Performance evaluation for Robotic Mobile Fulfillment Systems with time-varying arrivals

Abstract

This paper analyzes the performance of Robotic Mobile Fulfillment Systems with a semi-open queueing network (SOQN) with time-varying arrivals. We use a Markov-modulated Poisson process to characterize the variability in the order arrival rate and offer a matrix-geometric method (MGM)-based approach to solve the SOQN. Simulations are used to validate the analytical model. We verify that assuming time-homogeneous order arrivals will underestimate order throughput time. Numerical experiments are also conducted to explore the effect of peak order arrival rates, peak order arrival durations, and average order arrival rates on system performance. The results show that peak order arrival rates and peak order arrival durations have the most significant impact on order throughput time. Moreover, we compare the order throughput time of various workstation layouts with different width-to-length ratios.