InnerWeb
Skip to main content

One card for access control and RCP: implementation order and pilot pitfalls

The plant has a working access control system and wants the same card to serve time attendance and trainings. The order of the work decides whether the pilot ends with a result or with weeks of corrections. Below are the stages, conditions and pitfalls the IT manager should know before the first gate.

3 characters
minimum for the person search at a station with a card
Source: description of the card search function
13
maximum number of search results
Source: description of the card search function

Stage 1: tidy up the accounts before the first punch

The first pitfall is two active accounts of the same person. When that happens, the system moves the training result to the account with the card, so the gate stops reporting missing training. It does not merge accounts by surname alone, so the HR data must be put in order by the plant before the punches start flowing.

Run a review: how many employees have more than one active account, whether card numbers are assigned to the right people, and whether lost cards have the status inactive. Without this, the attendance records will reproduce errors that already exist in the data.

Stage 2: specification for the access control supplier

InnerWeb neither produces nor replaces access control readers. The access control supplier receives the punch event specification and configures the sending on its side. The event goes to a webhook authenticated with a technical account and a JWT token valid for 12 hours.

The message contains the punch time from the controller, the reader identifier, the person and the card number in the form object_code-number. Optionally, the direction of passage and the event identifier are added, which provides idempotency. It is worth agreeing that this identifier is mandatory, because it protects against double punches after a retransmission.

Example pilot schedule in weeks (example calculation)

Example pilot schedule in weeks (example calculation)Example schedule: two weeks for accounts, three for the KD configuration, four for the pilot and five for the extension.012345weeksOrder in accounts: 2 weeks2 weeksOrder in accountsKD configuration: 3 weeks3 weeksKD configurationPilot: 4 weeks4 weeksPilotExtension: 5 weeks5 weeksExtension
Source: Example calculation: assumed stage times, to be verified by you

Stage 3: RCP and trainings on the same scans

Gate scans form the working time records: daily, monthly and yearly views, exception queues, period closing and reminders about a missed exit. The module is disabled by default, so enabling it requires an HR decision and agreement on the period closing rules.

At a training station with RFID card integration, the employee punches the card or selects the company and types part of the surname. The search requires at least 3 characters, returns up to 13 results, and shows only active accounts with an active card, without revealing the card number. Effectiveness therefore depends on whether the cards are assigned and active.

What will block the pilot

From implementation experience, four things most often block the pilot. All of them can be removed before the start.

  • the access control supplier has no deadline for configuring the event sending,
  • the token of the technical account expires after 12 hours, and the supplier does not renew it automatically,
  • card numbers in the access control system and accounts in the system have different formats,
  • the RCP module was switched on without deciding who handles the exception queue.

Pilot scope and success criteria

It is best to start with one gate or one hall and a group of a few dozen people. Write the success criteria down before the start: a punch reaches the system within the agreed time, a repeated event does not create a second punch, and the exception list is handled by a named person.

After the pilot, decide on extension to further readers. Remember that the same punch feeds the evacuation list, refreshed every 13 seconds, so the tests should also cover an evacuation drill.

After the implementation: maintenance

Decide who monitors the sending of events and what happens when the connection fails. The access control supplier may resend events from the outage period, and the event identifier allows repeats to be rejected. Once a quarter it is worth reviewing the exception queue for recurring causes that point to errors in the card data.

Video

Frequently asked questions

Do we need to buy extra readers for RCP?

No. RCP uses punches passed on by the existing access control system. InnerWeb does not replace readers.

What if the same employee has two accounts?

The system moves the training result to the account with the card, so the gate stops reporting missing training, but it does not merge accounts by surname. Keeping the accounts in order remains the plant's responsibility.

Does the RCP module work immediately after implementation?

No, it is disabled by default. It is switched on after the rules have been agreed with HR.

Preparing an access control and RCP pilot?

We will help plan the stages and prepare the specification for your access control supplier. Book a meeting with the IT department.

Let's talk