Deploying Voice Interfaces for Real-Time Retail Inventory Management

Real-time inventory operations are central to agile retail management, but existing systems still face synchronization issues and human-driven errors. To overcome these limitations, this paper presents CAMVIS-RIM (Context-Aware Multi-Modal Voice Interaction System for Real-Time Retail Inventory Management), a unified framework that integrates voice AI, IoT sensor data, and predictive analytics to enable proactive, context-sensitive inventory control. The proposed system fuses data from RFID, barcode scanners, and shelf weight sensors with a voice-based interface capable of understanding multi-role interactions (store associates, managers, customers). By leveraging edge computing for low-latency processing and AI-driven forecasting, CAMVIS-RIM ensures faster, more accurate inventory updates and intelligent stock recommendations. Experimental deployment in a mid-sized retail outlet achieved a significant improvement in accuracy (+7%), precision (+6%), and user satisfaction (+10%) compared to conventional systems. The system demonstrates scalable potential for future retail environments where context-aware, voice-driven analytics redefine inventory operations.

Ҳали таржима қилинмаган