ABSTARCT : Accurate and timely stock control is one of the more persistent operational problems in small-scale manufacturing and distribution. Most small warehouses cannot justify the cost of enterprise inventory platforms, yet manual and barcode-based alternatives introduce systematic errors that distort purchasing and replenishment decisions. This work describes an RFID-based inventory management system built around an ESP32 microcontroller, an MFRC522 passive RFID reader, Google Apps Script as a cloud middleware layer, Google Sheets as a zero-cost database, and AppSheet for mobile dashboard access. A closed-loop tag reuse model is central to the design: rather than treating RFID tags as disposable, tags are returned with each delivery and reassigned in software, keeping ongoing hardware costs near zero. The system couples live RFID scan data with classical inventory methods—Economic Order Quantity, Reorder Level, Safety Stock, and ABC Analysis—so that replenishment decisions are driven by actual consumption rather than estimates. A three-month moving average provides a practical demand forecast for dynamic reorder-point computation. Testing confirmed end-to-end latency of 2–3 seconds from tag detection to dashboard update, 100% identification accuracy within the effective read range, and complete elimination of the duplicate-decrement problem through firmware-level debouncing. Total hardware cost is approximately USD 15–20, with no software licensing fees.
Praveena et al. (Fri,) studied this question.