Morning before the first shift
The first check concerns the readers. Each reader covered by the wave should make one test punch with a new card and one with an old card, if the old format is still active. The result has to be seen not only on the reader (signal, gate opening) but also in InnerWeb, because InnerWeb is what proves that the event arrived.
The second check is the transmission of events. The access control server sends an HTTPS POST message to InnerWeb with each punch, and the integration token is valid for 12 hours, so in the morning it is worth confirming that the token has been renewed and that the last punch from the previous day is visible.
During card issuing
During issuing, the gatehouse makes sure that the number of the issued card is assigned to a person straight away, either through the CSV import or by hand. A punch from an unknown card goes to the queue of cards waiting to be assigned instead of creating an account from an unknown punch, so the queue is the best measure of a gap in the assignments.
If the queue grows faster than the gatehouse empties it, stop issuing and find the cause: most often the card number is in a different form than object_code-number, or the card comes from a different batch than the one included in the file.
Checklist points by time of day
After the change: checking working days
After the end of the shift check several working days of people who received a new card. The controller time and the event identifier protect the time records from distortion by the offline queue or by repeated transmission, but it is worth seeing for yourself whether the entry and the exit gave one working day with the correct hours.
Pay particular attention to readers on which the access control system does not pass the direction. InnerWeb pairs entry with exit in the software, with a limit of 13 hours, so a person who left without a punch will appear as a day with only an entry.
Contingency plan for the day of the wave
The contingency plan is short: a pool of spare cards, a list of people whose collection has been postponed, manual opening of the gate by the gatehouse and a record of that person in a sheet, which later goes to the correction of the time records. There is one condition: the old format is not yet switched off, so a person whose new card does not work can enter with the old one.
If the transmission of events from the access control system is interrupted, the controller keeps the punches in the offline queue and sends them again after the line is restored. Thanks to the event identifier they are not counted twice.
Comment on the difficult points
The most often overlooked point is the card export: if the access control supplier does not deliver the list of cards and readers before the wave, the gatehouse assigns cards by hand and the pace falls. The second is the reader dictionary, in which separate readers for entry and exit must have the right entries.
At the close of the day, write down three numbers: how many cards were issued, how many are in the queue to be assigned and how many people did not collect their card. These three numbers make it possible to judge the pace of the next waves.
Checklist
Video
Frequently asked questions
Who should be present at the switchover of a wave?
The gatehouse, the access control administrator and the person responsible for the time records. The access control supplier should be reachable by phone during the issuing hours.
What do I do when the queue of cards waiting to be assigned grows?
Pause issuing, compare the number format with the import file and assign the cards from the queue. Resume issuing only after the queue is emptied.
Can a wave fall on Monday morning?
It is better to avoid it. Monday morning is the peak of entries, so card issuing is planned in low-traffic windows.
Let's go through the list together before the first wave
In a short conversation we will match the list to your access control system, the layout of the gates and the number of gatehouse workstations.
Let's talk