How do you use a hand-held RFID count? Move the inventory workflow.
Many people think that holding a RFID count is just opening the reader, and scrutinizing the shelf. In fact, it really takes time to prepare for and process an anomaly, and scanning actions are only the last steps of implementation. The complete mobile inventory process includes equipment and data preparation, field environment confirmation, tag reading, data return and reconciliation, and this is the five steps of the unusual processing. Either step is missing, and the final account may not be available.
Let's get to the point where we're holding a RFID.
Handheld RFID readers (also called mobile terminals, PDAs) are essentially a device that can send radio frequency signals. It sends an query signal to the surrounding space through antennas, and when it gets a signal on a commodity or shelf, the RFID label is sent back a data, and the reader reads out the data. The whole process is non-contact, and it can read dozens or even hundreds of labels at once, which is why it counts faster than a barboard.
But "readable" and "reads correctly." Handheld RFID is usually read at a distance of between 1 and 8 metres, depending on the type of label, the reading and writing power, environmental disturbance. Metal shelves, liquid packaging, the wrong angle of labeling, make it impossible to read or misread the label that you could read, or misread the next to it. That's why handheld points can't simply "go through."
We need to get the equipment and data ready before we go into business.
Several things must be done before the inventory is officially started, otherwise it will be stuck at the scene.
In terms of equipment: Reader and writer power is full, Wi-Fi or SIM cards are connected, and a self-check is performed if necessary. The RFID module is hot for long periods of time, and the performance will decline. This is more evident in summer in poorly ventilated warehouses. If the inventory is large, it is expected to sweep over several hours in succession, and it is recommended that the backup be prepared, or the device cooled down midway.
Data: Download this inventory tasklist from the back-office system before you go to work. The task list usually includes a repository number, the planned inventory of SKU lists, which corresponds to the scope of the EPC (electronic product code, the only identity in the RFID label). After downloading to the handheld terminal, the operator can see where he is responsible and sweep what range of goods. The task list is not right, when it is found at the scene that the data is not in the storage area for which he is responsible, it is dry.
Let's go to the scene and see if we can get this thing done.
Take a few minutes to observe the environment and save much time behind.
There are a few things that you can see: shelf density, metal ratio, liquid-type commodities, temperature and humidity. A dense metal shelf allows radio-frequency signals to reflect superheavy, the same label may be read over and over again (also called "reread" or "phantom reading," which is not easy to read. The liquid-rich area has a marked decline in signal, and the labels that could have been read could have been read very close to it.
The RFID reads the effects of environmental impacts, and it is common practice to: first scanning a small area, see whether power slots and reading modes need adjustment. The general handheld terminals have high power, medium, low, three-level, near distance, mid-field, far away. Instead, they are more difficult in their environment, with lower power in the immediate range, and more stable reading. This experience is more combative, and Shanghai-based Technologies Ltd. usually conducts an on-site environmental assessment at the beginning of the project, which will lead to a power, skyline, reading mode parameters that are set ahead and avoid frequent detorting during operations.
Routes and techniques during scanning
When the actual scan is performed, route planning and techniques have a significant impact on efficiency. A common error is that operators move in the warehouse in an instinctive manner, where they sweep, resulting in a leak or double-cleaning.
A standard inventory workflow follows bin numbers in order, starting at Rack 01 in Area A, for example, and proceeding row by row. Pause for three to five seconds at each location so the reader has time to read all tags there. RFID reading is continuous, not instantaneous. Allow enough time to capture every tag.
When hand-held, the antenna is critical. The RFID label is not as sensitive to the reading angles according to how the inner antenna is polarized, and it is basically readable with a non-clanted angle; the RFID label is sensitive to the angle, and the antenna is not readable. When actually working in the warehouse, it is recommended that the reader and the shelf be given a relatively fixed angle, for example, about 45 degrees tilt, then move slowly along the shelf.
We also found a recommended approach in practice: first, to do a blind sweep without relying on the information of the library, and then to scan the entire region, looking at the number and distribution of actually read labels, and then matching them with back-office data. This step quickly reveals which collections are problematic (label damage, masked, misplaced), so that the focus can follow.
Data returns and reconciliations
Once the scan is completed, data needs to be sent back from the handheld terminal to the back-office system. Common ways are Wi-Fi, 4G/5G mobile network, or data line import, depending on the conditions of the field network and the size of the data.
The data returned are usually: read the list of the EPC, read the time stamp, read and write, and the operator number.
- Compare the SKUs with the task orders, see which SKUs read and which did not;
- (b) To reprocess and remove data from the same label that has been read over and over again;
- (a) Tag abnormal labels, such as signal strength below threshold and too few readings;
- The variance report was generated, showing the number of items on the books that were not found at the site and were read more frequently on the scene.
This step is the "outcome" of the whole inventory process, which does not determine the credibility of the inventory report. Once the discrepancy report is made, the operator returns to the scene and double-checks or cleans out the abnormal labels.
Sweeping and processing of abnormal labels
Once the data is returned, it almost certainly happens. There are three common types of unusual:
- The labels are probably blocked, damaged or placed in a position that is too biased. The operator needs to take the handler and clean it up at close range to the specific storage area.
- Read multiple: The site was swept but the books were not. It may be that the labels of adjacent storage sites were read in a string or that there were temporarily placed goods that were not registered in the system.
- Incomplete reading: The label can read but the signal is weak, and it has been read ten times only twice or three times. This label is usually not far from being valid, so you can use this inventory to record it and arrange for a replacement.
Once the anomaly has been processed, updated data are returned and reconciliations are not completed until the discrepancy is within acceptable range (e.g. less than 0.5 per cent).
A few easy details to ignore.
One is not to stay within 0.5 metres of the metal shelf. When handhelds are close to large areas of metals, antenna performance is affected and the data cleared down.
The number of people working in the area at the same time is recommended to be no more than one reader.
The third is not to ignore antenna maintenance. If the RFID reader's antenna is covered with sweat and dust, it will read differently.
The core idea of holding the RFID count is one thing: the more elaborate the preparation in the first period, the greater the error space for field scanning; the more regular it is to sweep the scene, the less unusual it is for subsequent reconciliations. It's not a "button button" job, but a process that needs to be followed. Set every step, and the data that eventually gets the count report is really working.

Ms. Lei
Mr. Pang