Hey gang, I've got a chunk of free time lately, and I've been working on some of the backlog issues I've had with my HA instance. The one that is giving me trouble right now is my thermostat - I use Honeywell total connect (or whatever its called), and it works just fine when using the normal thermostat card or controls. However, I want it to be warmer in the day, and colder at night. So I had been using a scheduler entity from the HACS store. It always used to work, but lately I've been getting out of bed and realizing the temp is still set to the nighttime temp.

It's not every day, and it seems to work 90% of the time, but I had always thought that the scheduler entities did a periodic check to see if the thing they controlled was at the proper state? Seems like if the scheduler 'misses' the switchover time, it's just stuck at the night time temp all day unless I manually change it.

So this got me thinking... Is there a better, or more 'approved' way to do this sort of thing?

I have a few old automations that are designed to unlock my front door, or open my garage door, depending on how I leave. If I ride my bike or car, it opens my garage, if I leave via front door, it opens the front door. Pretty simple. Except!!! I did this in pieces, so it's 2 separate automations. One detects my phone entering the "home zone", the other detects an event firing (iOS triggered the event via 'shortcuts')

There is now "zone based" automations, and there is also my "person entity". So the way I see it, I have 3 different ways to tell when I come home:

1. My phone's gps enters the 'home zone'

2. My HA 'person' state changes from 'away' to 'home'

3. My phone's shortcut app fires an event that is detected.

I feel like #1 and #2 are the same, no? I only have one device linked to my person entity, and it's my phone. Is there any difference in this case? Is there a preferred choice?

I have an automation that when triggered turns on a scene that sets my lights a specific way. Then the automation waits for a second trigger.

I noticed that the automation doesn’t continue after the second trigger, so I looked at the trace, and it looks like the scene setting step of the automation times out: “error: Timed out when calling async_turn_on for bulb xyz: <class ‘TimeoutError’>

It looks like this bulb xyz was dropped from my zigbee network, which isn’t always the most stable. So it’s not exactly uncommon.

So how do I safeguard my automations from being broken by a single piece of a scene being offline?