I had a motor (that I hoped was my well pump) and another device that disappeared from my currently learning list without being added as new devices. 
I don’t suppose there is any action I can take other than waiting for them to show back up, but I’m curious about the process - why it stop tracking something it had nearly identified. (the microwave that was unplugged and moved out a month ago, dropped from currently learning to up next but is still there).
Had the same thing happen to me last night. A mystery heat device (950W) has disappeared.
The “Currently Learning” section is a (stalled) work in progress. Don’t read too much into what it says.
Infrequently, and especially early on, some devices that have been detected and named will get lost. Can’t say I’ve had it happen in almost a year, though.
well. they just popped in to my devices last night. I guess it just had to think about it for a bit. One was my space heater and the motor I haven’t ID’d yet. It gives an average of 473 watts, with an avg run time of 9 minutes, so it’s definitely not my well pump.
I have a followup question on this … actually more of a ‘duh’ moment than a question. I’m guessing that if I have one space heater and move it around, using different outlets, it will see it as a individual device at each outlet? It definitely popped up the new heat bubble with I used the space heater Wednesday, but not last night or this morning. … I’ll test it when I get home, but I’m wondering if this is the issue.
@njhaley is correct in that the ‘Currently Learning’ section is in need of something improvements and unfortunately, doesn’t always accurately display what is being learned.
In terms of the space heater question, it depends what phase it’s on. In other words, it depends what outlet in your house you move it to. Some outlets will see it as the same device.
Ben,
If “Currently Learning” is the list of mystery / motor / heat / named devices that I haven’t acknowledged yet, I would describe it as possessing excessive enthusiasm and a good imagination, a few weeks in.
I wonder if step one in sorting it out would be to put some checks in place that would reject new devices that fail some obvious tests. My experience in offering users software functionality is that, if found to be flaky on introduction to real users, remove it rather than leave it offering misleading information while being worked. Often that is not easy, but in this case I think use cases for certain devices could be written that eliminates some of the “noise”.
example I’ve just had my water heater identified. Consumption is 10W, which isn’t a water heater of course and we have a gas water heater. I do have a pseudo water heater as part of my humdifier (discussing lack of detection with Support off-line) but it has a very distinct sawtooth resistive trend which would not fit the pattern of a typical water heater.
I’ve just had a 10W Motor 2 identified this evening, which has been running for approx 25 mins. Feels like it’s in the noise. Even if there was a credible 10W motor detection, could loads below ??? be suppressed for some significant time or until a user had flagged that a primary list of loads had been detected.
Sense is a fabulous tool, isn’t it?
ok - I figured out the space heater mystery. It’s a little one that I carry around and use in different rooms. So - I had tried plugging it in different outlets but still didn’t know why it fired the bubble sometimes and not others. What I finally figured out was that if it was turned on and I unplugged it and plugged it back on, it popped the bubble. If it was plugged in and I used the switch to turn it on and off, no bubble. So Sense is seeing different startup patterns?? i think in this case, I’ll just have to wait until it recognizes all the various instances and merge them.
This sounds a little like my water heater, which is on a timer. Sometimes it gets it, sometimes only half (one 120v leg), sometimes not at all.
My hunch is that the timer affects how it starts up each time.