The RFID technology ' s powdered milk system has begun to expand in China.
RFID can link milk powder packaging to batch, source and flow records, but the quality of the retroactive system should be judged in case of recall or information dispute.
Retroactive, thermostatic and judgmental food safety are different capabilities
If you want to obtain data such as temperature, you need the corresponding sensor components and collection conditions.
The hardest part about the milk powder is to split and recall.
The relationship between a single can, box and tray changes with the unbundling and reorganization.
The draft does not identify a national promotion event under its original title, nor does it describe general electronic labels as equipment capable of detecting powdered milk or safety.
Recalls can be traced from batches to individual sources. Dismantling, mixing and returning shipments interrupt a simple correspondence. A mock recall is more likely to indicate whether the system can take on real retroactive tasks than if it shows product presentations after scanning.
Put this in a specific process.
Recalls are done, a batch of goods is broken down into multiple orders, and if the system can only find the original storage box, it cannot confirm the current destination. Backtracking requires recording changes in relationships and downstream receptions. Allowing consumers to scan pages, complete information displays and do not prove that the recall chain is available.
Write out the reasoning between information records and quality judgements.
From label identity to food findings, there may be a batch of linkages, environmental records, data calibration and determination models that are applicable at every step. The external measurement of packaging does not necessarily represent the internal state, nor does the complete recording of the process mean that each input is true. The programme should disclose what it has obtained directly and what conclusions depend on additional assumptions. It is recommended to retain the missing data and the situation beyond the model, moving to an application test if necessary, rather than imposing them in order to give the user a simple answer.
Let's look at the input and the results.
| Observation dimensions | Evidence to be retained |
|---|---|
| Retrospective Particle Degree | Clear relationship between individual goods, containers and batches |
| Data is credible. | Sources, calibration and sampling intervals are available |
| Limits of liability | Recording or early warning is not a substitute for formal quality determination |
| This is the key trade-off. | Should the retroactive system prove that “this code is available” or should it prove that the products affected will not be left out when problems arise? |
It's recommended that this be verified.
Recalls are conducted in a simulator series to check whether the areas affected and evidence can be found after the unboxing, return of goods and transfer from the warehouse.
It is recommended that identification checks, environmental data and quality determinations be separately validated.
It's worth asking.
Should the retroactive system prove that “this code is available” or should it prove that the products affected will not be left out when problems arise?

Ms. Lei
Mr. Pang