InnerWeb
Skip to main content

Cold store biometrics replaced with a card and PIN: a food plant scenario

A fingerprint reader at the cold store entrance rejects staff wearing gloves, with damp hands and cold skin. The security manager sees a queue at the door and an open door in a zone that biometrics was meant to protect. Below is a scenario based on typical food plant conditions: how to assess whether biometrics is necessary and what can replace it.

120
entries per day in the calculation assumption
Source: example calculation
10
rejections per day at an 8% reader error rate
Source: example calculation
1
rejections per day at a 1% card error rate
Source: example calculation

Situation: a food plant and a cold store with a biometric reader

This is a scenario, not a description of a specific deployment. A food plant has a cold store with restricted access, because it holds raw material that needs constant temperature. A fingerprint reader was installed at the entrance, selected at the tender stage when the team considered biometrics naturally secure.

After a few weeks of use, staff take off their gloves to place a finger on the reader, and in damp air the skin gives a weak reading. Some people hold the door open or ask a colleague to open it. The mechanism meant to restrict access is in practice being bypassed.

Problem: false rejections and bypassing the control

False rejections are more than an inconvenience. Every workaround, such as a propped door or following a colleague in, lowers protection below what a simple card would provide. A biometric system with a high rejection rate becomes weaker in practice than an ordinary proximity card.

The second issue is legal. Article 22(1b) of the Labour Code links biometrics of employees to access control for particularly important information or rooms requiring special protection. The employer must demonstrate that such a solution is necessary, not merely convenient or cheaper. When a reader regularly fails, it is hard to argue that it is necessary.

  • protective gloves and humidity reduce fingerprint reading quality
  • a queue at the door encourages propping it open and following a colleague in
  • the rejection log stops being reliable evidence of who actually entered
  • biometric templates remain special category data even though the device does not do its job

Rejections per day at 120 entries, example

Rejections per day at 120 entries, exampleBar chart: about 10 rejections per day for the fingerprint reader and about 1 for the proximity card.0246810rejections per dayFingerprint reader: 10 rejections per day10Fingerprint readerProximity card: 1 rejections per day1Proximity card
Source: Example calculation: 120 entries per day, 8% biometric rejections for gloved staff and 1% card rejections are illustrative assumptions.

Deciding necessity: what to check before the decision

The data minimisation principle in Article 5(1)(c) GDPR says data must be limited to what is necessary for the purpose. The purpose here is controlling entry to the cold store. The question is whether a less intrusive measure achieves the same purpose.

In a typical plant the answer is yes. 13.56 MHz proximity cards with DESFire encryption and QR codes are a full alternative to biometrics for areas with a standard level of protection. A cold store holding food raw material primarily needs accountability of entries, not the uniqueness of a body feature.

Questions for the security manager

Before the team keeps biometrics, it should record answers to three questions and attach them to the impact assessment.

Does the cold store hold information or materials requiring special protection within the meaning of the provision? Is the actual rejection rate acceptable for work in gloves? Would a card with a PIN give the same level of accountability?

Solution: a proximity card and PIN instead of a fingerprint

An encrypted card works through a glove within the reader range, and a PIN adds a second factor in places where protection should be higher. Removing gloves is no longer needed, so the main source of rejections disappears.

InnerWeb does not manufacture or replace readers. It is a software producer and integrates with the plant's existing access control system. An access event can consist of a person identifier and a timestamp, so InnerWeb needs no biometric template for it. It receives the event via webhook and links it with qualifications, permits and attendance.

  • changing the reader type is a matter for the plant's access control (KD) system, not InnerWeb
  • biometric templates from the cold store are deleted after the switch to a card
  • entry to the cold store still records the event with time and person
  • a PIN is used only where the risk assessment justifies a second factor

Effect: an example calculation of rejections and queues

Below is an example calculation based on assumptions that the plant should replace with its own measurements from the reader log. It assumes 120 entries per day, 8% false rejections of the fingerprint reader for gloved staff and 1% rejections of the card. These are illustrative assumptions, not a measured result.

With 120 entries, 8% means about 10 rejections per day, that is 10 situations where a person tries again or someone opens the door from the other side. A card at 1% means about 1 rejection per day. The difference of nine events a day is nine fewer opportunities to bypass the control.

Conclusions for plants facing a similar problem

Biometrics in a cold and humid environment needs testing under real conditions before it goes into the specification. One week of rejection logs with gloved work gives more evidence than the device's catalogue card.

If the test shows a high rejection rate, switching to a card is not only practical but also supports the data minimisation principle and justifies why the employer does not keep a special category of data where it is not necessary.

Example calculation

Rejections at the cold store entrance, example calculation

Assumptions
ParameterValue
Entries to the cold store120 per day
Fingerprint reader rejections for gloved staff8 %
Proximity card rejections1 %

Result: About 10 rejections per day for biometrics against about 1 for the card, that is 9 fewer opportunities per day to bypass the control.

Note: Example calculation based on illustrative assumptions; the plant should substitute its own rejection rate from the reader log.

Video

Frequently asked questions

Can biometrics be used in a cold store at all?

Yes, if the employer demonstrates necessity and meets GDPR and Labour Code requirements. In a typical food plant a card or QR code is sufficient, so biometrics requires a strong justification.

Is a PIN required together with the card?

Not always. A PIN as a second factor is used where the risk assessment justifies a higher level of protection, for example for raw material requiring special supervision.

Does InnerWeb replace readers in the plant?

No. InnerWeb is a software producer and integrates with the existing access control system. Readers are selected and replaced by the supplier of the access control (KD) system.

What about templates collected earlier?

After giving up biometrics in a given zone, the administrator should delete the templates and document this in the data protection records.

Sources

Let us check entries to special zones

Book a short call on integrating card entry events with InnerWeb. We will show how to connect them with permits and attendance without biometric templates.

Let's talk