 2009). Cows may visit the robot without being milked (refusals), reducing robot capacity. Milking frequency tends to decline with increasing days in milk (Gygax et al., 2007). This may reduce milk production and potentially increase fetch cows (DeLaval, undated). How the ration is balanced is critically important to help maintain turns in late lactation and lower producing cows (Lely, 2013).

The milk-first system

In the milk-first cow traffic system, the cows are pre-selected on their way from the stalls to the bunk. Cows with milking permission are directed to a holding pen to be milked on their way to the feed bunk. If not, they are directed to the feed bunk. The cows are fed a partial mixed ration at the bunk with a minimum grain ration per cow per day fed through the robot. 

The main motivation for cows to visit the robot is the bunk mix (PMR). Managing the partial mixed ration well is extremely important. This system, with no refusals at the robot, accommodates 55-75 cows per robot. The reported number of fetch cows is 1 to 5%, possibly more labour efficient than free flow. The initial set-up costs are higher than those for the free-flow cow traffic system (Rodriguez, 2013).

Benefits include more consistent milking frequency between early and late lactation cows (Gygax et al., 2007) and the reportedly lower number of cows to fetch. The low amount of ration fed through the robot allows for more on-farm feeds with this cow
traffic system (Rodriguez, 2013). Research demonstrated that the number of visits to the bunk is lower than with a free-flow system, but the cows partially compensated by spending more time eating per feeding and ate a larger meal at a faster rate (Bach et al., 2009). There is a risk that cows may get acidosis, especially if they can sort the bunk mix. Low rank cows may wait for extended periods of time to be milked (Jacobs & Siegford, 2012), keeping them away from feed, stalls, and water, unless water is provided in the holding pen.

The feed-first system

In the feed-first cow traffic system, the cows are pre-selected after feeding at the bunk on their way back to the stalls. Cows with milking permission are directed to a holding pen to be milked on their way to the stalls. If not, they are directed to the stall area.

The cows are fed a partial mixed ration at the bunk. The minimum and average amount of grain ration must be greater than that fed in the milk-first system. For this system to be successful the palatability and quality of the robot feed is extremely important. The main motivation for cows to visit the robot is the feed delivered by the robot. The system, with no refusals at the robot, accommodates 55-75 cows per robot. Reported number of fetchcows is 1 to 5%. The initial set-up costs are higher than thosefor the free-flow cow traffic system (Rodriguez, 2013).

The cows have unlimited access to the feed bunk. Only cows with milking permission are in the sort pen waiting to be milked, which increases robot capacity. Cows follow their daily feeding pattern and present more often at the sort gate, resulting in more regular milking. There are reportedly few fetch cows (Rodriguez, 2013). Cows may spend longer than necessary in the feed alley, there may be congestion at the sort gates (Lely, 2013), and cows selected to be milked, particularly low-ranked cows, may spend longer waiting to be milked (Jacobs & Siegford, 2012). Cows will be without water unless it is provided in the holding area.

Advantages to all systems

Despite the reported advantages and disadvantages of the different cow traffic systems, in practice it is evident that all these systems can work very well when managed appropriately.

The sometimes reported advantage of forced traffic over free flow for milkings/cow/day and fetch cows (Svennersten-Sjaunja and Pettersson, 2008, DeLaval, undated) are not necessarily observed on farm.

The single most important factor in robot milking is milk speed, followed closely by milkings/cow/day (or total milkings/day) and kg milk/milking (kg milk fat/milking may be just as important).

Milk speed is primarily determined by genetics and udder preparation at the beginning of milking. Milkings/cow/day is affected by many factors, some of which are summarized in the diagram (Figure 1. Adapted from Hendrix UTD, 2007). The cow traffic or feeding system, although important, is only one of many factors that determines milkings/cow/day. Because of all the factors impacting turns, it is impossible to definitely say that one traffic system is better than another. It is difficult to control for all these factors in research (and surveys), so the results usually vary substantially. The averages may differ but are rarely significantly different.

milking robots barn setup barn style Canadian Dairy free flow feed-first

At a basic level, the maximum turns possible is determined by the total time available, the box time, and the number of cows per robot. If we assume 1,260 minutes per day and an average box time of 7.5 minutes, then we have 168 milkings per day.

With 60 cows per robot, the theoretical maximum number of turns is 2.8/cow/day. A survey of 10 U.S. and 15 Canadian automatic milking system (AMS) herds revealed that herds with >60 cows/AMS averaged 2.4 turns and herds with <60 cows/AMS averaged 2.8 turns. The average turns/cow for all herds was 2.6. However, there was large variation with 36% of farms with fewer than 2.5 turns, 44% of farms with 2.5-3.0 turns, and 20% of farms with more than 3 milkings/cow/day. There was no correlation between milkings/cow/day and the cow traffic system in the herds surveyed (de Jong et al., 2003).

Feeding fresh feed and pushing up feed appear to have a consistent benefit on turns/cow/day. In a survey of 13 free-flow cow traffic herds in Ontario, feeding fresh feed more frequently enticed cows to go up to eat and remain standing after being milked, as well as generally improved cow traffic, turns, milking intervals, and fewer fetch cows (Deming et al., 2013). A lack of feed (bare bunk) at times during the day increases synchronicity among cows. There is more crowding at the robot waiting to be milked, fewer turns, greater variation in milking interval, and typically more fetch cows (Svennersten-Sjaunja and Pettersson, 2008).

Many important factors

Regardless of the cow traffic system and the diurnal pattern, having feed available at the bunk all the time is very important as cows are milked and eat on different schedules throughout the day. This is especially important in the milk-first system. Producers that have push-up robots typically run them once an hour for 20 hours a day. These herds are achieving very high turns and production: more than 3 turns depending on the number of cows/robot (DNA Trip to British Columbia, 2013).

Proponents of the free-flow cow traffic system argue that more than 5% fetch cows (similar to the forced traffic system) means the ration is not correctly balanced for this system, or that there is a high number of lame cows, or both (Lely, 2013, Hulsen, undated). In a survey of 43 herds, the fetch cows in 35 free-flow traffic herds was 16.2% ± 10.8%, and in eight forced cow traffic herds it was 8.5% ± 5.9%. Even though the averages are different, the variation is so large that we can’t conclude that the number of fetch cows between these traffic systems is truly different. Of all the fetch cows, 19% were classified as lame and 57.6% as non-identifiable reason (Rodenburg, 2007), possibly “lazy cows,” a result of not having a correctly balanced ration for the system.

The low synchronicity in milking and other activities, regardless of the cow traffic system, makes it difficult to obtain good research results on cow behaviour, as well as feeding and lying behaviour in robot herds. The contradictory results and large variations often seen in surveys (and research) means that differences in farm management and individual farm variations are more important to production and efficiency than the specific features of the robot (Deming et al., 2013, Jacobs & Siegford, 2012) or the cow traffic system.

The appropriate cow traffic system and feeding recommendations will vary by farm. The best option is to choose the robot and cow traffic system that best reflects your philosophy, management style and goals. Understand the key functions, features, and characteristics of the robot and the cow traffic system.

Understand the many factors that affect milkings/cow/day (and kg milk/minute; kg milk/milking); these ultimately translate into kg milk per robot. Optimize each of these to facilitate the most efficient use of your robot, while promoting cow well-being and the bottom line.