Simulation study of e-commerce robotic warehouse systems

Abstract

As an essential component of any logistics network, the warehouse plays a vital role in realizing fast and reliable E-commerce logistics services. Moreover, order picking is a necessary process that will impact the warehouse’s performance and logistics service quality directly. The Amazon claimed that the robot-based-order picking system is flexible, economical, and efficient, the industry is keen to introduce this new solution to their E-commerce warehouse, but they found there are few theories to refer. We used Python, a scripting language often used for algorithm implementation and scientific calculations, to simulate the e-commerce warehousing system. We chose the method of simulation because the operation mode of the warehouse operation system is relatively complex, so it is difficult to extract a single mathematical model to observe all the changes. Individual factors, such as warehouse layout design, affect all other elements of warehousing and decisions. For such a high complexity system, only the simulation can realize the "what-if" demonstration, so that we can have a more comprehensive understanding of the advantages and disadvantages of the robot warehouse. We also introduce a matrix-based clustering algorithm for order allocation in the robot-based order picking system and verifies the algorithm efficiency by a simulation example. From the results, we can see that as long as the order size is large, the operation efficiency of the robot warehouse is much higher than that of the traditional warehouse. In the most extensive order simulation, the robot warehouse completed the order sorting 35 hours earlier than the traditional warehouse. Therefore, the application of the robot warehouse plays a significant role in improving the efficiency of e-commerce logistics. However, the energy consumption of the robot warehouse is also massive, so we optimize the operation of the robot warehouse from different aspects in chapter 6 and chapter 7. First, order clustering was used to reduce the running times of the robot. The results showed that the optimization effect was very significant, with the maximum optimization ratio reaching 50%. Another strategy was introduced to realize the synchronous sorting of the robot warehouse. A mathematical model of synchronization was established, and an index of synchronization was given. The results show that the specification of synchronization not only does not affect the on-time sorting of orders but also can significantly reduce the time occupied by the order sorting table. This work should be the first comprehensive analysis of the advantages and disadvantages of traditional and robotic warehouses. This thesis gave theoretical guidance to enterprises that want to upgrade from a traditional warehouse to a robot warehouse. Moreover, the operation of the robot warehouse had been optimized, both by the order clustering optimization and synchronous optimization, which had achieved a good improvement effect.