- The German d-you EUDI wallet will launch in 2027 to provide digital identities on smartphones, enhancing public and private service access.
- Physical cards using RFID, NFC, and secure chips remain crucial due to their security features and established infrastructure in access control and government applications.
- Accurate synchronization of data encoding and printing is essential to prevent identity misalignment during card personalization.
- Personalization speed often limits production throughput more than mechanical card handling, requiring trade-offs between verification depth and process efficiency.
With d-you, Germany will launch a government-run EUDI wallet in 2027. At the same time, RFID, NFC, and chip cards will remain important secure identity media. Armin Rinas, CEO of Rinas Gerätetechnik, explains from an engineering perspective how encoders, printing, cameras, and software work together precisely.
Germany is preparing for the next major step in digital identities. On January 2, 2027, d-you—the government-run German EUDI wallet—is set to launch. Identity data and other digital credentials will be stored on smartphones and can be used for both public and private services.
SPRIND is driving this development on behalf of the Federal Ministry for Digital Affairs. Bundesdruckerei provides the PID Provider, a central infrastructure component through which the core digital identity is generated based on the online ID function.
The wallet promises convenient, Europe-wide interoperable digital identities. At the same time, discussions regarding technical and data protection issues highlight just how challenging such an infrastructure is. The European Data Protection Supervisor cites, among other things, protection against unauthorized access, unnecessary data sharing, and the linking of transactions and potential profiling as key design challenges.
These are not arguments against digital wallets on smartphones, but they do show that trust in an identity must always be established technically, regardless of whether the credential is stored on a smartphone or on a physical card.
Why the Physical Card Continues to Play an Important Role
RFID, NFC, and chip cards will therefore not disappear with the digital wallet.
In many applications, physical cards offer features that remain attractive: A secure chip can store cryptographic keys and identity data securely. Contactless cards do not require their own battery. Possession of the card constitutes a clearly defined physical layer of security. Depending on the application, additional visible security features, laser personalization, and tamper-resistant materials can be used.
Especially in access control, payment, corporate ID, ticketing, government applications, and closed infrastructures, the card is also part of an already established technical process chain.
Digital wallets and physical cards therefore do not need to be viewed as opposites. Rather, both forms of credentials will coexist for the foreseeable future.
And this is precisely where Armin Rinas’s engineering perspective comes into play.
He does not deal with digital identity in the abstract, but rather with the machines and modules that actually manufacture and personalize physical identity media. Encoders, RFID readers, printing systems, cameras, card-handling equipment, and production software must function as a unified process.
Because before a secure card can be used, one thing must first be guaranteed: the correct identity must end up on the correct card.
There must be no data misalignment
Mr. Rinas, from a technical standpoint, what is the key challenge of an automated personalization line?
Armin Rinas:
Card personalization—that is, the data encoding—and the printing must align. There must be no data misalignment.
This statement describes one of the key tasks of industrial card personalization.
For example, a system writes an individual data set to an RFID or NFC chip and then applies personal or other variable information to the card. Both processes must remain precisely aligned with the same physical medium.
With thousands of consecutive cards, this becomes a challenging task for machine and software control.
If a card is diverted, processed with a delay, or reprinted, this must not cause a shift in the assignment of the subsequent data sets.
The system must therefore know at all times which data record is at which position in the production process.
Personalization often determines the machine cycle time
Where exactly is the throughput bottleneck in industrial RFID personalization today?
Armin Rinas: Not necessarily in the printing or mechanical transport of the card.
As a rule, it’s the time required for personalization. Depending on the volume of data and security requirements, this takes the longest.
Electronic personalization can thus determine the cycle time of the entire machine.
How quickly an RFID or chip-based medium can be processed depends, among other things, on how much data is being transferred and which security functions must be performed.
From a mechanical engineering perspective, this means that the theoretical speed of a printer or transport system alone says little about actual production output.
What matters is how quickly the entire process can be reliably completed.
Read-back is possible, but speed matters too
An obvious way to verify this would be to read the entire memory contents again after each personalization.
Why isn’t this done as a matter of course?
Armin Rinas:
Many readers are capable of performing a read-back. However, this slows down the process and is therefore not typically used.
This is precisely where an important technical trade-off between verification depth and throughput comes into play.
The reader in use can already detect whether the personalization process was completed successfully.
Our reader reports an error if personalization was not successful. We therefore assume that the data is correct in such cases.
This avoids the additional time required for a full read-back.
UID and camera verify the assignment
But a second question remains: Is the successfully written data record subsequently still unambiguously assigned to the correct card?
Armin Rinas: This is where another level of verification comes into play.
Using the UID, which is read by a second reader, and a camera, the software can check for data misalignment.
To do this, the machine combines multiple sources of information.
An additional RFID reader captures the unique UID of the medium. At the same time, a camera system checks the visible or machine-readable information on the card.
The software can compare this information with the expected production data record.
This means that quality control goes beyond simply verifying whether an RFID chip has been successfully programmed. It verifies the correct correspondence between the electronic identity, the physical medium, and the visible personalization.
This is precisely where the importance of system integration becomes apparent: the reader, camera, and production software must not only function individually; their data must be linked at the right moment.
Defective cards can be automatically re-produced
Can automation also be extended to handling errors?
Armin Rinas:
We have a mode in which defective cards are immediately reprinted, and the sequence must be maintained. However, this can slow down the machine quite significantly at times.
A defective card can thus be automatically ejected and replaced.
The technical challenge here isn’t just the production of the replacement card. The system must simultaneously ensure that the original sequence of data records is preserved.
This can involve both errors during personalization and problems detected by a scanner or camera system.
When quality control itself detects an error that isn’t really an error
Even modern camera inspection has its limits.
Armin Rinas:
In practice, sorted cards are sometimes checked by hand to determine whether they are usable. In some cases, pseudo-errors may occur, or the camera may have been unable to read the code.
This presents an interesting challenge for further automation.
A card may have been produced technically correctly and still be rejected by the inspection system—for example, because an optical code was not recognized under unfavorable conditions.
The next stage of development therefore involves not only performing more tests but also evaluating them more intelligently.
A fully automated line must be able to distinguish as reliably as possible whether a production defect actually exists or whether the inspection system has simply reached its detection limit.
Flexible systems must be capable of handling different technologies
Can personalization machines combine different processes?
Armin Rinas: Yes. Depending on the configuration, different approaches can be flexibly integrated into a single system.
These include RFID and NFC as well as contact-based chips, magnetic stripes, optical codes, and various printing methods.
Not every function is required for every job.
Generally, the machine can perform multiple functions if desired. However, they are not usually carried out all at once. It is up to the software’s process control to maintain the correct sequence.
This makes the software, along with the individual hardware modules, a crucial part of the machine architecture.
It determines which production job requires which station and the order in which the individual processing steps are carried out.
A system’s flexibility thus increasingly stems from the interplay of modular hardware and intelligent sequence control.
Laser personalization adds an extra layer of security
When is laser personalization the right choice for a card?
Armin Rinas: Laser personalization plays an important role when security requirements are particularly high.
The laser etches the data into the card or its surface. This cannot be tampered with—or can only be done with great effort.
That’s why this technology is used, for example, in bank cards and government cards. However, the additional security places higher demands on the card material. Rinas mentions, among other things, polycarbonate as well as corresponding overlays or additives.
Speed and costs also differ from other personalization methods.
Personalization is slower than with DOD. Compared to thermal printing, it is similar depending on the resolution and data volume. However, it is also more expensive than DOD printing.
DOD stands for “drop-on-demand” and is a digital inkjet process for the rapid personalization of variable data.
This makes it clear: Different applications require different technical approaches. A highly secure government card has different requirements than a ticket or a simple access card.
The machine developer’s task is to provide the right combination of security, speed, materials, and verification methods.
Ten cards or 50,000: The job must first be set up
Another trend is changing the requirements for personalization systems: smaller batch sizes and more highly customized products. Here, Rinas highlights a factor that is easily overlooked when focusing solely on production speed.
Armin Rinas:
Every job has to be set up once. If it’s needed again, the data has to be linked anew. The effort involved here is the same, whether for ten cards or 50,000.
With large production volumes, this effort is spread across many cards. With small orders, setting up the production job takes on much greater significance.
This also changes the definition of a high-performance personalization system.
It’s not just the number of cards per hour that counts. It’s becoming increasingly important how quickly a system can be adapted to new data sets, card types, personalization methods, and smaller batches.
Digital credentials and secure cards are evolving in parallel
d-you and the EUDI Wallet demonstrate how rapidly the world of digital identities is evolving.
However, this does not automatically make the physical card a technology of the past.
Rather, the requirements are also increasing in this area. Secure chips, variable personalization, lasers, optical inspection, and automated data matching continue to evolve. At the same time, production systems must be able to handle different media and smaller batch sizes with ever-greater flexibility.
Armin Rinas’s perspective makes it clear just how much engineering work is already required before a card even reaches the user.
Readers must write and verify data. Cameras must recognize the correct information. Printing and electronic personalization must remain synchronized. The software must track each card through the process and respond to errors.
In the end, it’s not just the maximum speed of a machine that matters.
What is crucial is that, even at high throughput, it is ensured at all times that the correct data set is assigned to the correct card.
This unique link between the physical medium and the digital identity is a central foundation for secure identification systems.
Sources:
