What are the fixed assets inventory methods? Manual, sweep and automatic identification
The stocking of fixed assets, which many people in the firm think is "just one at the end of the year," but it's often true: the card machine on the books has been in the corner for three years, hundreds of books and physical numbers are missing, the team is exhausted by a round, the data is not available. Manual, sweep, automatic identification of three routes, no absolute good or bad, the key is to see your size, the human budget and the tolerance requirements. Here's how these three counts work.
Point 1: Manual inventory - process is simple but human cost highest
The manual inventory is the type of asset list that is printed, the form is checked on a case-by-case basis and checked against the ticking or filling in. This is the starting point for the first time that most enterprises do an inventory, and it is the most primitive way to do it.
What's the scene?
The quickest-ever, relatively centralized (one floor or one workshop can walk) of a total of 500 assets and only once a year is the case for small and medium-sized office space, which requires little prior input.
Real workload reference
A hardware trading company had about 1,200 assets, mostly in two buildings. It used to assign four administrative staff in two teams to the year-end stocktake. Each team had one handheld terminal, essentially just running an Excel spreadsheet. The count averaged five working days, followed by two to three more days to organize discrepancies and locate assets. Total effort was around 28-32 person-days. The final discrepancy rate was around 3%, meaning 30 or 40 assets could not be found or their status did not match.
Several common problems with this method
- Data lag:The results of the inventory will not be aggregated until a few days, while the business sector is still using assets and the time frame for the inventory window has passed when the statement is available.
- Man-made Leaks:The assets labelled are not necessarily present, and the assets borrowed, redeployed or maintained are not in place on the day of the inventory count and can easily be considered "lost".
- Cost of review:When differences arise, it often takes two or three counts to locate problems, doubling time and manpower.
So a manual inventory is only suitable for situations where assets are small and not very demanding for data accuracy. Once more than 1,000 assets are located or distributed in multiple locations, the upgrade should be considered.
Element 2: Scanning inventory - the direction of most enterprises for their first upgrade
The most common way to do this is to put a bar or two-dimensional label on each fixed asset, and the inventory operator with a PDA or a mobile phone App, go to the device and sweep it, scan the results for real time upload.
The front-loading is not an exaggeration.
The bar code label is about $0.05-0.2 per specification, the two-dimensional tag is a little bit expensive but within $0.1; the entry level PDF equipment is singled out for $800-2000, while the App program can use the employee's own mobile phone directly, and businesses need to purchase tags and scanning software.
How much efficiency has improved?
Similarly, 1,200 assets, usually 2 people, can be scanned in half a day, and the overall manpower input is reduced to 4-6 person days, with a difference of less than 1%. That is, the workload is roughly one fifth to one quarter of the manual count, increasing by two or three times.
A few easy-to-penetrate pits when you land.
- Labels off or deface:Paper labels wear quickly in workshops and warehouses. Often 5–10% need to be reprinted and replaced within six months. PET or other synthetic labels substantially improve durability.
- Scan Deadhorn:When assets are embedded, the equipment labels behind are blocked and need to be moved in order to be cleaned.
- Only "in the presence" and "in the absence":The assets that are cleared by the sweeper are accurately recorded and are simply missed. A round of asset status confirmation (which is borrowed, which is maintained and which is scrapped) is conducted before inventory counts are taken, which reduces subsequent tracing.
The borderline of a sweep is whether it's in essence or "man-o-person, asset-free." If assets are liquid, count frequency wants to be monthly or even real time, this path will not be enough.
Element 3: Automatic recognition - RFID, Visual and Iot are the three main paths
Automatic recognition counts have been used increasingly in manufacturing and logistics over the years. The central idea is to make assets "talk": no man needs to go ahead of each asset, and the location and status of assets will be reported to the system.
Option 1: RFID (radio-frequency identification)
A RFID label is placed on each asset, several readers (i.e. RFID antennas) are placed in the warehouse or on the floor, the reader sends a periodic signal and the next range of labels is activated and automatically returns information.
The RFID labels are divided into several categories by frequency band: UHF tag unit price of 0.5-3 yuan, reading distance 3-10 metres, identifying hundreds of pieces at bulk at a time; HF tag unit price of 1-5 yuan, reading distance 10 cm to 1 m, suitable for the scene where precise location is required.
An auto parts customer had about 8,000 returnable assets in its warehouse. Counting them by hand-scanning took six people three days. After switching to RFID, two people could walk the warehouse with a mobile reader and get results in two hours, with a discrepancy rate that stayed under 0.2%.
Path II: Computer visual inventory
Loading cameras above the shelf or channel, combining image recognition algorithms, automatically identifying asset appearance, numbering and location. This way does not need to post anything on each asset, which is particularly friendly for historical assets. But the accuracy of identification is lighted, angles are more influential, and now the country ' s mature cases are concentrated in highly standardized scenarios, such as room equipment, retail shelves.
Path three: Iot and positioning technology
For high-value or high-security assets (servers, medical equipment, dangerous chemical containers), blue teeth, UWB or LoRA positioning modules can be added to the platform and the location and status signals can be sent back to the platform in real time. The cost of this type of program is 50-300 yuan for single asset hardware, with a significantly higher overall input than RFID, which is more suitable for fine-line management of critical equipment.
Automatically identify the three paths to the count.
The criteria for the determination are simple: whether your inventory needs "cyclicalally over" or "continue to track every item." The former selects RFID, the latter add IoT positioning; visual recognition is now a complementary tool and does not recommend it as a dominant option.
Point 4: Three approaches compared to which four indicators
The selection method is not to be biased by a single indicator and it is suggested that four dimensions should be seen at the same time. The following table gives the common reference value (based on general industry experience, specific items will float):
| Contrast Dimensions | Manual | Barcode/ QR code Scan | RFID Auto-Recognition |
|---|---|---|---|
| Assets 1000 count timed | 4-5 working days / 4 persons | 0.5-1 Working days / 2 persons | 1-2 hours / 1-2 persons |
| Variance | 2-5% | 0.5-1% | 0.1-0.3% |
| Single-assets hardware input | Ignored | 0.05-0.2 yuan (label) | 0.5-3 yuan (label) + antenna input |
| Implementation cycle | Almost zero. | 2-4 weeks | 4-12 weeks (including hardware deployment and system interface) |
| Support for the frequency of the count | Year | Quarterly to monthly | Real-time to day |
Several additional notes:
- The difference is not just about numbers, but also about "explainable." The difference found by RFID is mostly real and easy to trace; the difference in manual counts is mixed with a lot of human-made leaks and miscalculations that don't make any sense.
- The implementation cycle is underestimated by many companies, and the RFID usually takes 2-3 months from programme recognition to steady operation, and the time cost is factored into the budget.
- The support for the frequency of an inventory determines how much you can manage. The annual count only solves the question of whether the accounts match, and monthly or real-time counts allow you to see data on the size of these fine particles in terms of asset flows, idleness, maintenance time.
Point 5: How? Look at your "card point."
And to finish these three approaches, and finally to choose, my experience is that the problem starts in turn: taking stock of what this thing is about, deciding which one you should use.
It's stuck in human costs.
The inventory team is stifling, and they need to be co-opted every year with a temporary staff – the most common point of pain. Scanning can solve most of the problems, with a budget-critical first-time phone app transition, and a willingness to go directly to RFID.
It's stuck in data accuracy.
The book and physical numbers are not matched for a long time, and they are repeatedly explained, and then tracked over -- in this case, either process problems or method caps. If assets are liquid, cross-sectional transfers are frequent, sweep codes can only be improved but unsolved, and must be automatically identified at the RFID level.
It's stuck to manage the particle size.
What you want is not "the results of the inventory" but "know where any asset is, what status" at any time -- which means that annual or quarterly inventories are not enough in themselves to keep collecting position and status signals, RFID+ positioning or IOT is a mandatory option.
It's stuck in history.
The old asset is not labelled, information is missing, and the attraction of refilling costs -- visual recognition programs is here. But currently, the visual program does not meet the master-force inventory requirements for its identification accuracy and stability, but it is more appropriate as a support tool to use the sweep code or RFID.
Point 6: Which way to upgrade? The rhythm is more important than the selection.
From manual direct to full automatic recognition, the range is too large for most firms to digest once. The more steady rhythm is two steps:
- First step (0-3 months):Replace manual by bar code + PDA or mobile phone App to bring the account to a level that is no longer valid; and complete uniform standards and data governance for asset tags.
- Second step (3-12 months):For a core high-value or high-liquid asset pilot RFID or IOT, the decision to expand coverage is taken after running.
The advantage of this rhythm is that the first stage is small, low risk, and it allows a team to build up its habit of "the inventory being a routine rather than an annual event"; the second phase has a clear data base, choosing which assets RFID can be targeted.
Element 7: Beyond the inventory, systems and data governance is the bottom plate
Whatever the method of taking an inventory, one premise is often ignored: the bottom asset master data should be clear. A multicard, a non-reciprocal card, a missing status field -- these problems are not solved, and in the case of advanced identification, the data generated remains unsettled.
Two things are suggested before the official inventory is taken:
- Clearing the rules for asset master data: what counts as an asset, what number is used, how status is defined and how redeployments are recorded.
- The establishment of a life-cycle record of the asset "birth-use-change-death" allowing each count to be cross-checked with process data rather than looking at the physical in isolation.
When we did the asset inventory, we met with companies that were focusing on identifying hardware options, and the data bottom was not well-seted, and the results were increasingly confusing. The way to identify it was "on site capture efficiency," but not "coherent data per se."
The core difference between the three approaches to inventory is one sentence:The manual is "person to assets" and the sweep is "equipment to clean assets."The four variables determine which level you should be in. There is no one-step solution, only the most appropriate option for the current phase.
If you can't start with any step, you can take a first count of your own asset status and the current process — clear where I am now — more important than rushing to choose.

Ms. Lei
Mr. Pang