How does the RFID Asset Management System (AMS) construction programme work? Requirements and implementation points
How deep are we going to write this?
RFID, the most common problem with building an asset management system is not "do you want to write" but "write up to any level" -- it's shallow and unreadable; it's deep, and afraid of being tied to death. The test is just one:The program must serve both the "decision-making" and "executive" levels.The former needs to see input output and risk, while the latter needs to see specific actions and acceptance calibrations.
When we take the target apart, the bone structure of the program is clear: demand analysis addresses "why" and the overall design answers "how," the implementation plan deals with "who does what when", and transport and extension covers "what happens after." When we do a programme review for our clients, it's usually suggested that the space should be 2-4:3:1 distribution -- needs and design weigh, implementation follows, and transport leaves separate chapters without too long.
Demand section: The list of functions in the column is only the starting point
Many of the programs stop at the generic description "Automated Asset Identification, Increased Inventory Efficiency." The real pain point is not functional, and the data as it stands, the process nodes and division of responsibilities -- that's what makes the RFID asset management system so critical to landing.
Rules for classification and coding of assets
The particle size of the asset label needs to be specified: whether it is "one piece" or "lot" (e.g. low-value depleting material). The effect of the two particles on the label cost is completely different from that of reading and writing equipment and backstage data structure. We met with a client who was in the college lab, initially trying to mark all the individual items, and it took almost 900,000 (estimated at an average price of $4-5 yuan/mixed on the EPC UHF label market) to count 200,000 asset light tags, and then to replace them with a compromise solution of "one item at a unit price of more than $500, and the rest by batch or area."
Operational process interface with existing systems
The inventory cycle, new/transfer/waste process, responsible authority, which if not clearly written, is repeated during the implementation phase. It is recommended that the current and target flow chart be drawn directly into the programme, indicating the input output for each node. Common docking systems include ERP, finance module, OA approval stream, interfaces are typically done in an intermediate library (intermediate table) rather than directly linked to the counterpart ' s main repository, allowing for a separation of privileges and a back-rolling of failures.
Performance indicators
Three of the most critical:Number identified at the same time(Depending on the number of reader antennas and collision-proof algorithms, normal UHF channels can stabilize 200-500 labels/s),Identification accuracy rate(on-site testing is typically needed around metal or liquids; typical industry requirements are 95%-99%),Inventory Efficiency(The speed multiplier for the manual-held disk, vs RFID channel, is typically 10-50 times the range, but it is strongly related to the warehouse layout.
Selective comparison: self-study, procurement or secondary development
The choice between "self-study" and "redevelopment based on mature products" in the face of a large market volume of RFID asset management products is often more problematic than which equipment manufacturer.
| Dimensions | Self-study bottom + operating systems | Grown product + Secondary development |
|---|---|---|
| Cycle | 6-12 Months of Start | 1-3 months |
| Cost structure | Manpower costs are predominant, hardware purchased on demand | Software License + Small customization fee |
| Controllability | It's completely manageable, flexible. | Reliance on vendor version plans, limited depth customization |
| Application phase | Large asset volume, special process and long-term IT team | Medium scale, process standards, wish to be online |
When you connect with a customer, Shanghai-based SIT usually helps the client to measure assets, complexity of business, three points on the current information base, and is above a certain threshold to recommend self-study; otherwise, based on the configuration and small development of mature products, the overall value is more stable. Attention needs to be paid to the two categories of "maturity product" - pure RFID intermediates (which require a self-connection of business systems) and complete packages with business modules (both inventory, report, approval), which are fast but low in flexibility, choice on demand.
A few things that hardware selection can easily ignore.
The RFID device selection cannot read only the reader parameters, but also depends on the antenna, label and match of the field environment. The UHF (860-960 MHz) is suitable for long-range mass identification, often used in warehouses, corridors; HF HF (13.56 MHz) is good for metal interference, suitable for tool cabinets, small equipment intensive scenes. Microwave bands (2.4 GHz) are also available for asset management, but they are expensive, labeled, and industrial scenes are less useful.
And the other thing that's often overlooked isSeal form for labels: On-metal tags cost two to four times as much as standard tags, and liquid-resistant encapsulation adds another 30%-50%. If many assets have metal housings, such as cabinets and servers, this is not a cost you can cut.
Implementation: hardware, software, data sequence
The extent to which implementation is written in the programme determines whether the project can be delivered on time. It is recommended that implementation be broken down into four phases, each with a clear "signage of completion".
Phase I: Environmental preparedness and pilot validation
Select a small area of 200-500 assets to pilot RFID, verify label selection, antenna position, and read-and-write distance for performance indicators. This phase is usually 2-4 weeks, but it exposes 80% of the technical problems in subsequent large-scale implementation ahead of time. After the pilot is completed, you must output the Pilot Certification Report, and record the recognition rate, the missing rate, the error rate, etc.
Phase II: System integration and data governance
Test RFID middleware integration with ERP and finance systems using simulated data first. Run through three to five core workflows, such as receipt, stocktaking, transfer, and disposal. Clean asset master data too: standardize names, units, and value classifications. Accurate RFID reads cannot compensate for poor data quality.
Phase III: Batch implementation and training
The training is conducted in parallel with the training of operators on the ground, or there will be an embarrassment to "systems are online but no one uses them." Our approach is to move in batches, first internal inspection and then switch over to another batch after each batch has been completed.
Phase 4: Pilot operation and acceptance testing
The test runs a proposal covering a full inventory cycle for no less than one month.
Applicable scene: policy differences for HF flow vs low-frequency disk counts
The same RFID asset management system, "opening" in different asset management scenarios is very different. It's not like the plan is written out of the picture.
HF-streamingThe programme focuses on the deployment density of closed reading and writing equipment, antenna orientation angles and signal coverage testing methods.
Low-frequency Class(e.g. office equipment, fixed assets, archives): Focus on handheld terminal count+ periodic drawings, RFID is mainly addressing "get difficult" and "annual inventory time-consuming". The program focuses on the assignment logic of an inventory mission, moving end App experience and offline data synchronization mechanisms.
Mixed(large manufacturing enterprises, across geographical groups): There is a need to support both fixed access and mobile handheld, and to consider multi-level aggregation of data. The most vulnerable scene is network stability -- handheld terminals in workshops, warehouses that are often Wi-Fi unstable, offline caches and breakpoint transmission mechanisms must be written into the program.
Common pits: a few things that are abused when the program is evaluated
Ignore the label 's full life-cycle cost'
The plan lists only upfront purchasing costs and leaves out three to five years of tag reprinting, battery replacement (active tags generally need new batteries every three to five years), and tag application labor. Ongoing operating costs often exceed the equipment purchase price, so the plan must include an estimate of five-year total cost of ownership (TCO).
Ignoring electromagnetic compatibility and safety
Metals, liquids, humans all decay RFID signals, and the program sets the equipment model directly without field surveys, with a rate of return at an approximate time. In addition, extreme temperatures (coolers, high-temperature workshops), dust, oil pollution environments require different labeling levels. If the customer has special circumstances (e.g. cold chains, chemical) these constraints are written into the demand document.
Permissions and Data Security
The RFID system records the location of people and assets, and if it doesn't spell out who can find what data, how long will the data be kept, how much of an audit will leave behind, compliance problems, especially in hospitals, financial institutions, government offices, etc.
Extensions are verbally promised, but not quantified.
The phrase "for the future supports XX" is written without technical pre-engineering in a program, and often it is time to reverse the re-entry. The spectrometer is to expand the possible functionality into a list of compatibility assessments under the current architecture -- the smoother-up sign "supported," the second developed mark "needs input", and the non-supportive label "needs re-construction."
End of story: Three things before you write the plan
Before writing, it is suggested that the following three things be done, which would make the programme more persuasive:Field surveyat least once in the area where representative assets are located, recording metal/liquid ratio, humidity, network conditions; andData sampling100-200 asset information drawn from existing desk accounts to estimate the data governance workload based on field completeness;Case against the target, search for public information on 2-3 peer-down projects to avoid empty design.
A qualified RFID asset management system construction program, which is not a functional statement in essence, but a combination of "basis of decision-making plus implementation manual." The evaluator wants to see if it's worth doing, and the implementer wants to see what's next -- both sides can read -- that's the real way to land.

Ms. Lei
Mr. Pang