How do you choose the RFID tool management system? Functions and ways of identifying them?
Choose an RFID tool management system based on three things: the number of tools, the operating environment, and how important real-time location is. Passive UHF tags and fixed readers are enough for conventional discrete manufacturing and are inexpensive and quick to deploy. For more than 1,000 items where real-time location is needed, consider active RFID or UWB. Always test actual tools on site to check read rates and interference from metal; this step cannot be skipped.
First, three questions, then a comparison.
Three questions first go to the programme, which can filter out half of the unsuited manufacturers.
First, what is the total number of tools.For fewer than 200 fixtures and jigs, a spreadsheet and handheld terminal can be sufficient. RFID may not be worth the cost. Passive UHF is a good fit for 200–2,000 items, with tags at RMB 1–2 apiece and manageable costs for fixed readers. Above 2,000 items, consider splitting inventory across storage areas or installing multiple readers to avoid bottlenecks.
Secondly, what is the use of the environment.The metal tools are intensive workshops, oil-contaminated machine-plus areas and the proximity of highly electromagnetic interference transformer cabinets.
Third, do you need "identify" or "positioned"?Knows whether the tool is in a library or used, it's called identification; knows where the tool is at the moment, it's called positioning. The cost of the two is many times different, and you want to choose again so that you can avoid being biased by the program.
Identification: Passive UF is the mainstream, other as required
The vast majority of the current tool management projects are active with passive UHF (unsourced UHF) programmes, operating at frequencies ranging from 860 to 960 MHz, labels without batteries themselves and electric magnetic wave power activated by card readers.
Passive UHF advantageIt's a low unit price (a metal-resistant label is usually 2 to 5 yuan), long life (which can be ten years old), medium reading distance, and high mass reading efficiency. Fixed multi antenna readers can read dozens to hundreds of labels at a time, doing warehouse counts or access libraries very quickly.
It's shortboard.It's also clear that the metal surface will fall below 30% with a normal label, and it must be marked against metals (the type of suction material); the error rate in the magnetic environment will rise; only "in or out" can be identified.
If the business requires real-time positioning — for example, to know which position a wrench is working at this time — it needs active RFID (2.4 GHz or 433 MHz active tags) or UWB (superbroadband). Active RFID can read 30 to 100 metres, and can locate at the regional level; UWB can do centimetres with precision, but single label costs dozens of times, and full system deployment costs 5 to 10 times as much as UHF.
In practice, most customers choose a combination of UHF identification with manual inspections or fixed cameras to help locate assets. One investment addresses most of the problem.
The functional module is not as good as the scene.
The functions of the tool management system appear to be quite numerous and can actually be divided into three layers.
Base LayerThis level is mandatory, with all options and differences being smooth in interaction.
Operational level: Maintenance records, Validity Reminder, Loss of alarms, Waste Processes, docking with MES or ERP. This layer depends on the existing system - If an ERP or EAM already exists, prioritize the program to support a standard interface (RESTful, database middle table, WebService) and avoid data isolation.
ExtensionReal time positioning, electronic fencing, video connection, smart cabinets (lock-based and identification). This layer is budgeted to add, but not needed.
It is common to run the base and core business layers, then expand them in batches. Too big a one-time bunk can be more troublesome — systems are less volatile than they are without them, and staff will go around the system.
The interface is easily ignored. Ask the manufacturer three specific questions and sift out the wrong deal: what's going on with ERP? Is data synchronizing real time or time? Can historical data move in bulk?
A few of the ones that can easily step on.
Pipe I: Metal tools with general labels.The test is to be done with your field tools, and the recognition rate is below 95% direct pass. The metal labels vary widely in their glue and thickness, and the oil-resistant and heat-resistant models are twice as expensive but worth input in the workshop environment.
Pipe II: Ignores the location of the card reader.The fixed-reader is above the door, side by side is inside the cabinet, reading the same labels 20% less. Let the supplier produce a plan of installation, and go to the scene and re-examine the bridge, air conditioning the windways.
Pipe Three: Ignore the durability of labels.The tool shop environment is a test for labels: cutting fluid soak, friction and high temperature disinfection.
Pipe Four: Over-exploring real time positioning.Unless the tools are safe (e.g., if they leave a vehicle in the air) they'll be enough for regional positioning and manual review. UWB has seen many projects that have stopped, mostly with input and scenes not matching.
This is more important than the bidding.
No paper or electronic solutions are convincing on the ground. It is suggested that 5 to 10 representative tools (metal-containing, oil-contaminated, different sizes) be taken to the vendor's sample room or to its own workshop for three sets of tests: single label reading distance, batch recognition accuracy, and missing reading in metal-intensive environments.
The test environment is consistent with the actual use.
As a long-term service provider, Shanghai Select Tech Ltd.'s project delivery process usually involves two to three days of measuring links: first one or two tools to validate technical feasibility, then a whole series of tools to simulate the flow of access to the library at peak, and finally to sign the business terms. This rhythm avoids delivery before it finds that the recognition rate is not up to standard and returns to work.
Summarize the selection order.
First choose the identification technology (90% of projects use passive UHF). Then list features: essentials and core workflows first, extras according to budget. Screen suppliers for integration capabilities, references, and responsiveness after purchase. Finally, take the equipment on site and test it. Reversing the order makes it easy to be swayed by a demo, and it costs more to switch after spending money.

Ms. Lei
Mr. Pang