Process automation and quality check for Secrid wallets

Students: Roel op ‘t Hoog, Theo Tromp, Jorge Luis Díaz Onofre, Florens Vernooij

Screen Shot 2018-06-20 at 13.21.56

Figure 1. Secrid logo



Figure 2. Secrid wallet

For the minor Smart manufacturing and robotics we did a project for Secrid. Secrid makes card protectors which consists of an aluminum casing in which cards are stored. A mechanism at the bottom of the casing allows for the cards to be slid out. With more cards being equipped with wireless communications, both wanted and unwanted connections can be easily established. These chips can be read, activated and cloned from a meter away, without you noticing. These wallets are designed to block RFID (Radio-Frequency IDentification) or NFC (Near Field Communication), protecting cards from unwanted wireless communication.

Global overview current assembling process

The current process starts at an extruder in Helmond, this company extrudes aluminum billets into profiles of 6m. These profiles are shipped to The Hague to finalize the end-product. When they arrive at the workshop of Secrid a quality check happens by sampling the batch. These samples undergo two different checks, deciding if the batch is accepted or needs to be repaired. The most occurring problem is that the internal height of the profile isn’t according to the design dimensions, which results in the cards being unable to be put in the protector.

The approved profiles get anodized to give them strength for the cutting, they are cut into the desired length. The semi-finished card protector gets de-burred and de- anodized before they are anodized into the wanted final color.
If these dimensions cannot be achieved, the products will be bent manually by pushing the piece in a rotating device and rotate it.

When the card protectors have been anodized, they are assembled by a social workplace in The Hague to get the final (basic) product as can be seen in figure 2.

Problem description

The consistency of the inner dimension, is something that Secrid wants to improve. For the quality check Secrid picks now a random length out of the truck and measure that one. If this one is accepted, then they assume that the rest of the batch will have the same dimensions. This working method is based on assumptions, and therefore they do not have their production line perfectly under control.

Our solution

For the project, of the minor smart manufacturing and robotics, we want to help Secrid to control their production line better. One way to gain more control over their production line is by making sure that all the sawed & de-burred aluminum pieces are according to the same size. Our way to do that is by making every aluminum extrusion profiles standard a bit too small, and afterward deep draw or stamp them into the right shape. Afterward, the pieces will be checked with a vision system, and the pieces will be placed into the corresponding output buffer.

Mayor decisions

At first, we started with the idea to make a working machine which was placeable directly into the production facility at Secrid. After 2 weeks we concluded that this was too much to get done within our timeframe, so we switched to making a proof of principle, of which a design can be seen in the figure below. The key points that we would like to show were the singulator (1), automation of the bending system (2) and a vision system (3) in a way that they are applicable so that they can be integrated into high quality and high-speed system. These 3 key things are demonstrated on the presentation day.

Screen Shot 2018-06-20 at 12.59.46

Figure 3: Design of the end-system


The singulator is made in such a way that the system could easily be implemented into a system like the design. A stepper motor rotates two holding rogs 90°, dispensing one card protector. Afterwards a piston pushes this profile out into the bending system.

Screen Shot 2018-06-20 at 12.59.54

Figure 4: Singulator

Automation of bending system

The system used to push the profiles into the desired shape comes from Secrid, this system was used to fix the rejected batches. Since this process is being done manually, it had to be automated. This meant that the mechanical button had to be replaced for an electrical one which could be connected to the process controller.

Screen Shot 2018-06-20 at 13.22.15

Figure 5: Proof of principle automation bending system

Vision system

Secrid currently uses a vision system that has to be loaded manually. An employee puts an aluminum profile in the setup and presses a button to check it’s values. Their system utilizes backlight to see a clear contour of the profile which then can be measured accurately. For the design mentioned under ‘Mayor decisions’, this method cannot easily be used, as pictures will be taken while the profile is still on the bending system. While not required – as using a robot for transportation circumvents this problem – a new system was created to work towards a solution that could be used in a final setup.

This new system uses the reflection of light on an object to get a clear view of a face of the object. With the clear distinction between the face and the surrounding, measurements can be made through software.

Screen Shot 2018-06-20 at 13.22.24

Figure 6: Image of the card protector made by the vision system

Through some image manipulation all visual noise can be deleted and a clear outline can be found.

Screen Shot 2018-06-20 at 13.22.30

Figure 7: Contours of the card protector