Abstract
This paper studies a three-dimensional automated storage and retrieval system that uses conveyors for depth transportation and a crane for both horizontal and vertical transportation. While previous studies put input and output points together at the system bottom, we consider separating and elevating them by proposing four I/O point policies: input and output points at two sides, input and output points elevated at one side, input and output points elevated at two sides, input and output points elevated at a midpoint. We build travel time models to analyse these policies and use simulation for validation. We theoretically find the optimal system structure and compare these I/O point policies, in terms of expected S/R machine travel time. The results show that the square-in-time structure is optimal in most cases and the mid-point elevation I/O point policy performs the best in terms of expected travel time. Moreover, we compare our I/O point policies with the lower mid-point policy. The results show that the mid-point elevation I/O point policy outperforms the I/O point at the lower mid-point by about 6.07% in terms of expected travel time, while other I/O point policies are dominated by the I/O point at the lower mid-point.