InnerWeb
Skip to main content

Entry control pilot at water network sites: implementation notes

Entry control at water network sites is best introduced at a few chosen sites first, and only then extended across the whole network. Below is the order of steps for the maintenance head: choosing the sites, agreeing with IT and the KD vendor, a parallel period with paper, criteria for moving on, and the pitfalls that surface in the first weeks.

3–4
sites in the pilot in different conditions (recommendation)
Source: sample calculation

Stage 1: choosing the sites for the pilot

Choose three or four sites that together show all the variants you will meet in the network. One site with connectivity and a permanent network connection, one without coverage, one often visited by external companies and one with an existing access control system. Such a mix reveals problems on a small scale.

Do not choose the easy sites, because a pilot that passes without objections teaches nothing. Nor should you choose the hardest site in the network, because the first experiences will discourage the team before the system has a chance to prove its value.

Stage 2: technical agreements

For a site with access control, the first step is a conversation with the KD vendor. InnerWeb receives card swipe events as an HTTPS message initiated by the plant's site on port 443, without opening incoming ports. Points still to settle are the possibility of sending HTTPS POST requests, the list of entry and exit readers, and the card export. The integration does not require replacing cards or readers.

For sites without KD, a QR code at the entrance is enough. A scan of a QR badge or an RFID card gives a verdict on whether the person may be on site, and the scan itself records the entry. Prepare the badges and the places for the codes before the first service technician arrives with the new rule.

Sample calculation: discrepancies between the register and the logbook during the pilot

Sample calculation: discrepancies between the register and the logbook during the pilotLine chart: discrepancies of 12, 6, 2 and 0 in each of the four weeks of the pilot.02.557.51012.5discrepanciesWeek 1Week 2Week 3Week 4Week 1: 12 discrepancies12Week 2: 6 discrepancies6Week 3: 2 discrepancies2Week 4: 0 discrepancies0
Source: Sample calculation: an assumed decline in discrepancies at four sites over a four-week parallel period

Stage 3: work without connectivity and the firewall

At a site without Wi-Fi, actions that change data are queued locally in the device's browser and replayed once the network returns, and repeats do not create duplicates. Check this in practice during the pilot: a technician makes a visit without coverage, and after returning you confirm that all the entries arrived.

Agree with the IT department what happens to alarms when the plant's network blocks outgoing traffic. Alarm webhooks are then stored durably, and an external poller collects them over an incoming channel. The alarm is not lost, but the poller has to be configured.

Stage 4: a parallel period with paper

For two to four weeks, keep the entry logbook and the digital register side by side. Compare them every week. The discrepancies will show where people have not yet got used to the scan, and where the system has a gap. Record every discrepancy with its cause.

Do not extend the parallel period indefinitely. Once discrepancies stop appearing for two weeks in a row, the logbook can be withdrawn at the pilot sites.

Stage 5: criteria for moving on to the whole network

Before the rollout, set the criteria in writing. Examples: full agreement between the digital register and the paper for two consecutive weeks, no lost entries after visits without coverage, a satisfactory entry time at the gate, and confirmation from the IT department that the connection needs no changes.

If any criterion is not met, the rollout waits. Changes to access rules made during the pilot are saved with their author and time in the log, so it is easy to show later what was changed and why.

  • agreement of the register with the logbook for two weeks in a row
  • no lost entries after visits without coverage
  • confirmation from the IT department on the connections
  • acceptance of the procedure by the dispatcher and the crews

Pitfalls of the first weeks

The most common pitfalls are insufficient training of technicians from external companies, QR codes damaged by weather, and no one responsible for the pilot on a daily basis. The team should have one coordinator with the authority to see the discrepancies and settle them.

Also prepare one page of answers for the technicians: what to do when the scan does not work, where to report a problem, and who the dispatcher is. This page saves more time than any training meeting.

Video

Frequently asked questions

How many sites should the pilot cover?

Three or four, chosen to show different variants: with and without connectivity, with external companies, with and without existing KD.

How long should paper and the system run in parallel?

Usually from two to four weeks, until discrepancies stop appearing for two weeks in a row.

Does the pilot require changes to the firewall?

Usually not, because the connection is outgoing on port 443. If outgoing traffic is blocked, the poller collects alarms over an incoming channel.

Do new readers need to be bought?

No. The integration with access control does not require replacing cards or readers.

Let's plan your pilot

We will help you choose the sites, write down the criteria for moving on and prepare a one-page guide for the technicians.

Let's talk