I am on the dog food version, it works just fine for me. Try disabling
or deleting the routine, delete the speaker group. Then create a new
speaker group re-enable the routine or create a new routine wait 5
minutes and try again. I find that sometime...
My methodology involves the function of a routine that individually
controls speaker volume. With a keyword to invoke the routine. It is a
workaround and not part of the patent fight because I am not using my
voice to invoke a change to the speakers ...