The Apple TV report and why it doesn’t settle the question
A headline this week reads: “Apple @ Work: Apple TV proves to be more reliable than Android in new digital signage report.” The detail isn’t public, but the claim isn’t surprising. Apple’s tvOS is a closed garden. Fewer models, fewer firmware branches, fewer third-party apps competing for resources. Reliability is easier when you control the whole stack.
That control comes at a cost. Apple TVs are not already on the wall in most offices, schools, or hospitals. Android TVs, Fire Sticks, and repurposed tablets are. They are already paid for, already networked, already mounted. The question isn’t whether Apple TV is more reliable; it’s whether the reliability gap justifies buying new hardware when existing screens can do the job.
What an Android TV can actually do as a signage endpoint
An Android TV or tablet can:
- Show a full-screen media player that survives a reboot.
- Cache content locally so playback continues through a network outage.
- Interrupt whatever is showing with an emergency alert, without user interaction.
- Send a screenshot back to the server so the dashboard shows what is actually on the wall.
- Poll the server every 60 seconds, so it works on ordinary shared hosting with no WebSocket support.
It cannot:
- Run as a system service, so it is visible in the recent-apps list.
- Start automatically on every possible device without at least one manual setup step.
- Guarantee the same performance across every model from every manufacturer.
- Prevent another app from being launched and taking over the screen.
The permissions that matter and how to grant them
Android’s “Display over other apps” permission (SYSTEM_ALERT_WINDOW) is the documented exemption that lets an app start in the foreground from the background. On a television it is granted once at setup, either by hand or by one adb command during provisioning. Without it, the screen will not interrupt a broadcast for an emergency alert.
| Permission | What it does | How it is granted | What happens if it is missing |
|---|---|---|---|
| SYSTEM_ALERT_WINDOW | Lets the app start in the foreground from the background | Manual tap or adb command | Emergency alerts will not interrupt what is showing |
| FOREGROUND_SERVICE | Keeps the player alive | Granted by the OS when the app starts | Player may be killed by the system after a few hours |
| INTERNET | Lets the app poll the server | Granted at install time | No content updates, no screenshots, no heartbeats |
On a television, SYSTEM_ALERT_WINDOW is the only permission that needs human intervention. On a tablet or phone, the user may also need to disable battery optimisation for the app, or the device may kill it after a few hours of inactivity.
How to work out the real cost against dedicated players
Dedicated signage players start at £150 per screen. They are purpose-built, so they are more reliable and easier to manage at scale. But they are not the only cost.
| Cost factor | Android TV/tablet | Dedicated player |
|---|---|---|
| Hardware | £0 (already owned) | £150+ per screen |
| Setup time | 5 to 10 minutes per screen | 5 to 10 minutes per screen |
| Network | Ordinary Wi-Fi or Ethernet | Ordinary Wi-Fi or Ethernet |
| Power | Uses existing TV power | Needs a socket |
| Content caching | Yes | Yes |
| Emergency interrupt | Yes | Yes |
| Remote screenshot | Yes | Yes |
| Per-screen licence | No | Usually yes |
The real cost difference is not the hardware. It is the licence. A per-site licence for self-hosted signage software costs the same whether there are 10 screens or 100. A per-screen licence for a cloud service or a dedicated player scales with the number of screens. For a site with 50 screens, the difference can be thousands of pounds a year.
When Android TV is the wrong choice
Android TV is the wrong choice when:
- The screens are in a public space where anyone can reach them. A locked-down dedicated player is harder to tamper with.
- The content is mission-critical and must show 24/7 without fail. A purpose-built player is less likely to crash or be killed by the system.
- The site has no IT staff to handle adb commands or troubleshoot permissions. A dedicated player is simpler to deploy.
- The screens are running other apps that must not be interrupted. Android’s permission model is not designed for absolute control.
When Android TV is the right choice
Android TV is the right choice when:
- The screens are already on the wall and already networked.
- The budget does not stretch to £150 per screen for new hardware.
- The site has IT staff who can handle adb commands or manual setup.
- The content is not mission-critical, or there is a fallback plan for outages.
- The licence cost must not scale with the number of screens.
How to decide for your site
- Count the screens you already have that run Android TV, Fire OS, or Android tablets.
- Check if they are in a secure location where only staff can reach them.
- Ask your IT team if they can handle adb commands or manual setup for permissions.
- Work out the licence cost for a per-site solution versus a per-screen solution.
- Decide if the reliability gap between Android and dedicated players is worth the hardware cost.
What to do next
If you already have Android TVs or tablets on site, install the software on one screen and test it for a week. Check that it can interrupt a broadcast for an emergency alert, that it survives a network outage, and that the dashboard shows the correct screenshot. If it works, roll it out to the rest. If it doesn’t, the dedicated players are the safer choice.
