Random missed automation actions, logs look fine though

[VanderMueller]

Has anyone dealt with that annoying situation where your automation is clearly triggering, but the device acts like it didn’t hear you? I’m trying not to overthink it, but it keeps happening just enough that I’m losing trust. I’ll see the automation ran, and then one device will randomly skip the command even though nothing else seems wrong.

This happened to me earlier this week when I walked into the hallway and the motion automation fired, but one light stayed dark and made me feel like I broke something. Then the next day it worked perfectly and I felt silly. I’d love a simple troubleshooting order so I stop guessing. Do you usually look at cloud dependence first, then hub performance, then mesh signal quality, or do you start with the device itself like a flaky plug or bulb that needs a reset?

 

[ValimarKnight]

I’ve seen this and it usually ends up being device level, not the automation logic. My quick order is check if the missed device shows as unavailable or has weak link quality at that moment, then check if it was recently power cycled, then try sending the same command manually right after the miss. If manual also fails once in a while, I stop blaming the automation and start watching that device or that part of the mesh.

 

[Alan]

This is super common and usually ends up being one flaky endpoint, not the automation logic. Quick order that saves time: verify if the miss is always the same device, then check its power state and connectivity, then remove and re-add it. If it’s the same bulb every time, treat it like a device problem first, not a rule problem.