How does the asset management system access RFID?
RFID Knowledge

How does the asset management system access RFID?

PublishedShanghai Zejie Smart Technology Co., Ltd.

The asset management system accesss RFID, which is based on a middle layer between the original business system and RFID hardware. Readers only return original label data, operations only concern asset numbers, locations, status fields, hardware differences and data format discrepancies are isolated from the intermediate. Specific access involves four jobs: Middler deployment, EPC code and asset number mapping, interface protocol design, reading and writing rules are fine. Below is a few dimensions.

The asset management system is connected to RFID. How does the overall architecture work?

The typical three-tiered structure is: the bottom is RFID readers and antennas, which collects the physical layer; the middle is the intermediate service, which is responsible for data weighting, event resolution, status determination; the upper is the asset management system, which receives structured events through RESTful API or message queue. The benefits of this segment are that hardware selection and business system adaptation can go in parallel, without hindrance.

Shanghai-based Scratch technology uses this structure in a number of projects in storage and manufacturing. Readers choose which, what labels, midway through Windows or Linux, are the middle end; only the interface fields on the side of the business system need to be agreed, whether Impinj is the device below or Alien.

What are the specific responsibilities of the RFID intermediate?

The middle is not a simple data feed, but it does four things:

The intermediate is usually deployed in a stand-alone server or container, and exposes external information channels such as HTTP API or Kafka, RabitMQ. The deployment is based on the real-time requirements of the business system - the asset inventory scene spectrometer response is sufficient, and the product line in process requires a millisecond event to be sent.

What type of reader protocol does the intermediate require?

Common protocols are in two categories: UHF-ISO 18000-63/EPC Class1 Gen2 standard, which is much more used for logistics warehousing; HF-ISO 14443 or ISO 15693, door-ban, tool management is common. The intermediates need to support both LLRP (Low Level Reader Protocol, for controlling reading and writing) and ALP (Application Level Event Protocol, for receiving label events). The details of the two agreements vary greatly from one manufacturer to another, and the fit-for-work of the intermediates is mainly spent here.

How does the interface between RFID and the asset management system be designed?

The core of interface design is to define the type of event and data format.

RESTTful API is suitable for request-response type operations, such as searching for an asset in the current area; WebSocket or message queues are appropriate to keep moving event streams. The data format is light on JSON compared to XML, but consistent with the previous regulations of the asset management system, avoids two sets of field names.

What's the key interface field look like?

A standard asset movement event usually consists of: epcCode (single tag code), assetId (assets number in the operating system), localcode (location code), timestamp (read time, ISO 8601), readerId (collect device number), rssi (sign strength). The rssi field, although not necessary, is useful for the view of the type-reading rate and the ablution antenna, and it is recommended to retain it.

How does the EPC code and business asset number correspond?

The EPC code is the only ID that is burned when RFID labels are out, usually 96 binary, and the hexadecimal is 24-bit string; while asset numbers in asset management systems are generally business-specific, such as "EQ-2024-001." The two need a map sheet to link.

The mapping strategy has three ideas:

The first option is the most widely applicable and makes migration and auditing easier. The third suits very large tag populations where database query load must be reduced. We default to the first option unless the customer has more than 100,000 tags and needs the highest possible performance.

What are the most common problems with physical access?

Several high frequency problems can be avoided earlier:

Read rate fluctuations.The RFID signal near the metal asset declines, and the distance of reading may be reduced from 5 m to 0.5 m when the label is placed directly on the metal surface. The solution is to use a metal resistant label or a layer of absorbent material on the metal surface.

Error and missing.When the two readers next to each other overlap, a label may be read at the same time by both readers, which causes confusion in location determination. The middle is determined on the basis of readerId and signal strength rather than simply in chronological order.

Inventory performance.When you scan thousands of labels once, the reader may be thrown out. The middle is to support the Bulk Reaing mode and configure a reasonable Inventory Cycle parameter, which is usually configured at 50 to 100 tabs per second.

System decoupling.The RFID event cannot be lost when the business system is upgraded or restarted. The message queue is introduced to understand the chords - the middle is only responsible for throwing events into the Quarter, and the business system is consumed at its own rhythm.

How many steps does the access process usually take?

Phase one is requirements and tag selection. Define asset types, range, and quantities, then decide between UHF and HF and between standard and on-metal tags.

Phase II involves intermediate deployment and interface connection: the reader is completed with a test tool for the intermediate self-hatted to verify the validity of the original data, followed by the docking operational system API.

The third phase is pilot and rule-making: first run through a warehouse or workshop, verify the map accuracy, trigger logic of events, and then gradually move it.

The full process generally takes four to eight weeks. Most of the time goes into integration testing in phase two; getting the hardware in place is actually the quickest part.

In general, the difficulty of RFID access to asset management is not in hardware, but in the case of the middle-case incident processing logic and interface field design. Planning these two blocks clearly, the operational system is actually limited in terms of adaptation, more at the rule-tuning and pilot validation stages.