How does the RFID fixed asset management solution be designed?
The RFID fixed asset management program is designed to be the core of "Access, Labeling, Run, Processes" that goes down. This idea applies to organizations with more than 500 assets, a tight count cycle, and a much-mistaked manual inventory. The program works well, does key view data work, processes are not tired, and can you trace people out of the ordinary? Below is a reset of projects, untraceable.
First step: do a thorough manual inventory, and don't expect RFID to settle historical debt once and for all.
Many people are wrong about RFID, thinking that the label system automatically knows where the asset is, and how much it is. No, if the bottom book itself is a lot of mistakes, like the non-released equipment, the new purchase is not recorded, where no one can tell, the data will be even more confusing -- the system will "oppose" the wrong physical and wrong account, and it will harden dirty data.
So we usually recommend a full manual inventory to the client at the beginning of the project. This sounds dirty, but it's the bottom of the whole program. The inventory addresses three problems:Physical storage does not exist、It's a book and a thing, right?、Who's taking care of each asset, where?. The results of the inventory are deposited in an Excel or a simple desk account as baseline data for subsequent labelling and construction systems.
This step often takes longer than expected. A thorough count at a factory with 5,000 assets normally takes three to five people about a week of continuous work. The time is well spent: every later step relies on this clean dataset.
Hardware selection: scenes determine configuration, not frequency as much as possible
The RFID hardware consists of three main types: electronic tags (on assets), readers (identify labels), antennas (launching and receiving radio frequency signals).
About TabsThe vast majority of fixed asset management scenarios are filled with passive labels, without batteries, small sizes, low costs, and the unit price usually ranges from a few centimetres to a few bucks. The focus is on "matching labels and objects." Metal surfaces need to select metal tags that cannot be read or deflated at a distance; wood, plastic, media like glass are relatively friendly, and common labels can be read a few metres away. In an environment that is bad, it takes high temperature tolerance and a level of corrosion resistance. In doing such projects, the hardware selection typically recommends to the customer for a small batch trial -- to test reading distances and stability in the actual assets, to confirm that there is no problem of mass procurement.
About Frequency: The main domestic asset management use is UHF (UHF, 860-960 MHz), which identifies distance, single-volume readable, low label cost, and meets inventory needs in warehouses and corridors. HF, 13.56 MHz, is suitable for small-scale, dense-spaced scenes, such as archives, tool cabinets. The two frequencies are not absolutely good or bad, and the mix is common in many projects, with critical view of the distance and density of asset distribution.
About the reader and antennaThe method of inventorying varies considerably. Handhelds are suitable for situations where assets are scattered on multiple floors or areas, and require flexible mobile inventories, which is often a handheld that works with several employees and scans along the route. Channel reader (installed at the door or in the corridor) is appropriate for asset access control, such as machine rooms, archives entrances, and this scenario requires antennas on both sides and even on top to ensure that the entry and exit direction is correctly identified.
System architecture: layered design, first to get the core running
The whole architecture of the system can be understood on four levels: sensory layers (labels and readers), network layers (data transmissions), platform layers (data processing), application layers (e.g., inventory, query, report etc.). These four layers are in a fixed logical order, but the specific technology selection can be gradual.
Many projects have just begun to try to fully function — positioning, visualization, depreciation and automatic warning — with the result that the development cycle is often delayed more than six months and business has become impatient. The practical approach is to do three things first:Batch count、Single Location Query、Actual discrepancy report. These are the core values of the RFID programme, which can be used in a linear manner and then gradually added to functions according to operational priorities.
The middle is a very easy layer to ignore. The RFID reader has different protocols, data formats are not uniform, and the middle is the conversion of these raw data into system-recognised information, such as "what type of labels read by which writer at what time." The middle is not well chosen, and then the quality of the data is problematic, missing, repeating, over-reading, and cross-regional misreading is not controlled.
The interface with the ERP, the financial system is another key point. Asset number, name, specification, user, location where these fields need to be connected to the ERP, otherwise the search trails are just a bunch of numbers that do not see what is needed in business. The docking plan is discussed ahead of time with the IT department, whether it's an interface, or a one-way sync, and avoid data fight later.
Process retrofitting: The hardest part is not technology, but to get the first line to work with it.
The real decision on RFID is to succeed or fail, not hardware and software. We have seen many project systems that are fully functional, but the inventory of data after a few months is more messy than it was before it went online, and problems are being encountered in the process.
Several clear process elements:Who's in charge of the tag?• It is recommended that asset management should be centralized to avoid the risk of individual departments being singled out for damage to their labelling;Count frequency— Not as often as possible, the key is to form rhythms such as monthly voltage, quarterly total, annual large disks with different levels of asset interest;Unusual treatment• Clear reporting and back-up process for the gain/loss, damage to labels, change of location;Responsibility rests with the people.— Each asset must have a clear liability, and the RFID data can record who was last approached with equipment and where there is evidence to track it.
The acceptance of liners is a hidden barrier. When they start to expand, asset managers or warehouse people will think that "a table can be drawn before it's done, and now more time." The answer is to minimize the operating costs in process design -- the more assets are counted on hand, not to require each asset to check against the screen. There is also an incentive mechanism, such as linking the count accuracy rate to performance, so that one line is dynamic.
After the line: Put the data decay on the tube.
The RFID system is on line. This does not mean that the work is over. There are various situations in which data quality is being reduced: labels are scraped off, the reader antenna is malfunctioning, asset positions are not updated and new assets are not labelled. These problems accumulate for months, and discrepancies in the system start to appear again.
A mechanism for monthly data health check is proposed: randomly extracting 5 to 10 per cent of assets for manual or handheld double-checks, comparing system data and physical consistency. If the difference exceeds a predefined threshold (e.g. 1 per cent), then sort out the problem of labels, the problem of reading and writing, or the process is not in place. Additional plans for damage to labels are also made - typically after two or three years, 5 to 15 per cent of labels can cause physical damage or read abnormalities, which will require fast access to relabelling, which cannot be processed for six months.
Data are valuable. In addition to the inventory function, it is recommended that several readable statements be exported on a quarterly basis: asset usage, long-term inactive asset lists, assets up to depreciation period, inter-office transfer records, etc. These data combine financial information from ERP and help management make decisions about asset acquisition and disposal. If RFID's data are used only for "searching things," it will be less valuable, and more permanent to use it as input into operational decision-making.
Summary: What are the options for use?
The usefulness of a RFID fixed asset management programme depends on three things:Is the bottom line clean?、Whether or not the hardware configuration fit the scene、Is the process a match for the first line?The technology itself is ripe for the task of establishing data habits and accountability mechanisms in the organization.
Consider a small pilot if you have more than 500 assets, spend more than two weeks a year on stocktaking, and consistently see a discrepancy rate above 3% between physical assets and your records. Start in one area with a high concentration of assets and straightforward workflows, such as a server room with clustered racks or a densely shelved spare parts store. Run the pilot for four to eight weeks to verify that workflows run smoothly and the data is accurate before deciding whether to roll it out companywide. This is less risky than starting with a full deployment.

Ms. Lei
Mr. Pang