Can dedicated circuit monitoring (DCM) be switched to another device (by physically changing the circuit the flex sensors are on within the panel? If so, how well would Sense handle the flex sensors being moved around between circuits? Ideally the accuracy of the detection would simply just be more improved for that device during the time it was under DCM. No prior device history recorded through native detection would be lost and the ‘record’ for the device would just go back to native detection once the flex sensors were moved to another circuit. i.e., an existing device can just be allocated or “merged / unmerged” into the DCM device w/o having to ever start fresh. Is this possible? I haven’t yet purchased the flex sensors and don’t have any feel for how the DCM interface and setup works.
Why I ask… From what I can tell, most of my “Other” from the winter months (with A/C off) comes from my dryer and over/stove combo; the latter seems to be the worst contributor. However, in the warm summer months I assume applying DCM to the A/C compressor would have the most value in terms of limiting the amount of “Other”. I believe I’d get the most value out of DCM if I were able to switch the device/circuit it was on in a fairly seamless manner that didn’t disrupt the history and stability of my device records – just allowed me to intuitively isolate the device contributing the most to “Other” as error at a given time.
I apologize if this has been asked previously but I didn’t see anything on the forums…
My take, based solely on a few different runs through DCM setup, is that that should work. In my mind, you would need to run DCM setup every time you did the seasonal change, and you would have to decide whether to delete the previously recorded DCM devices from the previous year (delete AC2020, when you create AC2021 ?).
Thanks @kevin1. From what you describe, it sounds like DCM may be setup like the smart plugs, where recorded usage data remains linked to the plug and not to the device the user denoted was connected to the plug during that time (an issue I would like to see Sense address). Unfortunately I think this may have been what I was hoping to avoid. I place some value on keeping a long-term record of device usage. I wouldn’t want to have to delete the history of a device previously established under native detection when converting it to DCM and/or lose the history recorded through DCM if I moved the flex sensors elsewhere. I was hoping to use DCM as a simple means to strategically improve the accuracy of recorded device history for certain periods, with the main goal being to reduce the amount of overall usage being attributed to “Other”.
I was envisioning being able to select DCM under my existing A/C device, and all data gathered from the DCM signal would be added to that same device record until I deselected DCM within it’s setting. This A/C device record would then go on behaving as it did before the DCM period, gathering what records it could from native detection. The accuracy of the device-level power meter and usage data would just be more accurate for the period it was being gathered through DCM.
Hopefully there is, or soon will be, a way to use DCM in the way that I described.
Gotcha ! You’re thinking of using DCM in a time-shared supplemental way, which indeed would be interesting. As you suggest, the usage model today is very similar to a supported smartplug. In that smartplug like usage model DCM is very useful for reducing Other, plus you would get to keep all of your histories (DCM and native detections). It’s just that they would be fragmentary. You might still get benefit from seasonally changed DCM.
Exactly… my “Other” seems to be mostly usage that’s not properly allocated towards major consumption devices. I don’t have solar or any clear main usage device like an EV, so I want to use DCM to help reduce the portion of “Other” that results from error in the recognition of other devices. During summer, A/C seems like the clear choice since it’s the biggest user (I also enjoy experimenting with small A/C efficiency improvements like cleaning/straightening condenser coils, insulating outdoor portion of suction pipe, etc. and want an accurate record of compressor usage to help measure any marginal benefit - which Sense isn’t great at since it takes some time for the waveforms learned for the A/C to update and reflect any improvement). During the winter with the A/C off, most of this error seems to come from my electric oven/stove combo (periods of high “Other” usage seem to correlate with periods of some detected oven or stove usage, or at least occur during a time we’d be using the oven/stove) . I want to be able to view DCM as an option for improved measurement accuracy of one device, and be able to switch around the device I use this improved measurement accuracy on at will.
I do recognize the issue you pointed out about this resulting in a sort of fragmented history. In other words, it isn’t necessarily proper to view usage statistics over time if they weren’t all recorded by the same method. Overall I’d say this is no worse than how we do this today despite changes always being made to the Sense algorithms. But I agree that in the long-term I would ideally make a point of switching the DCM between the A/C and the oven/stove at consistent dates in fall & spring so I could more easily consider it’s influence when looking at patterns.
I’m happy to see another user thinking in the same way as I am once I heard about DCM. We all so want to have these devices detected, we’re prepared to go to the DCM mode if we can be sure the system will “appreciate” the added detection and treat it as if it were native, but with some “tutoring” along the way.