An AGV system has been widely used by organization to supply components to production systems, and retrieve manufactured good from those production systems to warehouses. Its benefits are widely recognized by its cheaper operational costs (being that they are automated, hence, no human salary is directly attached to it), being able to work 24/7 without breaks, and algorithm-based decisions. However, some challenges may arise upon its inception, namely guidance and its own algorithm-based decisions may not retrieved the best results – whether by design limitations, or by decisions made by the specific organization which installed it. In this paper an AGV system is modelled using the SIMIO simulation software, taken advantage of its intelligent object, Vehicle. A decision taken by top management to prioritize two particular routes was assessed, which was found to be inducing a lower than expected AGV system performance. A new prioritization solution is suggested.