Stochastic modeling of conveyor systems applied to the chain conveyor at Hollander Barendrecht

The chain conveyor in the distribution center of logistic service provider Hollander stagnates many times each day. In this thesis we investigate if the continuity of the chain conveyor could be improved, using an operational approach. With regard to this improvement, we model the chain conveyor as a queueing network. The queueing model is a general distributed, closed, finite capacity, multi-class queueing network, consisting of FCFS disciplined finite capacity nodes with either transfer or recirculation blocking, allowing head-of-line-priorities and multiple server nodes. We propose the MVABLO-m algorithm, in order to approximate the mean number of jobs for each node in the network. This algorithm is based on the MVABLO algorithm. We conclude that for multiple server node networks, deviations can raise up to 30%, and for multi-class networks with two job classes, deviations can raise up to 80%. Therefore, we conclude that the MVABLO-m algorithm is not suitable for modeling the chain conveyor. Finally, we incorporate the queueing model in an integer program, and solve this integer program using simulation. For all tested scenarios, we conclude that stagnation can be reduced using implementation of proposed loading strategies. No drastic negative side effects of this implementation are expected.