Autonomous Edge Intelligence for Real-Time Inventory Management an Adaptive Zone-Aware IoT Mesh Framework

Modern intelligent factories are struggling to achieve real time visibility of the inventory because of the latency, flaws in the human intervention, and bottlenecks in the verticality of the traditional systems that encompass RFID-ERP integrations, barcode-MES models and manual periodical stock valuations. Such constraints cause poor decision making, poor stock control and insufficiency in regard to responsiveness to changing production requirements. To meet the issues described above, though, the application of Adaptive Zone-Aware IoT Mesh equipped with Autonomous Inventory Decision Agents (AZA-IMAD) can be proposed on this paper; it is a full edge-centric inventory control framework. AZA-IMAD implements the IoT sensor nodes in a truly dynamic form of zoning that are governed with autonomous decision agents that have the capability to perform real time sensor fusion, anomaly detection as well as reconstruction of replenishment at the edge itself. A zone-aware communication protocol will further ensure the vital changes will be transmitted to the network to reduce network congestion but a lightweight blockchain ledger will provide auditing to diverse changes that are decentralized as well as tamper-resistant and will not need web connectivity on an ongoing basis. The overall evaluation in a virtual smart manufacturing arrangement reveals AZA-IMAD achieves an desired inventory monitoring of an 91.75 that is above the conventional process of 65% - 85%. Also, specifically, the framework recognizes stock change at a rate of 93,90 and induces timely replenishment decision at 91 of the stock at 93% accuracy. In addition, AZA-IMAD saves on 38 percent of bandwidth in addition to activation response of decisions in milliseconds since the centralized processing bypasses are eliminated. AZA-IMAD introduces two core innovations: (1) a dynamic zoning mechanism that adapts sensor clusters based on real-time inventory density and flow, and (2) autonomous decision agents (AIDAs) operating at the network edge, enabling immediate anomaly detection and replenishment without central dependence. These innovations jointly deliver over 10% higher accuracy and 38% bandwidth savings compared to RFID-ERP systems.

Ҳали таржима қилинмаган