Is there a delay in Sense App

Sense found a new device. Said it was the microwave. Wanted to test it so I heated some water in the microwave. Almost immediately it said it was on. When I turned it off it said it was on for at least another 5 minutes. When it said it was off again I retested. As soon as I turned it on it acknowledged that it was on. When I turned it off it said it was still on. I tripped the breaker and it still said it was on for several minutes. Does anyone know why this would be happening? Thanks

Sounds like Sense is still learning the off-signature and missing, at least sometimes. If Sense misses an off-signature, it eventually times out and turns the bubble off, with the time depending on the type of the device (Microwave might be 5 minutes). Also, please be aware that flipping a breaker off looks different than a microwave turning off normally to Sense, so you shouldn’t expect flipping the breaker off (especially if there are other device on the breaker) to make the bubble go away.

As with many other operations of the Sense Machine learning, it is best to give the system time to continue learning the signature of your device(s) to improve detection accuracy. I’m with @kevin1 that the system needs a little more time to detect the power down signature of the microwave.

My microwave appears to be fully detected and if I have notifications on I get alerted within a couple of seconds of the microwave shutting off.

Thank you all for your responses. It is nice to have an active community to bounce ideas and questions around. Sounds like I just need to give it time. Also good to know that flipping a breaker may not be the best way to find/test what Sense sees. Any other ideas or best practices on how to validate Sense findings. Right now it is like an easter egg hunt. Very happy with the product though.

1 Like

I have used the unplug/breaker trip method to figure out which one of my refrigerators and refrigerator defrosters was being detected so I could label them correctly, but I try not to make that a habit and confuse the machine learning.