Best Practice for Alarm Integration

I have an Elk M1Gold security/automation system successfully integrated with my RadioRa2 system. I can turn lights on and off, etc. based on rules on write in that system.

However, I was wondering what are the best practices for automation.

For example when my security system is set to be armed away, I can

• write individual rules turning lights off
• Toggle a RadioRa2 keypad button "All Off"
• Toggle a phantom button on the repeater for "All Off"
• Set a mode such as "Away" in RadioRA2 and set my lights based on that

Is there a recommended best practice?

Quote:

---

Originally Posted by johngalt

I have an Elk M1Gold security/automation system successfully integrated with my RadioRa2 system. I can turn lights on and off, etc. based on rules on write in that system.

However, I was wondering what are the best practices for automation.

For example when my security system is set to be armed away, I can

• write individual rules turning lights off
• Toggle a RadioRa2 keypad button "All Off"
• Toggle a phantom button on the repeater for "All Off"
• Set a mode such as "Away" in RadioRA2 and set my lights based on that

Is there a recommended best practice?

---

I'm planning on replicating the functionality described in Application Note 488, which describes the integration with the Honeywell Vista system. I'll have to have my HA system send the correct commands to the Lutron system since I won't have the direct integration. Since you can program the M1 system you can have it send the commands directly.

One thing to consider is making effective use of scenes. Instead of telling individual devices to change, create a scene for them and use that instead. This way you avoid the 'popcorn' effect of individual lights changing sequentially. Put them in a scene, send one command and they'll all change to their scene-defined settings together. Even better is you can control how long a light takes to make the transition, and how long it delays before doing so. As in, turning off a path of lights. The scene might have immediate 'zero fade' for certain lights, so the user sees immediate action. But some other lights along the way might have a delay of a few seconds before going off over a slower 5-second fade. Then a few others might have longer delays to allow typical 'travel time' to walk the path. All of which would be triggered by sending just one command to the repeater.

You can, of course, just control the lights directly, but you'd really miss out on some nice effects.