FAQ

Short answers to what comes up on nearly every project. If yours is not here, the help desk answers the same day.

1.Running a house

Does the house still work without internet?

Yes. Everything that matters runs on the server in the house: panels, scenes, automations, all bus traffic. A house can be offline for months and the customer will not notice.

Only these need internet: remote access from outside, voice assistants, cloud backups, and software updates.

Do the wall panels work if the server is off?

No. Panels are screens for the server — if the server is off they have nothing to talk to. Physical KNX wall switches keep working, because that logic lives in the KNX actuators themselves, not in the server.

This is worth designing around: on a KNX job, wire the lighting so the basics work from the wall switches alone. Then the server adds scenes, automations and screens on top of a house that is never dead.

What happens during an update?

The server restarts, which takes well under a minute. Panels reconnect on their own. A panel-only update is quicker still — a few seconds of black screen, with the server untouched.

Can the customer break the configuration?

They can change what they are meant to change — scenes, schedules, favourites, names. Devices, protocol connections and flows are behind the installer login. You can also put a passcode on individual devices, which is the right move for door releases and gates.

2.Commissioning

Which systems can I mix in one house?

KNX, Z-Wave, Zigbee, Shelly, Ebelong, Modbus, and anything with an HTTP or MQTT interface — all at once, in any combination. The customer never sees which protocol a device uses.

A device switches, but the panel does not update. Why?

Almost always a missing status address. The platform wrote the command successfully, but nothing told it what the actuator actually did, so it shows what it last sent. Bind the actuator's status object to the same property.

The giveaway: switch the light at the physical wall switch instead of in the app. If the screen does not follow, it is the status address.

A blind runs the opposite way to what I asked.

Set the reverse flag on that device's position property rather than re-wiring or swapping addresses. One flag fixes the panel, Apple Home and Google Home together.

My device shows the wrong kind of card.

The type is decided from the device type you chose, its name, and the properties you bound — first match wins, and dimmer is tested before plain light, air conditioner before fan coil before thermostat.

Check the device-type name first: a type named "Fan Coil Controller" behaves differently from one named "Dimmer Module", and renaming a type changes every device using it. The full rules are in the device reference.

How long does commissioning take?

With an ETS project to import, a typical apartment is a day: import the addresses, build floors and rooms, create devices, bind properties, then scenes and automations. The binding is the slow part, which is why naming your KNX objects properly in ETS pays for itself.

Can I prepare a project before going on site?

Yes. Floors, rooms, devices and the ETS import can all be done in advance. On site you then only connect the bus and verify, instead of typing while standing on a ladder.

3.Apple Home, Google Home and voice

Does it work with Apple Home and Google Home?

Yes, both. You expose the devices you want from the panel's HomeKit and Google Home pages — not necessarily everything, and usually you should not: a short, well-named list is far more pleasant to use by voice than eighty devices.

Do I configure the bridges separately?

No. Both read the same device configuration, so a device that works on the panel works in Apple Home and Google Home with the same capabilities. Fix it once.

Does voice control need internet?

Google Home does — it goes through Google's servers. Apple Home works locally on the same network, and only needs internet when the customer is away from home.

Can voice disarm the alarm?

Only with a PIN. Google asks for a spoken code before disarming, which you set per installation. Arming needs no code — locking a house down is never the dangerous direction.

4.Cloud, remote access and data

What is the cloud portal for?

Keeping track of every house you look after: which servers are online, what version they run, and their backups. It is for you, not the customer — their experience is the panel and the app.

How does remote access work?

You request activation for a server from its page in the portal. Once it is active, that house gets a secure address you can reach from anywhere, with no router configuration and no port forwarding at the customer's end.

Where is the customer's data?

In the house, on their server. The cloud holds what you registered — the server's name, address and status history — plus any backups you chose to store. Day-to-day device states and usage stay local.

The privacy policy is the binding description of what we hold and why.

How often should I take backups?

After every visit where you changed something, and always before an update. It takes seconds. Restoring is deliberately safe: it adds what is missing and corrects what differs without deleting or duplicating, and tells you exactly what it changed.

5.When something breaks

The server shows offline in the portal.

Check power and network in the house first — that is the answer most of the time. The portal re-checks every server automatically, and the card tells you how long it has been unreachable.

If the house has no internet, a server can be perfectly healthy and still show offline in the portal. Remote access is what makes that distinction visible.

The server hardware failed. How bad is it?

Recoverable, if you have a backup. Install a server on the replacement hardware, restore the backup, reconnect the bus. The configuration — every device, scene and automation — comes back.

Without a backup it is a full re-commission. This is the entire argument for taking one after every visit.

An automation stopped working.

Open the flow and check it is deployed and enabled — each one has its own switch. Then watch the block values live on the canvas while you trigger it by hand; you will see which input never arrives.

Add an Error monitor block wired to a notification and the system will tell you about a failing flow before the customer does.

Is the alarm a certified security system?

No, and it should not be sold as one. It is automation: it arms, watches sensors, sounds a siren and sends notifications, which is genuinely useful and covers what most homes want.

What it is not is a graded, monitored, tamper-protected installation with a certificate behind it. Where a customer needs that — insurance requirements, commercial premises — install a certified panel and integrate it, rather than replacing it with this.

Same reasoning for life safety: smoke, gas and water sensors can be read and can trigger automations, but they are not a substitute for the certified detection a building code requires.
Still stuck? Ask the help desk — include the server name and what the protocol's connection page shows.