Fire Safety Alerts: Where Screens Work and Where They Fail

Your screens might not interrupt ads when fire strikes, here’s why

A fire alarm triggers, but the nearest screen still shows an ad for a staff discount. The delay isn’t down to slow servers or weak signals, it’s because the screen’s software treats alerts like any other scheduled content. Some digital signage systems queue messages, waiting for the current loop to finish before interrupting. A 30-second promo becomes a 30-second barrier. Even if the system is legally compliant, the law doesn’t dictate how software handles interruptions, only that an alert eventually appears. By the time it does, the building may already be evacuated through the wrong exit.

The problem worsens with ad overlays. Many screens use third-party apps to display promotions, news feeds or wayfinding. These apps often request the same permission that lets fire alerts take over: display over other apps. If the app isn’t granted that permission, it runs in the background while the screen’s native player shows its own content. But if the app is granted it, and many are, it can override the alert system entirely. A screen set to show a rotating ad feed might never acknowledge a fire message, because the ad app has priority. The server logs will show the alert was sent. The screen will show nothing but the latest promotion.

Auditing for these failures starts with checking which apps are active on each device. Some signage platforms list installed software, but others don’t. A manual check, unplugging the screen, booting into safe mode and reviewing installed apps, reveals what’s competing for control. The second step is testing. A simulated alert must interrupt every screen, not just the ones running the signage software. If a device ignores it, the cause is either an app with higher priority or a misconfigured permission. The fix isn’t always removing the app. Some overlays, like digital clocks or temperature displays, are critical. The solution is to ensure the alert system has the highest priority in Android’s overlay hierarchy, or to replace the device with one that can’t be overridden.

The risk isn’t theoretical. A fire in a London office block saw staff follow signs to an exit that had been blocked for maintenance. The digital wayfinding screens, set to cycle through ads and directions, never updated to show the change. No law required them to. No one tested whether they would. The delay didn’t come from poor wiring or a failed server. It came from software that treated fire alerts like background noise.

The Fire Safety Act 2021 added ‘mass notification’, but who actually tests it?

Mass notification testing is where most estates fail, not because the screens cannot interrupt an ad or a safety loop, but because the act of testing itself reveals hidden costs. A single dry run across 50 screens, each running a different app or locked into a playlist, turns into a week of chasing down permissions, rebooting devices and reconciling logs. The responsible person must confirm every screen received the alert, but if half the estate runs third-party software or custom kiosks, that confirmation becomes a manual audit. Most managers skip the test until the fire drill, when the screens that should have shown the alert instead display yesterday’s meeting schedule.

The problem is not the screens themselves. It is the estate’s inability to prove, in advance, which ones will actually interrupt. A network of mixed devices, some with the overlay permission, others without, means the panel cannot tell, until it is too late, which screens will fail. The choice is between two bad options: either trust that every screen will work (and risk the audit later) or treat each device as a black box (and accept that testing takes longer than the law demands). Neither approach scales.

Here is how the trade-offs compare:

The only way to avoid the second-order cost is to know, before the test, which screens will work. That requires a panel that flags screens that cannot draw over another app before you need them. Without it, the dry run becomes a gamble: either you spend the time proving compliance, or you spend it fixing the screens that did not show the alert when it mattered. Most estates choose the latter.

A single screen in the wrong place turns your system into a liability

A screen in a corridor behind a closed fire door does not help anyone. The alert arrives in seconds, but the door blocks it. The same happens if a screen is mounted above a high counter, or if it faces a wall instead of a clear line of sight. A single misplaced screen turns what should be a safety system into a false sense of security.

The problem is not just that the alert is hidden, it is that the system cannot prove it has worked. If the screen acknowledges receipt but no one sees it, the panel shows a green tick. The responsible person checks the panel, sees all screens reporting, and assumes the alert reached everyone. They have no way of knowing that half the building missed it.

You do not need a full site audit to find these spots. Start with the evacuation routes. Walk them as if you are leaving the building for the last time. Look for screens that would be obscured by doors, furniture or blind corners. If a screen is mounted where staff or visitors would not notice it in an emergency, it fails before the fire starts.

A quick test confirms it: ask a colleague to stand where a member of the public would be during an evacuation. Close any doors that might be shut in a real alert. If they cannot see the screen from that position, the screen is useless. The same applies to screens in stairwells or lobbies where lighting or reflections reduce visibility.

The fix is not always to move the screen. Sometimes it is to adjust its position by a few centimetres, tilting it down, swapping it to the other side of the wall, or adding a reflective surface to direct attention. The goal is to ensure that when the alert appears, someone in the room sees it. If they do not, the system has not done its job.

Power and network failures won’t be your problem, until they are

When a screen loses power or its network connection, the system switches to cached content. That is not a failure, it is the design. The problem comes later, when the backup itself starts to degrade without anyone noticing.

Batteries in uninterruptible power supplies lose capacity over time. A UPS that held a charge for two hours in its first year might last only 90 minutes three years later. The screen itself has no way to warn you: it simply plays its last downloaded content until the battery is empty. By then, the alert you needed has already passed, and the screen is stuck on a stale loop while the building empties.

WiFi credentials expire too. A screen that cached its media in January will still try to reconnect in April, but if the SSID password was changed in February, it will fail silently. The device has no log of the attempt, only the empty screen, which now shows nothing at all. Even if the network returns, the screen remains offline until someone manually checks each one.

Automated checks can catch these issues before they matter. A daily poll from the server forces every screen to report its battery level, its last successful connection time, and whether its cached content is still valid. If a screen’s UPS drops in capacity, the panel flags it. If a device has not refreshed its WiFi handshake in over a week, it triggers an alert. The system does not guess at failures, it measures them directly. A screen that cannot report its own state is treated as broken, and the panel shows exactly which ones need attention.

The trade-off is clear: you cannot predict every hardware failure, but you can make sure the system tells you when the backup it relies on is no longer working. The difference between a screen that plays cached content and one that plays nothing is not the outage, it is the check that follows.

The ‘responsible person’ clause hides a paperwork trap for facilities managers

The ‘responsible person’ clause in fire safety law does not just require evacuation drills, it demands proof they worked. That proof starts with records of every test, including those using digital screens. The moment a drill is logged in a spreadsheet or printed form, it becomes a liability. A single misplaced sheet, a lost file or an unchecked box turns a compliant system into a gap in the audit trail.

Digital signage avoids this by tying compliance logs directly to the alerts themselves. When a screen displays an evacuation message, it also timestamps the event and marks it as tested. This record is not stored on a separate system or buried in an email chain. Instead, it updates automatically in the same software used to manage the screens, linking each drill to the device, the time and the person who approved it. No manual entry means no missed entries.

The risk comes when managers assume the software handles the paperwork. They schedule drills, watch the screens flash, then move on, until an inspector asks for the logs. Without an integrated system, those records may sit in a folder labelled “2023 Fire Tests” while the actual test dates are scattered across calendars and notes. A digital signage system that logs tests as they happen eliminates this gap. The software does not just show the alert; it proves it was delivered, when, and by which screen.

The trade-off is visibility. A manager must configure the system to require approval before clearing a test alert. Skipping this step leaves the log incomplete. But the alternative, relying on memory or separate forms, is far riskier. The law does not care how busy you were. It only cares whether the evidence exists.

Three months in, your screens will start ignoring fire alerts, here’s how

Three months after installation, screens that once interrupted everything for fire alerts will start ignoring them, not because the system fails, but because something else changes. The most common cause is an update to the device’s operating system. Android’s security patches or manufacturer firmware often reset permissions, including the one that lets alerts take over the display. Without it, the screen plays its scheduled content and treats fire alerts like any other notification: a brief banner that fades away.

Staff also disable alerts after false triggers. A sprinkler test, a faulty smoke detector or even a misconfigured server can flood screens with alerts. Someone in the building will see the interruption as unnecessary and turn it off. The device then remembers that choice. A single screen in a quiet corner, left alone, may still work. The rest follow the first one that was silenced.

Neither issue is hard to fix, but both require action. The first step is to check permissions after every update. A locked-down device will not let users revoke the “display over other apps” setting, but some do. The second is to audit screens regularly. A screen that shows its last received alert as “evacuation drill” instead of “fire alarm test” has been disabled. The panel in the control room will flag it, but only if someone looks.

The trade-off is clear: a system that works only if someone maintains it. Three months in, the screens will stop working unless someone ensures they haven’t.

The hidden cost of ‘compliant’ screens: maintenance you never budgeted for

A fire alarm test once a month is the easy part. The hidden costs come after, when screens stop working as expected. A mount fails under the weight of a heavy television. A staff member leaves without retraining their replacement on the new alert tone. A screen that once showed evacuation routes now loops a broken video, and no one notices until the fire drill. These are not hypotheticals, they are the gaps that turn a compliant system into a maintenance burden.

Replacing a mount is straightforward, but it exposes another issue: the screen itself may no longer be in the right place. A television shifted by a cleaner, or a monitor unplugged during a power test, means the alert no longer reaches the people who need it. The panel shows the screen as "online," but the overlay permission, what lets an alert interrupt anything playing, has been reset after a factory reset or a botched update. Without a way to check which screens can still cover another app, you won’t know until the next drill, when half the estate ignores the test.

Retraining staff is worse. The person who triggers the fire alert must recognise the tone, the flashing screen, and the exact steps to confirm it. If that person moves departments, or worse, leaves the company, no one documents who takes over. A single missed training session means the system fails when it matters. Even worse, if the alert tone changes (because the manufacturer updated the app, or because the responsible person insisted on a different sound), no one tells the night shift. The result? A silent screen during an actual emergency.

The real cost isn’t the screens themselves. It’s the time spent chasing failures, the paperwork to prove compliance, and the panic when the system doesn’t work as planned. A screen that loops an ad during a fire drill isn’t just non-compliant, it’s a liability. The only way to avoid it is to track which screens can still interrupt content, who is trained to use them, and where they physically sit. That starts with a panel that shows you exactly what’s on every display, not just whether it’s connected.

Ask for a live demo where you can see the overlay permission status for every screen in your estate. If the system can’t tell you which ones will ignore an alert, it’s already costing you more than you realise. The next step is simple: book a trial on your own hardware. Only then will you see the difference between a screen that looks compliant and one that actually works.