Data Watchdog
Know when your Homey data stops arriving.
Data Watchdog is a new Homey community app by Arrow87 Home, focused on monitoring the reliability of device and integration data.
I built it after running into a simple but difficult smart-home problem: a device can keep showing its last known value even when the underlying integration has silently stopped delivering new data. At a glance, everything can still look perfectly normal.
Data Watchdog is designed to detect that situation using explicit evidence of activity or data delivery, rather than assuming that an unchanged sensor value means something is broken.
What it does
- Monitors selected devices and integrations for proven activity or data delivery
- Detects devices that become stale
- Can group related stale devices into an integration-level incident
- Provides Flow triggers for failures and recoveries
- Includes a local simulation device for testing watchdog behaviour
- Runs locally on Homey
Platforms
The primary target is Homey Self-Hosted Server, with support for current local Homey Pro platforms as well.
Languages
Data Watchdog is available in English and Dutch, including the app interface, settings and user-facing text.
Current status
Data Watchdog is currently in Test and is being prepared for Homey App Store certification.
The first release is intentionally focused on reliable monitoring and clear failure states rather than automatic recovery or cloud monitoring.
Why this app exists
A sensor value that does not change is not necessarily stale data. For example, a temperature of 20.5 °C may legitimately stay at 20.5 °C for quite some time.
Data Watchdog therefore monitors evidence such as an explicit delivery timestamp, device activity reported by the source integration, or a confirmed heartbeat.
The goal is not to guess whether data is fresh — it is to only alert when there is enough evidence to justify it.
Links
Arrow87 Home: https://arrow87-home.github.io
Developer: Arrow87
I’ll use this topic for development updates, release notes, questions and feedback as the app moves toward its first public release.