How does RFID asset management work? Equipment selection and data acquisition
An RFID asset management rollout hinges on four tasks: identifying asset categories, selecting suitable tags and readers, designing the data collection workflow, and integrating system interfaces. Hardware selection accounts for only about 30% of the work; the remaining 70% is workflow design and integration. Too many projects fail because they buy equipment before defining the process.
Point one: Clear asset classes and management of particle size before implementation
The more expensive the RFID is, the more expensive it is, the more important it is to match your asset scene. A sophisticated manufacturing plant and a storage complex have completely different claims -- the former has to run hundreds of thousands of units, the latter has to reach the shelf or tray level.
So, the question is three questions: whether to deal with fixed assets or low-value, easy-to-use items? What level do you need to track -- single stations, components, storage space? The count frequency is monthly, quarterly, or as required? The answer to these questions directly determines the subsequent equipment selection and acquisition process.
Element two: Label selection revolves around three dimensions - frequency, environment, reading and writing distance
RFID tags fall into three main frequency bands. LF (125–134 kHz) penetrates well but generally reads no farther than 10 cm, suiting applications involving metal or liquids. HF (13.56 MHz) reads within 1 m and is mostly used for access control and ticketing. UHF (860–960 MHz) is the main choice for asset counts, with 3–10 m range and rapid bulk reads of hundreds of tags.
The selection depends on the characteristics of the asset itself:
- Metal surface: Anti-metallic labels must be selected, and general tags will reduce or even fail to read the reading distance significantly.
- Outdoor or wet environment: labeled shell ABS (engineered plastic) or PPS (polythalaid sulphate-resistant high temperature plastics) avoiding paper labels.
- Long-term read and write: Select a label that can read and write more than 100,000 times.
- Low value and one-time use of assets: paper inlay (an antenna on a layered core) label costs as low as a few cents and suitable for large volume labelling.
When designing a solution, we usually test on a few representative assets first, such as metal cabinets, equipment with plastic housings, and cable trays. We apply the candidate tags and measure read success at 2, 5, and 8 m. Field data are more reliable than any spec sheet.
Point three: reader.
Readers are mainly fixed and handheld. Fixed-type readers are placed at the entrance or through the entrance, in conjunction with antennas to form an identification area, and assets are automatically collected as they pass through, suitable for the factory workshop, room access, data centre cabinet channel. Handheld reading is used for a desk check, and inspectors can read all labels within the range of lots by taking the equipment around the warehouse.
Several parameters to be noted in the selection:
- Fixed-readers look at the number and angle of antennas, with limited single antenna coverage, and route-type scenarios usually require 2-4 antennas to form an identification area.
- Handheld readers will see the fusion rate and re-routing, and equipment with a group reading rate below 50 per second will be very slow in large warehouses.
- Communications interface: Fixed general Internet access or Wi-Fi, handheld with Wi-Fi or 4G to facilitate real-time data retrieval.
- Agreement standard: Prioritize equipment supporting the EPC Class 1 Gen 2 (ISO 18000-63), which is the main agreement of the UHF RFID and is best compatible.
Both types of equipment are usually used in combination: a fixed-form system for automated entry and exit records, a handheld system for periodic inventory counts and unusual check-ups.
Point four: Is the process design system for data acquisition working?
Hardware is only the first step, and the real difficulty is to collect a process. A good set of processes usually consists of four elements:
The first step is to initialize the label.Each label is coded with the only EPC (electronic product code, corresponding to the tag ID number), and it is suggested that the rules for coding be harmonized, such as those of the previous indicates asset class, the middle indicates year of purchase, and finally the flow number, which will facilitate subsequent classification.
The second step is asset sequestration.Links the labeled EPC code with the underlying information of the asset (name, specification, location, value). This step suggests that the machine should be supported by a desktop haircard to avoid manual entry errors.
The third step is the initial count.The initial inventory data accuracy directly determines the credibility of the follow-up system.
The fourth step is daily collection.There are two scenarios: a fixed-reader automatically records the movement of assets and exits, with handheld for periodic counts. Data collected is time stamped and position information, or subsequent retroactive information will be difficult.
Point five: System interfaces begin with a protocol of the data interface and code rules.
Many projects are being implemented half-calculated to the system, and problems are often less detailed than those agreed in the previous period.
- Data format: The original data collected by RFID is the label EPC, which requires mapping to the enterprise's internal asset code and a table of comparison between the two sets.
- Interface protocol: Fixed-style readers typically push data to the middle with HTTP/MQTT (two common network data transfer protocols), and handhelds can be imported through SDK (Software Development Toolkit) or files. Intermediates are interfaced with ERP, EAM systems.
- Unreserved labels, unread assets that are not found in the system, multiple labels but only one object - these cases need to be defined in advance in the process.
- Data storage: The original logs recommended for RFID collection are retained for at least one year to facilitate retroactive and audit purposes.
When Shanghai Picket Technologies Ltd. does this, it usually takes a week or two to run through the smallest closed ring -- from stickers, issue cards, and collect into systems libraries -- to confirm that data flows are smoothing and spreading on a large scale. The rush to open up in one-off times tends to be concentrated, but slows down.
Point six: Several common pits avoided earlier.
Several issues warranting early attention in the implementation of RFID asset management:
Interference of metals and liquids.If there are large metal shelves or liquid containers, the label selection and antenna location will be specially processed and field surveys will be conducted if necessary.
Tab pasting position is not uniform.The label of the same asset should preferably be placed in a fixed position, for example, by placing it at the upper right corner of the shell or next to the signboard. Otherwise, the error of the count will affect reading distance and success rates.
The government has been over-reliant on RFID and neglecting manual review.Technical means cannot be a complete substitute for the system and it is recommended that manual signatures or photo-reviews be retained at key points as the basis for an audit.
No training to go online.The training is included in the implementation plan and at least a round of live exercise is arranged.
In summary, the RFID asset management path is to take an inventory of assets - the selected label - the selected reader - and design the pick-up process - to circumvent common problems. Each step suggests that it be small and large, first measuring the back-slip, and avoiding the cost of returning work from one-time inputs. The first period takes another week or two to do a selection test and process combing, which will save several months of adjustment.

Ms. Lei
Mr. Pang