Designing the optimal robotic milking barn by applying a queuing network approach

I. Halachmi, I.J.B.F. Adan, J. van der Wald, P. van Beek, J.A.P. Heesterbeek

Research output: Contribution to journalArticleAcademicpeer-review

19 Citations (Scopus)

Abstract

The design of various conventional dairy barns is based on centuries of experience, but there is hardly any experience with robotic milking barns (RMB). Furthermore, as each farmer has his own management practices, the optimal layout is `site dependent¿. A new universally applicable design methodology has been developed, to overcome this lack of experience with RMBs and to facilitate the designing of their optimal layout. This model for optimizing facility allocation, based on cow behaviour, welfare needs, and facility utilization, uses queuing network theory, Markov process, and heuristic optimization. The methodology has been programmed into a software application, supporting the design process. On a particular farm, presented later as a case study, numerical results include: if the herd contains more than 50 cows, the forage-lane utilization is greater than 70% (or idle time is less than 30%). To meet animal-welfare demands, the herd size should not exceed 60 cows. Therefore, the herd should comprise 50¿60 cows. In the second scenario examined, the average robot idle time was 25%, queue length was three cows, and each cow waited for about 3 min at the robot. It is still uncommon to apply techniques from queuing-network theory to livestock housing, and this study demonstrates their potential as practical design tools that meet both economic and animal welfare needs.
Original languageEnglish
Pages (from-to)681-696
JournalAgricultural Systems
Volume76
Issue number2
DOIs
Publication statusPublished - 2003

Keywords

  • dairy-cows
  • social hierarchy
  • system
  • behavior
  • metabolism
  • water
  • potassium
  • lactation
  • chlorine
  • queues

Fingerprint

Dive into the research topics of 'Designing the optimal robotic milking barn by applying a queuing network approach'. Together they form a unique fingerprint.

Cite this