SPATIAL-TEMPORAL DATA-DRIVEN CYBER-PHYSICAL SYSTEM FOR WAREHOUSE OPERATIONS

ABSTRACT

As logistics increasingly play a significant role in business competition, it necessitates more efficient warehouse operations, which deal with the location of man, machine, and material in many instances. If where each object is located can be captured timely, the efficiency of the operational conduction will be considerably enhanced. In my study, a spatial-temporal data-driven cyber-physical system (STDCPS) for warehouse operations is established to achieve indoor localization and apply the real-time location information to warehousing. Concerning different moving patterns and actual needs, particular approaches are implemented to position each sort of object. The deep learning method and SWIMMING algorithm are devised to locate massive products and moving workers, respectively, both based on Bluetooth Low Energy (BLE) on account of its low cost, low energy consumption, and compatibility with smartphone. And UltraWideband (UWB) is harnessed to detect the material handling equipment that moves fast considering accuracy and responsiveness. The application of IoT technology enables the self-learning of the positioning models and alleviated offline signal map training. Besides, the correlated and traffic balanced storage assignment policy (C&TBSA) and diverse order picking plans for pickers are adopted to reduce travel time and enlarge the overall throughput. In the presence of the cyber-physical system, the storage allocation can be adjusted in due course pursuant to the correlation between goods stemmed from the updated customer orders, and each worker’s picking route can also shift dynamically in light of real-time conditions sensed by the system. A case study of a textile and apparel manufacturer’s warehouse is carried out to demonstrate the system and validate its feasibility in practice.