How does the RFID smart-ass management system work?
RFID Knowledge

How does the RFID smart-ass management system work?

PublishedShanghai Zejie Smart Technology Co., Ltd.

The RFID smart cabinet is loaned and spot-on surveillance, the core of which is to have a high frequency reader real time to identify the movement of each label item in the cabinet. To make it really useful, you have to consider all three things: hardware selection, tag strategy, access flow, or otherwise not, the data count is not accurate, the system is detached. And then, by rewinding, what should be noted at every point?

Hardware: The cabinet is wrong. The recognition rate won't be restored.

We did a multi-plant RFID tool cabinet project, and the last count was not a good reading machine, but an unreasonable tectonic structure and antenna. The money and energy invested in this part are more important than choosing which brand reader.

A complete set of RFID smart cabinet hardware includes: UHF reader (generally ISO 18000-6C/ EPC C1G2 protocol, output power 0-33 dBm adjustable), near or flat antennas within the cabinet, 12V or 24V electric control locks (recommended to break out electrically, locks safer), identification module (IC card, second-handigrapher, fingerprints, face-to-faces) and working control master plate touch screen.

A few structurally easy points:

We met with a firm that made the exact mold, and the cabinet was put in two months without a bottom count, and then we took off the metal partition. The extra money spent back to buy two more cabinets. So the structural problems must be solved at the selection stage.

Loaning process: "Who, what time, what" all left behind.

The loaning process appears to be three steps – open, pick up and close – but the system must record all the operators, time, item code, reasons for lending, or follow-up is a bad idea.

  1. The user brushes the card or face of a door reader, automatically opens the lock after the system has verified permission and the screen is synchronized to show the person ' s loanable range.
  2. The system takes a snapshot of the list of items in the current cabinet as the baseline for this operation at the moment when the door is open.
  3. After closing the door, an inventory was completed within 3-5 seconds of the inside antenna and a new list of positions was generated.
  4. The system automatically compares the two lists and counts the mission collection — that is, the items borrowed — in the lending records, which includes the operator's number, the EPC code, the name of the item, the quantity, the time of loan, and the expected return time.
  5. In the case of valuable or controlled items, system jump-through approval flow is valid only after confirmation on line by the supervisor and no record of entry into to-do alarms is confirmed.

The problem with the operation is that "the door is open for a long time, and the count benchmark is not credible." It's suggested that the door be opened and that the high frequency (HF) count be activated every second, and then the last result of the closing be used as a reference. So that even if someone opens it for ten minutes, they can get back to you.

Return and on-site monitoring: let the system keep its own accounts

It's easier to return than to lend, because it's an "additional operation," and the system just needs to recognize what's more, and then mark the corresponding lending record as returned. The point is "automatic," so no operator can go to the "return" button -- just one thing will be forgotten.

After the door is closed, the system automatically counts every 15-30 seconds (rearable) and compares with "should have an inventory":

In order to make the identification more stable, usually by adding an antenna to the side of the cabinet, you should look at "the object is attached to the door and the main antenna cannot be reached." When the top and the side antenna work together, the count is up to 95% and above 99%. If this detail is not done, the rain is damp or the labels are slightly skewed, it's gone.

Permissions and anomalies: Don't let the system go offline for six months.

The RFID cabinet is the most easily ignored, but the most problematic part of the problem. It is based on a role classification: operators can only use the tools that are commonly used in this group, and team leaders can borrow all the tools in this workshop, supervisors can move over workshops, and high-value items require more than class leader privileges. Each user binds a permission group, local cache at the end of the cabinet, and the network can be verified.

Time limits are often missed. For example, night shifts prohibit the lending of precision measures worth more than $5,000, and only managers are allowed to open doors on weekends. Without writing death in the system, people can't catch up with it.

The irregularity must be pre-planned and cannot be remedied until something happens:

Rewinding projects for several years, three of the most easy places to catch up with.

The items were not marked or were not labelled.The labels are put into a few cents to a few dollars each, but the time saved for hand-counting is far greater than this cost.

Blindly pursuing the full range of functions.The first time you go to the RFID cabinet, you suggest that you do only "Leave + Return + On-Site" and then put it aside, on a statement, on an inventory plan, and on an inventory alert. When the process runs, you start to open new modules. The more than one training cost and the probability of error increases in a straight line, the operator is more willing to go back to the old path of writing and writing records.

No re-entry was left for operations.The system is long, and it's always "Leave something to pay but the loaner has left," "The good is dead, but the system shows a place." The design must be done by an administrator, or the system will become a data black hole for six months, and nobody dares to use it again.

If you're assessing whether to go to the RFID smart-box, we suggest that you try to test one of the most obvious cabinets in pain -- like the workshop locker where tools are loaned frequently and lost at a high rate. Run data for a month or two and decide to spread it all. From a single point pilot to full deployment, it usually takes 3-6 months, but not so much. The projects that Shanghai-based Scrubbery has been doing over the years on the management scene of manufacturing tools, which is basically a pace move forward, first to run out of a cabinet with a level of recognition, stability and user acceptance, then to replicate them horizontally.