How do we build a fire-fighting asset management system? Equipment and material control
What are the criteria for a system?
A truly operational RFID fire asset management system, with only three items for inspection and inspection: equipment and materialsConcordance(A long-term physical error rate of less than 1 per cent between book and site),Inspections are retroactive.(Every count, every change in position is recorded with time and responsibility)The exit is restricted.These three items cannot be easily removed from the warehouse without approval or time-out. The other three, which are just a look at the way they are put in front of them. We use these as hard bars for the end-of-takes when we do projects for the fire brigade. Here's to reverse the judgment logic of several key steps in the build.
Think first: What kind of assets are we talking about?
The fire protection system is divided into at least three types of assets, each with a very different treatment approach, and it is recommended that a full inventory be performed before the system design:
- High-value single-unit equipmentThe high unit price, low frequency of use, and security for every use are the regulatory requirements of “one size, one life cycle,” for these assets.
- Bulk consumptionWater belts, filter tanks, fire suits, oil, etc. are large and fast. The management objective is “Bile by batch/box, early warning by valor”, and each should not be labelled or labeled.
- Emergency reserve and combat materielThe speed of response is more important than the daily count.
The most common reason for returning to work is when the classification is not clear. A fire brigade labeled all materials in a single format, and then dropped them in one year, reticked them, and the cost of reposting was higher than the system itself. Then the water belts, filters, etc., were changed to "Classes by batch + box-checking" and the problem was solved.
Hardware selection: tags and read-and-written logic
RFID looks like a lot of hardware, and the real impact on the system is that it's not working.Frequency MatchwithAnti-metallic/anti-liquid capability。
Frequency Selection
The metal parts, liquids, oil-filled environments are many. Conventional UHF (UHF, 860-960 MHz) read long distances, lot recognition is fast, but attached to the metal surface will be shielded, and reading and writing distance will shrink from 5 metres to dozens of centimetres. We encounter metal casings for hydraulic breakup tools that can hardly be read by conventional soft labels, which require special attention.Anti-Metallic TagsHF, 13.56 MHz, is slightly more metal-resistant but within a short distance (generally within 1 metre) and suitable for close-range verification in small items of assets or toolboxes.
Environmental adequacy
The normal method is to change the water.ABS or PC material envelopewater-resistant tags, with an epoxy coating if needed. For textile items such as firefighting gear and combat boots, sew in flexible laundry tags rated for the highest drying and disinfection temperatures the gear will face, usually at least 90°C.
Reader Deployment
Proposed deployment of warehouse entrances, vehicle loading and unloading points, equipment door-to-door closuresChannel ReaderThe contents of the storeroom are attached to a wall or ceiling antenna for each 5-8 metre of shelf, avoiding blind reading. Handheld terminals are used as a bottom-dips tool for temporary count-taking, equipment for repair and so forth, and for leaving the fixed reader area.
Software architecture: smoothing data flow
The most vulnerable aspect of software is not the lack of functionality, but the fact that it's a problem.The data source is not working.. The system's ideal data stream is this:
- When equipment/materials are in storage, the RFID channel reader automatically reads the label code to match the purchase order generationLibrary records。
- When the application/repatriation is taken, the passper captures the movement and the system determines compliance with the approval flow and the unapproved call produces a warning.
- Daily inspections are scanned by handheld terminals along pre-established routes, account data are sent back via Wi-Fi or 4G and anomalies are forwarded to the equipment manager.
- The inventory supports two models: the storage room.Quick Inventory(channel reading + shelf antenna), and on-site equipmentCheck Item-by-Search(handheld read at close range)
It is important to note that RFID is a collection tool.Real accounting relies on back-office data management.. The tag EPC code (Electtronic Production Code, the only ID number stored in the label chip) must be numbered with the equipment, the purchase order number, the responsible unit bound, and cannot simply leave a non-linked electronic label on it, otherwise all subsequent data will float.
Business interface: process over function
The real test of the system landing at the fire department is...Bite-up with established business processesWe work with a municipal fire division, and the original online version is completely unusable. The problem is that it's connected to the process:
- The approvals are not connected to the existing OA.Equipment acquisition is still subject to paper-ticked approval, after-action re-entry only in the system and with significant delay in the inventory data.
- The inspection rules are not embedded in the system.: Fire protection equipment has clear periodic detection norms (e.g. annual check of respirators, filtration tank validity) and if the system only records " in/in place " during inspection, without triggerEarly warning dueThis is equivalent to cutting off half the core value of the system.
- The battle assignment is missing.The movement of equipment during fire rescue was sudden, with pre-positioned list of equipment, vehicle loading positions, and the person responsible being secured in advance of the system and moved out by one key at the time of dispatch, much faster than a temporary check.
The system will become a daily tool only after we have approved the current OA for the power stream, made the due warning automatic and made the battle plan a one-click assignment order.
A couple of easy-to-touch pits.
The project that has been developed over the years is re-engineered with several common problems which deserve to be avoided earlier:
- One-time full-brand tag. The stickling strength of the label requires batch verification, especially in high temperature, damp and oil-stained environments, which should be piloted in small lots for 3-6 months before being fully spread.
- Ignores the interference of a reader. The metal shelf is a very dense storage room, and the UHF antenna needs to stagger or increase space between them, otherwise it can interfere with each other, and the reading rate will drop from 99% to less than 70%.
- No offline cache.Fire protection units often have a dead end to the signal in their garages and training grounds, and handheld terminals must support local cache data and return to the signal area for upload, otherwise there will be “no memory on board, no thought of the disk”.
- Lack of awareness about data security. The RFID label itself is accessible by the close-to-field device and key equipment numbers should be done if they involve sensitive informationData Encryptionor use a password to protect the label.
Follow-up: Can the system last three or five years?
The system is not over.Transport systemWe suggest that at least three things be created:Tab InspectionQuarterly check-up of 5%-10% label readability and find that the drop or failure is timely to refill;Reader maintenanceBiannual inspection of antenna interfaces, power supply modules and channel readers recommend a reading rate calibration once a year;Reconciliation of accountsOnce a month, small and large, quarterly inventories are conducted to keep errors within the traceability range.
After all, RFID is just a tool.Conformity of accounts, closed loops, accountability. The system is alive only when hardware, software and process systems are aligned; it is always the beginning of a failed project that focuses on the price of equipment lists.

Ms. Lei
Mr. Pang