I have had sense for over a year. When I first installed it detected my submersible pump right away. I replaced the pump last summer and it was detected a few days later. Two pumps both detected within a few days.
In June we had a new well drilled with a new pump. For whatever reason it has still not detected the pump. It had been about a month.
Is there anything I can do to help it recognize it?
I have a deep well constant pressure pump and, not only has Sense been unable to detect it, but Sense Engineering tells me that device “noise” is interfering with detecting most everything else in my home. After 3 years, only 17 or my 176 devices detected, and those not consistently /accurately.
I’m relatively unfamiliar with well pumps (apartment dweller here), but do you know what type of well pump you have? Is it powered by a variable or single-speed motor?
Even if it’s not a variable speed/constant pressure pump, many modern pump controllers ramp up slowly. My understanding is that makes it difficult for Sense.
Well, I don’t have any useful suggestions here other than to either wait a while longer, or delete the old device and see what happens. If it were me, I’d care more that my primary well pump was detected properly than that the backup/secondary pump is tracked. That said, Sense should be able to differentiate between a 1/2 and 3/4 HP pump load pretty easily. In fact, I have two well pumps, one 3/4 submersible and one 1/2 HP jet, and Sense detects both reliably. Any other detected motor loads that could be similar to the 3/4HP pump?
I have no problem deleting the other pump, but maybe I’m confused on how that would help find a different one. They have completely different signatures.
I’ve monitored for a month to see if it was incorrectly assigning labeling the pump a different device and there has been nothing.
Like I said I’ve had multiple pumps, 2 other submersible, a pressure pump, hot tub pumps etc. And they are all detected in relative short order.
I really want to find it, because the sense actually alerted me to a leaky pipe in my lawn when my old well sprung a leak last year. The pump ran longer than normal and how I found an issue.
I don’t have knowledge of how the Sense local data processing works in the background, but my observations as a user lead me to believe that the Sense monitor looks for “interesting” power transitions and first compares them to its internal database of locally discovered device. It then uses those models to determine whether any known devices have turned on or off. If the waveform doesn’t match any local device, it then sends the data to the cloud for processing, which could eventually result in a new device model being generated and pushed down to the monitor.
What I’ve seen is that if a not-discovered device is similar to an existing device model that’s presumably loaded into the monitor, perhaps from a defunct appliance that’s similar but not exactly the same as its replacement, Sense won’t discover the new device until the old one is deleted. I suspect that the transition event is close enough to the old device model that some internal process prevents the data from being uploaded to the cloud, but it also doesn’t trigger the existing model to activate. In the past when I’ve seen this, I’ve just deleted the offending device model, and then Sense eventually will rediscover the new device(s) and push down new model(s).
Hi @pswired. I checked with Data Science on this. Based on how the device was detected on our end (we have several different methodologies working in the back-end), an existing model should stop new duplicating models from showing up even if the existing model is off slightly. So, your suspicion is correct.