Semantic Digital Twins for Retail Logistics

摘要

Abstract As digitization advances, stationary retail is increasingly enabled to develop novel retail services aiming at enhancing efficiency of business processes ranging from in-store logistics to customer shopping experiences. In contrast to online stores, stationary retail digitization demands for an integration of various data like location information, product information, or semantic information in order to offer services such as customer shopping assistance, product placement recommendations, or robotic store assistance. We introduce the semantic Digital Twin (semDT) as a semantically enhanced virtual representation of a retail store environment, connecting a symbolic knowledge base with a scene graph. The ontology-based symbolic knowledge base incorporates various interchangeable knowledge sources, allowing for complex reasoning tasks that enhance daily processes in retail business. The scene graph provides a realistic 3D model of the store, which is enhanced with semantic information about the store, its shelf layout, and contained products. Thereby, the semDT knowledge base can be reasoned about and visualized and simulated in applications from web to robot systems. The semDT is demonstrated in three use cases showcasing disparate platforms interacting with the semDT: Optimization of product replenishment; customer support using AR applications; retail store visualization, and simulation in a virtual environment.