“Always On” Stats doesn't include the “always on” smart plug usage

When I look at the “Devices > Always On” “comparison”, it shows only the usage that wasn’t tracked by my smart plugs (it shows 94 watts, rather than the actual “always on” of 140-150).

The itemized breakout below that “Stats/comparison” shows an itemized breakout of roughly 48 watts (meaning that my [unidentified] “always on” of 94w + [itemized] “always on” of 48w + <a few watts of “Itemized always on” devices being slightly above their norms>: equals: my typical “always on” of roughly 140-150… which is the number shown counting my live “always on”).

However, the number shown in the Stats and the comparison (roughly 94w) does not match my true “always on” at all.

Instead: that number is the unidentified “always on” number… which would be cool if I were competing to get the lowest “unitemized” usage… but I’m not particularly interested in that… rather: I’m interested to drive down my actual “always on” usage)

When this feature was initially introduced, I thought it was maybe just a temporary thing, but this has persisted quite reliably for several months now.

I understand the concept of “always on” as a kind of “baseline” usage that Sense will “never” have insight to (and I fully recognize the importance of the difference between the “always on” usage that Sense does not have granular tracking for vs. the “always on” that Sense does have granular tracking for [presumably via smart plugs or such])…

However, if I am working to minimize my net “always on” power usage, then just putting a bunch of smart plugs around the house (on “always on” devices) will misleadingly indicate that my [Sense unitemized] “always on” has decreased considerably: when, really my “minimum power usage” [what I would call my actual “always on”] is entirely unchanged (technically would be slightly increased because of the few watts that each smart plug consumes constantly)…

Is this something unique to me, or am I misunderstanding something?
Or is this a bug/feature (and is Sense aware of this?)?


I think that this only appears wrong if you view the Compare metric as a competition for the lowest usage. With smartplugs in the mix, Always On and identification/detection are orthogonal measures. You have smartplug devices that have separate identified Variable and Always On components (Blue). You can have Sense detections that only identify a Variable component of the device, completely missing the Always On of that device, plus you have the Always On (and Variable) of all the other unidentified devices in the house.

If you think about your example, and the usage / billing accounting, the 48W of Always On detected by the smartplugs could either be associated with the the smartplug devices or Always On, but not both (that would be double counting that usage). There’s nothing wrong with peeling off the 48W of Always On detected by the smartplugs from the whole house Always On, to get a more accurate view on how much Always On is still unaccounted for.