What's good about using the RFID system for asset counts?
The RFID inventory is a mass non-contact recognition and long-range automatic reading, with a core benefit that is 5-10 times more efficient than manual scans. But three things need to be clearly thought out before you land: frequency selection, tag matching with asset material, system alignment. Either part of the link is wrong, often "not readable" or "not readable."
Frequency preset: HHF for mass, HF and NFC for single product management
RFID is divided into three main groups by frequency band: LF (near 125kHz), HF (13.56 MHz), UHF (860-960MHz). The three are reading distance, label costs, and the ability to penetrate is very different, not "what good" but what your assets are, how to manage them.
The UHF is the main option for current asset inventories, usually at 3-10 metres of distance from which dozens of labels are read in bulk with a direction antenna, and the fastest. The UHF tag unit price is now between 0.3 and 2 bets, the most expensive frequency range for large-scale asset inventory counts. But the UHF is sensitive to metals and liquids, metal surfaces need metal resistant labels, and the liquid environment is compensated by adjusting the antenna angle and power.
The HF (including NFC) reads between 10 cm and 1 m, the label is worth 0.5-5 pm, which is more appropriate for a "one-to-one precision" scenario, like tool loan, file management, precious small equipment. It has the advantage of antimetallic resistance, strong liquid resistance, and easier to fit labels on the side of the curve.
The LF reads the shortest distance, usually within 10 cm, and is common in animal labels and some industrial identification scenarios, which are not used much in asset inventories. A simple test for selection: a single count of more than 50 items, a "readable through" search, a UHF; an emphasis on precision to a single piece, reading a close view, or a choice between HF or NFC.
The label is wrong. The perfect reader is useless.
The most easily pedals in the asset inventory project are not errors in the reader, but rather labels and asset materials that do not match. RFID tags work with electromagnetic waves and antennas, where metals are shielded and liquid signals are absorbed.
Metal assets such as cabinets, servers, and equipment housings need on-metal tags. These use an isolation layer between a conventional tag and the metal surface and a specialized antenna design. They are typically 2–5 mm thick. An ordinary UHF tag on these assets would drop from a 5 m read range to less than 10 cm, making it effectively unusable.
Liquid environments (chemicals, liquid raw materials) are either labeled with HF or the antenna is attached to the bottom of the container and debugged in the UHF program. We have experienced a similar situation in the actual project: the original HHF tag was completely unreadable on the surface of the metal storage tank, and then replaced by a HF tag with handheld terminal, each can is labelled, and the Clerical Crew is more stable than the remote solution.
Paper assets such as files, documents, and manuals offer the most tag choices. Basic self-adhesive UHF tags are sufficient and can cost as little as CNY 0.20-0.30 each, making them suitable for large-scale tagging.
Batch reading is a core gain, but it also raises the question of "column read."
The RFID's greatest advantage is to read in bulk: a warehouse entrance with two UHF readers and antennas, the vehicle pushes through the shelf, hundreds of labels are all read in two or three seconds, and the data automatically write backstage. This efficiency is not possible by hand-cleaning or barcoded discs.
But there are side effects of mass reading: interference between adjacent labels, leading to certain labels not being read, or the misreading of asset information – often referred to as “crash” or “string reading.” The problem is solved mainly by impact-proof algorithms (such as Q-value adjustments in the EPC C1G2 protocol) and on-site antenna layouts. Lower the reader's power, narrow the antenna angle, and stretch the distance between labels.
Also, the "Left Read" mechanism should be set aside: 3-5% of the labels are always not read because of angles, masks and power problems. The system is designed to support automatic re-testing or manual clean-up, and cannot expect 100% accuracy at one time.
System interfacing: Don't let the inventory count and ERP run on each other
Many projects RFID hardware is accepted and accepted, but the business sector is reluctant to use it on-line because of poor data inventory and existing systems or the need for re-entry. Problems often arise in the integration chain.
The flow of inventory data is typically: RFID Reader and Writer RFID Medium Quest Business System (ERP, Asset Management System). The middle is responsible for filtering duplicate data, translating the EPC code into asset numbers, making data caches. If this layer is not working well, the business system receives thousands of duplicate records per day, which becomes a burden.
There are three main types of docking:
- Direct interlocking: the reader provides SDK, the business system is directly mobilized and suitable for a self-developed enterprise;
- (a) Intermediate docking: deployment of professional RFID intermediates, standardized processing data before being rolled over to the operational system and extended;
- API docking: Transfer data via RSTful or WebService interface, suitable for cloud-based deployments or scenarios where multiple systems coexist.
We generally recommend that clients use the API to run the flow process during the pilot phase, and then scale up before considering the intermediate, as deployment and transport of the intermediate will require dedicated responsibility.
Horizontal comparison of several mainstream programmes
| Programme | Single Read Distance | Batch capacity | Label Costs | Typical scene |
|---|---|---|---|---|
| UHF RFID | 3-10 meters. | John, hundreds at a time. | 0.3-2 dollars. | Warehouse, machine room, equipment inspection |
| HF RFID / NFC | 10 cm-1 m | Usually, a few readings at a time. | 0.5 to $5. | Archives, tools, valuable pieces |
| QR code / barcode | We need to be on a case-by-case basis. | None, dependent on manual | How much money? | Low-value, limited budget scenario |
| Bluetooth / UWB Positioning | 10 to 50 meters. | Strong. We can locate it in a specific position. | A few hundred to a hundred. | Real-time positioning of high-value equipment |
Up to down, precision and positioning are increasing, and costs are rising. The asset inventory is by nature a trade-off: UHF RFID is the most widely covered, with the lowest combined cost scenario, but if your asset is of high value, then it's more appropriate to know where it is in real time, and the UWB position is just one order of magnitude higher than the UHF for individual labels.
Let's try this before we land. Don't start all over the place.
Whatever the solution, start with a two- to four-week pilot for one warehouse or asset category. Verify three things: whether the read rate reaches 98%, whether tags read reliably in different positions, and whether the system fits existing processes. Buying and deploying at scale after a successful pilot substantially reduces rework costs.
RFID is not a technology that fits everything, asset material, and the way it's set up, the electromagnetic environment around it can affect the real effects. Put the expected management ahead of you, put the pilot in place, more important than which brand. Shanghai-based Technological has been trying to do this for the client, so that one time they don't get the expected results.

Ms. Lei
Mr. Pang