The SmartThings Development team is ready to begin the first phase of local scene execution on a hub for current enrollees of the Hub Firmware Beta Group on Thursday, June 30. The second phase will be available soon and support routines that execute locally when all associated scenes are also running locally.
With this beta release, you can start testing this new feature by creating a new scene using the SmartThings app and then running it.
This feature is enabled using the Location for your hub. If the Location is changed, the feature will not be automatically enabled at this time.
Scenes are eligible for local execution when all associated Hub Connected Devices can execute locally using Edge drivers and a limited set of DTH devices. They include Zigbee, Z-Wave, and LAN devices as well as other Hub Connected Devices using Drivers. Many of the existing scenes have already been examined for eligibility, and if they are not enabled for local execution, you will need to recreate or modify them.
You can use the SmartThings app or the Scenes tutorial to create, view, and execute scenes. The second tutorial shows you how to verify that the execution location is “Local”.
Not quite. I get that the API can tell me whether a scene is running locally. But how do I actually test that? If I disconnect my hub from the internet it can no longer receive API or mobile app commands to run. So how do I trigger a scene to run if my hub doesn’t have internet to validate it is indeed running locally?
I think the easiest way to do it is to create a routine that calls the scene. It should also run locally because as you mention, the app and API don’t support local connection.
I’ll confirm with the team.
We used the list of hub IDs for the current enrollees of the Hub Firmware Beta Group and manually enabled this feature for its current Location. If you were enrolled in this group as of yesterday, this feature is enabled. It’s a manual process to verify that your Location was enabled correctly.
The feature will improve performance and reduce latency when performing 1 or more actions in a scene. In order to test this feature, you could create a scene to turn on multiple lights and then observe a smaller delay between each action.
This is actually a good point: will local scenes work with local Groovy SmartApps like SmartLighting or only local Rules API automations like routines? (I would guess only the latter since it old va new architecture).