How does RFID-based asset management work?
RFID asset management typically takes five steps: assess the current inventory, select tags and readers, deploy and tune equipment on site, integrate the software platform, and operate and improve the system. RFID makes sense above roughly 500 assets, when counts are needed at least monthly and bulk identification or location tracking is required. Below that scale, QR codes and a mobile app may be better value. No one design works everywhere, and a mistake at any step often means starting over.
Status inventory: Base was not finished, all back to work.
Many projects run into trouble because this step was not done properly. Before the count, review three things: how assets are classified by value, shape, material, and frequency of use; whether the site has metal shelving, liquid containers, or strong electromagnetic interference; and where the current count process falls short. A factory comparison gives a sense of the difference: a manual count of 500 assets usually takes two people a full day. RFID can cut that to about 30 minutes, an order-of-magnitude improvement.
The RFID signal is strongly reflected in metals, the UHF (UHF, common work band 840-960 MHz) signals are absorbed by liquids, and these two types of assets are returned to work again and again if they are not identified earlier. A small-scale test is recommended: 10-20 different labels are placed on typical assets, actually read at the planned deployment location, recording pass rates and drop rates.
Type: labels and equipment should not be price-only
The main choice is the UHF non-active label. In mainland China, the available frequency range is 920-925 MHz, with a maximum equivalent radiation power of no more than 2W ERP (effective radiopower, which is the rigid limit of the Ministry of Industry and Communications to the UHF band). The passive tag does not have an built-in battery, it costs less (commonly 0.5-3 yuan one), it has a long life span (more than 10 years) and it is usually 1-8 m to read the distance, suitable for most asset management scenarios.
Active tags have built-in batteries and a read range of 30-100 m. They cost several tens of yuan or more, last 3-5 years, and require periodic battery replacement. They are suitable for large equipment, long-range identification, or applications that need real-time location tracking.
Metal assets must be labelled as resistant to metals, which are barely readable on the metal surface. The anti-metallic label usually contains a suction material of 2-4 mm thick at a cost of $3-8. Liquid assets do not have a specific label, and read and take out loss by optimizing their deployment position and antenna angles.
Readers are in fixed and handheld categories. Fixed readers are typically equipped with 4-8 antennas, deployed at the entrance to and exit, access channels, warehouse doors, and do unmanned auto-identification. Handheld palletators are used for periodic count and unusual sorting, working in a similar pattern as swipe gun plus RFID reading and writing modules. An easy principle of an antenna selection: round polarized antenna does not require labels to be directed, tolerance for installation is high; line polarized antennas are read more distant, but labels are required to go further.
Deployment debugging: more complex than expected
The core of deployment is coverage and reading rates, not the number of equipment. The common error zone is a large pile of sensors and readers at first sight, which results in high cost, but rather heavy read-through.
The standard configuration of the entrance channel is a fixed reader with 2-4 antennas covering 3-5 metres in width. A master view inside the warehouse is more suitable for handheld, scanned by series. A peri-printer at desktop level (reading and writing distance 10-30 cm) is used for asset registration stations to manage single pieces.
The label is marked in terms of location. Metal asset tags leave the metal surface at least 3-5 mm, with a middle cushion of cotton or a direct anti-metal label with a back-of-the-back tape. Plastics, wood assets are attached directly. The sticker position is selected where it is not easy to be touched, and where it is difficult to spray liquid, such as a handle on a work-pack, a bottom or back of office equipment.
The debugging phase involves several things: power from low to high, gradually finding critical points without a string reading; screen tests to confirm that adjacent areas are not misread; speed tests to simulate the speed of movement when actually taking an inventory, and to confirm that read is stable. Shanghai-based Quest Technologies has seen debugging cycles in multiple plant projects that are two or three times as expected -- often at the top of the power range, which results in inter-tralinking, and then re-reverses.
Software integration: last kilometre of landing
The software layer is three blocks: the middle, the asset management platform, the top operating system docking. The middle handles data reported in a reader, filters over and over again, makes collision-proof processing (more than one label is also queued to read when responding to a data confusion), and caches anomalies. This is the data hub of the entire system, which requires the highest stability.
The asset management platform is a core functional vehicle, including labelling, asset files, location tracking, inventory statements, and abnormal alarms.
The interface with ERP (Enterprise Resource Plan) is the last kilometre to land. There are three common docking methods: API real-time synchronization (suitable for a high-real-time scenario), intermediate library step-synchronous (suited for batch data exchange), file interface (suited for report type data exchange).
Tragedy: The real test of the project
An RFID project is not over at handover. Operations and maintenance are the real test. Typical tasks include monitoring tag failure rates by sampling 5%–10% of tags monthly and promptly replacing those with falling read rates; inspecting readers and antennas for loose connectors, misalignment, and damage to the feedline connecting reader and antenna; and reconciling platform data to make sure RFID counts match the financial records.
The first year of the process, the second year of digging up data value -- for example, optimizing the procurement plan through asset idleness analysis, optimizing the layout of the warehouse by flow-path analysis, and rechecking the gaps in management by taking an inventory of anomalies. These are the RFID’s greatest advantage over two dimensions: automatic collection, process trailing.
Returning to the conclusion at the outset, RFID asset management has no shortcuts but clear paths. The project can run steadily by mapping the current situation, matching options in the medium term and keeping pace with events later.

Ms. Lei
Mr. Pang