Validation of simulation model for robotic milking barn design

As milking robots are a recent technological development, there are few precedents and little experience to draw upon when designing robotic milking barns. There is wide diversity among farms, so the optimal layout of the robotic milking barn (RMB) varies accordingly. Using a behaviour-based simulation model, the design focused on optimal facility allocation and its relation with feeding routine, herd size, management practices, etc. This paper applies validation study, compares data on real and simulated RMBs. Measurements from a real robotic farm with 10, 20, and 29 cows are compared with simulation data. The simulation model appears to be a valid, accurate representation of the real system, under commercially feasible conditions. This hypothesis is tested statistically and is not rejected at α=2.5%. So, the conclusion is that the model is a practical design tool, enabling the designer to optimise facility allocation and barn layout.