plug in watt meter for appliance for sense. tell sense what the meter is recording to help it learn a complex device then when it learns it you could move it to another device or put it in breaker box on individual breakers to
help identify loads on that breaker
I like this suggestion. This sounds like a reasonable request but Iâm sure that it is at too simple a level for Sense to be able to incorporate the data. It is, what I consider a usable approach I will follow independently. For years I have used Kill-o-Watt meters to track power usage by appliance. Prior to Sense, I have added more & more wireless devices to better control my power usage. Now with Sense I am interested in IDing the devices in âAlways Onâ. Currently I will incorporate new plug sensors that offer scheduled switching along with tracking power consumption, all under wireless control.
I also have a number of ânoisyâ devices I want to track. I can get some good information from these new sensors but only the 110v devices.
Hi guys,
Question for you. If Sense were to build a wi-if connected plug-in watt meter that connects back to the Sense mothership, would you support a 250$ model that samples at the same rate as the Sense probe or a 50$ model that is essentially only useful for telling the Sense mothership whether a particular device is on or off, plus the steady state power levels associated with that device ? A full sampling model will have a much larger bill of materials and associated cost, but would be able to provide exact âground truthâ which would likely greatly speed machine learning.
I raise this question as it represents a likely trade off Sense would need to make. I personally am in favor of a âground truthâ model if it truly does speed learning of my devices.
@kevin1, I too would also be particularly interested in a device that would be able to uniquely ID the attached/isolated device and make itâs signature compatible with the Sense network. The sole intent would be to speed up learning of my devices. I anticipate that it would be available only for 110v devices but that could greatly improve the %ID of my devices; greatly reducing âOtherâ & âAlways Onâ (Unknown). I can see it doing a lot to âcombineâ the various phases a device may have. The home has few hardwired devices on 110v; the furnace is the largest. Coordinating these device signatures with the Sense team, should improve their references library and allow more time for work in the 240v world; EV, AC, Stoves & ovens, etc.
As a consultant, I also have an ulterior motive. I am usually assisting with proprietary pilot operations. I would be interested in including a Sence monitor in each system but since it canât identify specific users for almost 3-5 months, the pilot term would expire before the data was in. A âsetupâ monitor would pay for itself quickly and eliminate a number of other measurement devices.
I would jump on something like this as well. Two locations specifically in my house would benefit if for no other reason to feed sense info to ignore for its detection. Itâs seems both my main TV entertainment center as well as my NAS and home network equipment generate a lot of electricity fluctuations that have to make things more difficult for sense. If an outlet wifi monitor were depleyed at those locations that sense could recieve patterns from to be removed from the house total, for detection purposes, I would think the rest of the house would be simplified greatly.
This idea is just another version of training sense. I would prefer that time is spent fine-tuning current device detection rather than building out a whole new product that is limited to such a small demographic.
Continuing the discussion from Plug in watt meter:
I was just thinking of this concept and seen this post, that way it would see the full duty cycle (ON-OFF), for the correct wattage usages for the month for that device. As it stands now they discover components of a device within the duty cycle. I would like to see that too. Plus I think It would give more data back to the main Sense data base. On the signature wattage cycle per device.
I would prefer that time is spent fine-tuning current device detection rather than building out a whole new product that is limited to such a small demographic.
Is it a small demographic?
So far to me it seems that Sense has issues learning equipment. It we could speed up that learning for the 110v devices that would be great.
plug in watt meter for appliance for sense. tell sense what the meter is recording to help it learn a complex device then when it learns it you could move it to another device or put it in breaker box on individual breakers to
help identify loads on that breaker
This is a great idea, I am surprised that Sense already doesnât have something that would do this.
I suppose at the end of the day it depends on what the end game is.
I do sometimes wonder what Senseâs business model is outside of the nifty geek factor. We pay nothing for subscriptions so once the device is sold what other revenue streams may our data help with? Would building up an entire new product line be beneficial?
I think partnering with solar and generator companies would be a marvelous thing for them to do.