Good job!
Do you know about the built-in alert integration?
Some recommendations for you:
- Rather use state triggers and conditions, instead of the device ones. It’s more robust should you need to replace devices and makes a for a cleaner config.
- Use timers if you want this to be robust against an HA restart. Your automation would start a timer instead (one for each device). When it lapses, notify and restart the timer (another automation). A third automation would stop the timer once the issue is resolved.
- Your automation runs in single mode, so if you have the scenario where two sensors are triggered, this won’t work as expected. I would set it to parallel.