An Evaluation of Shuttle Selection and Storage Assignment for AS/RS

• Main Authors: Tsai-Pei Liu, 劉彩霈
• Other Authors: Shin-Ming Guo
• Format: Others
• Language: zh-TW
• Published: 2011
• Online Access: http://ndltd.ncl.edu.tw/handle/01619087516574684675

博士 === 國立高雄第一科技大學 === 管理研究所 === 99 === An Automated Storage and Retrieval System (AS/RS) operates a shuttle with computer process control to execute the storage and retrieval of warehouse stock. It can be equipped with the traditional single shuttle or the newly developed twin shuttle. There has been a significant literature on the comparison of single shuttle systems versus twin shuttle systems. Researchers agree that twin shuttle systems have better operational performance. However, a twin shuttle system costs about 40% more than a comparable single shuttle system. Therefore, given the same operating environment, we believe that the capacity or effectiveness of a twin shuttle system should outperform a single shuttle system by at least 40% to justify the cost. This study has interviewed equipment vendors and users in Taiwan to undertand the operating requirements and managerial concerns. Then we develop Markov processes and use a system of state equations to describe the dynamic behavior of single/twin shuttle S/R machines performing storage and retrieval transactions in a single aisle. Processing times are mixtures of exponential distributions or mixtures of Erlang distributions. We find that the twin shuttle is better but does not outperform the single shuttle significantly. The study also develops Arena 12.0 simulation programs to compare single shuttle and twin shuttle systems under various operating environments. The comparison is based on average storage/retrieval waiting time, the effective storage/retrieval capacity, and the ratio of dual command. We assume different time periods of storage and retrieval operations may or may not overlap, input and output stations may not be in the same location, and there may be batch demands for storage and retrieval operations. Simulation results of single-aisle suggest that twin shuttle systems would be the better investment if the input/output stations are in separate locations, time periods of storage/retrieval operations do not overlap, or batch demands are significant. Simulation results of multi-aisle suggest that twin shuttle systems are more cost-effective when the dedicated aisle rule is implemented. If I/Os are located at the same side and S/R periods overlaps, the class-based storage rule will significantly improve the relative performance of twin shuttle systems. We believe that users will need more sophiscated systems and techniques in the future. Our results based on the cost performance ratio can provide useful information to practitioners planning to implement AS/RSs in real world applications.