Abstract
The traditional barn design is a milking parlour oriented. To integrate a milking robot the barn should be redesigned according to the robotic milking concept. The entire system (barn design, feeding and cow-traffic routines, management practices) should encourage 'voluntary milking', i.e., it should ensure sufficiently frequent visits of the cow to the robot.
An optimal layout should balance animal welfare, on the one hand, and the economic need for high facility utilisation, on the other hand. These two conflicting forces (which are to be optimised) should be incorporated into management practices and physical layout. However, the actual capacity performance) of each facility (such as robot, forage lane, concentrate feeder) in the robotic milking barn (RMB) depends on cow access (animal behaviour), barn design, farm routine and management practices. There is also a wide diversity among farmers and local conditions, therefore the optimal layout may differ among farms. Numerous important factors are evaluated as a system when RMB is designed, but it is usually difficult to quantify how consideration of these factors may affect production or income. Taken together these factors and their variations mean that we are dealing with a quite complex system.
The optimal RMB layout (the solution) has to be matched to individual farm conditions, adjustable for any farmer or site, but the design methodology should be universally applicable. Therefore, the objective of this study was to develop a design methodology for finding the optimal layout for a robotic milking barn before the barn is built, and to implement the methodology into a practical design tool, embedded in a user-friendly software application, ready for use in the barn during a consulting session.
Four experiments were conducted, two under research conditions and two in commercial farms. They aimed to explore the stochastic nature of the facility utilisation in a robotic milking barn, and to validate the model under a variety of scenarios.
A closed queuing network model for a robotic milking barn was developed, and a behaviour-based simulation (BBS) model, which enables a designer to optimise facility allocation in a barn, was developed and validated. Having been validated, the simulation model becomes a practical design tool for optimising a barn layout. The design methodology was finalised by integrating the queuing network model, the BBS model, a regression metamodel, full factorial design, and optimisation algorithms. By using the proposed design methodology, a model of a future barn can be created, which will help to make effective decisions. It is possible to predict how the barn will respond to changes in design or operation, and to compare what will happen under a variety of scenarios. Among other things, it is now possible:
- to predict facility utilisation and cow queue length;
- to calculate the optimal facility allocation: the numbers of robots, cubicles, forage lane positions, water troughs and concentrate feeders that are needed;
- to advise the individual farmer on the choice of robot location, cow traffic routine, required floor space in front of each facility (waiting area), feeding routine, separation area, automatic cleaning; and
- to gain assurance before building that the proposed design would actually meet the specified requirements.
In general, this research has reached a stage at which a behaviour-based simulation is adjustable for any farmer or site. The onus is now on the industry to implement this proposed design methodology on a daily basis.
Original language | English |
---|---|
Qualification | Doctor of Philosophy |
Awarding Institution |
|
Supervisors/Advisors |
|
Award date | 10 Nov 1999 |
Place of Publication | S.l. |
Print ISBNs | 9789058081308 |
DOIs | |
Publication status | Published - 10 Nov 1999 |
Keywords
- milking parlours
- animal housing
- milking machines
- robots
- design
- layout
- methodology
- mathematical models
- simulation models
- optimization