Architecture Design of a Smart Logistics Warehouse Management System based on Multi-modal AI

Abstract

Addressing the inefficiency and poor real-time performance of traditional warehouse management systems, this paper proposes a design scheme for a smart warehouse management system based on multi-modal AI. The system achieves intelligent management of the entire logistics process through the collaborative decision-making of multi-modal AI (DeepSeek-R1 language model and GLM-4V visual model). Combining a modular micro-services architecture (SpringCloud) and distributed technology, it constructs a closed-loop system of "perception-decision-monitoring." The design integrates two-factor authentication and edge-cloud collaborative computing, supporting dynamic inventory optimization and real-time data analysis, providing a lightweight and scalable solution for the intelligent upgrade of logistics warehousing. Future work will further explore the deep integration application of digital twin technology.