I’m interested in understanding better what devices you end up bucketing under “always on”. There has been discussion that only constant draw devices are included in “always on”. Is this correct? That would mean charging adapters for mobile devices are not included. Nor would devices with a standby mode and an active mode. It would be helpful to know what you include and exclude so we can have a more accurate list of devices to match up to the devices Sense is able to “learn”.
Hi Brian, thank you for the question! We actually have a blog article on this topic: What is "Always On" Power? - Sense Blog that I would recommend taking a look at first. Please let us know if you have any other questions after reading it, we’re happy to clarify further.
Thanks, I’ve already read this blog article and while it does provide a loose definition, estimates of the overall cost, the average for “always on”, and a call to action to reduce it, it doesn’t address what kinds of devices are included, and what kind are not, in any useful detail.
Sense has two “black box” numbers - “unknown” which hasn’t been allocated into a learned device, and “always on”. “Unknown” is self-explanatory and the goal for it is clear, but “always on” is hard to feel good about. We can’t see what devices make it up. There’s no way to know or verify whether everything that is always on is being shown as “always on” because the devices that it comprises aren’t individually identified.
Is the “constant drain/sleep” load for the device shown under “always on”, but the active/awake load shown under either unknown or a different learned device? Does “always on” show all devices that are always on regardless of their sleep/active state? In other words, will any “always on” device ever be individually learned by Sense? And if there is a split load approach for sleep/active, how to we understand the energy impact of a single device?
It would be great if you could take a few example house-level models and explain what devices appear under “always on” and in what states they appear there, and show where the active loads for these same devices are being reported. We know that Sense won’t recognize every device, so at some point we have to make peace with “unknown” remaining unknown, but if “always on” devices won’t be individually learned then we need to make peace with “always on” being unknown too.
Well I have found that my computer which is technically always on is showing up in Unknown. I say technically as it draws between 250-350 watts, but when I tell it to sleep it is still drawing 39 watts. I know this from the Kill-A-Watt meter that the system is connected to. To further complicate it the system is behind a UPS device. So what is actually being measured by the Kill-A-Watt is the draw from everything connected to the UPS.
That kind of makes my point, Howard. I don’t know what devices in my home are part of “always on” and neither do you. It would be helpful to understand better (or at all) what Sense thinks they should be.
It appears that “Always on” is an average, can you also explain what it’s an average of and over what period? I’m looking at metrics I’m calculating from the data in your app, and if you are already producing a 7-day rolling average I can just report that. Thanks.
Hi guys, so the “always on” estimate is pretty simple at the moment. Not quite as simple as just an average, but a rolling estimate of over the past 24 hours or so of the baseline power. In most houses, this is mostly due to devices which are drawing power all the time (or, the minimum power draw for a device which is sometimes active and sometimes sleeping). We don’t have a good way to break this out yet, but do have some ideas of how to help break this apart that we will be working on soon. We’ll let you know once we get closer to working on this to gather any ideas or suggestions you might have about this. As a side-note, we also discussed a related topic below a forum post here: Timeline callout does not match graph.
Thank you for the description of always on, now for some minutia. If something runs for more than 24 hours, does it all of the sudden become part of the “Always On” bubble (since it’s usage would be part of the previous 24 hour average)? For instance, I have a computer that I use to play video games. I turn it on and off usually once per day. Occasionally, I leave it running (usually unintentionally), and it could run for a day or 2 consecutively.
So, first, I wish Sense could identify my computer (its one of the hogs that I want to graph usage over time).
Second, assuming Sense eventually identifies my computer, what will happen to it’s usage graph if the device runs for over 24 hours? Will it disappear into Always On? Will it stay with Computer 1? Will it report in both? Will Computer 1 be forgotten, and have to be reidentified at a later date?
Did something change over the weekend? My “Always On” has been around 1,000 watts from beginning of December until this past Saturday. As of Sunday morning, it dropped to 5 watts and has been stuck there every since.
Anyone else see a change on their “Always On” device?
p.s. My “Unknown” went up about about 1,000 as would be expected since I didn’t magically turn off all my “always on” devices.
Not seeing anything here. Always on (49w) and unknown (100w) running about normal. Sounds buggy?
I’ll see your Status change and raise you the following:
As you can see, “Always On” used to make up a significant portion of my usage, now it is like nothing…