Research question
Can unit-level tracking make recurring inventory variance traceable?
For four consecutive months, the operator's stock count showed an unexplained shortage of $250 to $500. Repeated manual counts did not identify whether the variance came from counting, receiving or unrecorded movement.
The operational problem was therefore larger than the loss itself: reorder and inventory decisions depended on a stock figure the owner could not verify.
Working hypothesis: identifying every unit and recording its movement would replace disputed aggregate counts with a traceable inventory history.
Intervention
Direct UHF RFID integration inside the operator's workflow
UHF RFID was selected for its read range and ability to identify multiple tagged units in one pass. Before implementation, tag economics were assessed for the Egyptian market to verify that unit-level tagging was proportionate to the inventory value.
- 01Hardware selection
Specify the UR-3ED4 reader for bulk shelf and stockroom reading.
- 02Protocol integration
Map the reader protocol directly over BLE at the GATT layer because no usable platform SDK was available.
- 03Mobile workflow
Build a Flutter application that surfaces variance against expected stock during the count.
- 04Unit accountability
Assign each item an identity and record movement time and handler without vendor middleware or manual re-entry.
Observed data
Inventory shortage result
4 consecutive months
$250–500 unexplained shortage per month
$0 unexplained shortage from month one
Variance could be reconciled against recorded unit movements
The movement data exposed a specific internal cause. The owner identified and resolved it without requiring the cause to be disclosed publicly.
Interpretation
Traceability resolved both the shortage and the uncertainty.
The system did more than accelerate counting. It changed the available evidence from an aggregate discrepancy to a unit-level movement record, allowing the operator to investigate variance and act on a defined cause.
RFID economics remain category-dependent. The value of the protected inventory must justify the tag, hardware and implementation cost.
This is an anonymized operational before-and-after result, not a controlled experiment. Catalogue size, location count and the underlying internal cause are intentionally not asserted.
Need reliable inventory data across shelves, stockrooms or field operations?
Discuss the workflow