According to the warehousing introduction in the previous article of this series, we already know that there are four major scenarios in warehouse management: warehousing and exiting scenes, sorting scenes, inventorying scenes, and cargo tracking and positioning scenes. This article will mainly introduce the design of IOT+RFID solutions to achieve warehouse management in and out of the business.
Description of warehouse entry and exit scenes
1.1 Scene description and comparison
Inbound and outbound management, that is, through modern digital technology to complete the effective identification and monitoring of warehouse goods outbound and inbound, as well as the check of manifests and goods.
In a general scenario, the truck loads the goods that need to be managed and arrives at the gate of the warehouse. The traditional warehousing solution is compared with the RFID solution.
In the traditional scheme, after the truck arrives at the warehouse door, it needs to be unloaded on the spot. The administrator holds a scanner to identify the goods, and after the goods are compared and checked, the warehousing operation starts, which lasts about x hours.
Under the RFID scheme, after the deployment is completed, the goods are directly unloaded and put into storage. After warehousing is completed, it can be automatically checked, eliminating the steps of unloading inventory and manual scanning.
1.2 End-to-side deployment design of RFID solution
Under the RFID end-to-side technology, radio frequency can solve the excitation and signal recognition of RFID tags, but the core problem is how to complete the judgment of the incoming and outgoing direction of the finished goods.
The radio frequency is divergent, and the antenna excitation signal can be abstracted into a sector shape. Within the sector range covered by the antenna, the RFID tag may be excited and identified. However, it is impossible to produce a sense of “direction” only by relying on a fan shape. It will be recognized when entering the fan-shaped coverage from any direction. It is also impossible to recognize whether the goods enter or leave the warehouse door. You can only know that the goods are being moved around the warehouse door. Recognized.
According to the theory of determining a straight line from two points, we deployed a radio frequency identifier (helper) inside and outside Kumen to solve the problem of determining the direction of goods identification.
Understandably, if a goods affixed with an RFID tag is first stimulated by the helper on the outside of the warehouse door, and then by the helper on the inside, we believe that the goods have been put into the warehouse in this short period of time; if the goods are first on the inside of the warehouse door The helper is motivated and then motivated by the outside helper, which is considered to be an outbound operation.
As shown in the following diagram of the end-side deployment architecture:
With Huawei’s RFID transceiver separation technology, helper devices need to be deployed on the inner and outer sides of the library door or channel to stimulate RFID tags.
Deploy receiver equipment in a larger area, responsible for receiving the signal after RFID tag excitation, so as to identify and obtain useful information.
Difficulties in the warehouse scene
2.1 Large amount of data
The RFID device emits radio waves to excite the tag, and there is a difference in intensity in the specified frequency band. In order to cover a larger identification range, the intensity of the general radio frequency will be appropriately increased to ensure that the RFID tag is activated enough times, and the scan of the redundant tag is increased to improve the accuracy. Therefore, during the passage of goods, the number of signals induced by the tag is very large.
The general warehousing scenario is to solve the problem of manual efficiency. At the same time, there will be a lot of goods identified and a lot of tags, so the amount of data will increase exponentially.
2.2 Data cleaning and analysis
The real model of warehousing management on the equipment side is that the RFID tag is scanned by the helper at a certain time, and our business is actually concerned with the goods being identified by the warehouse at a certain time. RFID tag data needs to be converted into cargo data, and helper identification needs to be converted into helper-related facilities, such as empty doors in and out of the warehouse.
On the other hand, the complete outbound/inbound of goods is analyzed by multiple RFID scanning events, which need to be scanned by InHelper and finally be scanned by OutHelper to generate stable state transitions to analyze. In addition, in actual situations, Inhelper and OutHelper will have cross coverage, and their state transitions are not linear, and more complex analysis algorithms are needed to realize the state transitions.
Finally, when there are multiple warehouse doors in parallel, there may even be mutual interference scanning between doors. The same cargo label will be scanned by door 1 and door 2, and it is easy to judge the abnormal event of multiple doors in and out. data.
How to use the IOT platform to solve
3.1 Device access service
Solve the problem of RECEIVER accessing the platform, using massive data uplink and high concurrency capabilities to solve the problem of massive data upload, and realize real-time data upload for subsequent analysis modules.
3.2 Data Analysis Service
Quickly docking equipment access services can naturally obtain equipment data and perform effective analysis.
The asset modeling module can be used to complete the cleaning and conversion of equipment data; the real-time stream analysis job can be used to complete the filtering, denoising, and state inference of RFID data, realize the access analysis, and generate event data. And the real-time stream analysis job can support the docking of various output components, such as DIS data access service (kafka), such as SMN message push service (can push SMS, email), etc.
In addition, data analysis services can provide capabilities such as real-time analysis and offline analysis to help users complete basic statistics and big data analysis capabilities of Internet of Things data, such as reports on the total number of daily incoming and outgoing goods at warehouse doors.
After the equipment-side networking is completed, the helper stimulates the goods tags in the coverage area, and the receiver device is responsible for collecting the signals generated by the RFID tag excitation in the field, and connects to the industrial computer through the serial port. Use the supporting program on the industrial computer and integrate the IOT Device SDK, convert the RFID signal from hexadecimal message to json, connect the IOT platform device to access the service, and report the data.
At the IOT platform layer, the equipment access service is responsible for the accounting and management of equipment, and receives equipment data; the data analysis service performs data conversion and analysis.
The data analysis service can analyze the original data into event data and send it to the message middleware such as DIS service, and the upper application consumes the event data to complete the corresponding business.
Conclusion
The above is based on the analysis and design of the warehousing scene in warehouse management based on RFID+IOT technology. Next, we will introduce in detail how to use IOT technology for access, modeling, and algorithm analysis to achieve data cleaning and event analysis. @
