Just curious as to why we couldn’t plug in all our major appliances down to make, model number to assist Sense in scanning for said devices or matching signals up with unknown devices.?
+1 and some thoughts.
There are two modes of operation here a bit:
Mode 1: Only apply models that look for these specific devices. This may significantly speed up device identification.l and assist in the development of new models but limits detection of non-detected devices.
Mode 2: Focus on these identified models and then switch to other models after a majority of these devices are detected and we are failing to detect other devices.
Prime example: to “truly” list all devices is not as easy as people think. Were you planning to list:
- The light in your fridge?
- The light in your oven?
- The water heater in your washing machine? (Assuming you have a newer one)
- The Garage door?
There are lots of little things some people may not address.
Funny, but done right, as models are put around devices and devices are detected in a house, the ability to detect additional devices should get faster and more accurate within said house. By identifying “noise” we strengthen the signal of remaining devices.
Thanks for the suggestion, @ahrenh
As @vDoubleShot said, it is a little more complicated that it seems. That being said, this is something we are looking into as we do think it would be helpful for our data science team.
I have a question. It’s going to sound bad but it’s really just an honest question. Is Sense making 1) an energy monitor to help people know and understand their electrical usage…or is Sense 2) developing a machine learning algorithm/model to be applied on many applications…energy modeling being one. The reason I am asking…I signed up for the first. I want an energy monitor. Many of the ideas offered on this board will get that done quickly and efficiently. It will make many people happy. I don’t want my $300 to go for a research project. If you cannot (or will not) provide 1 until 2 then you should tell us and figure out a way to compensate those “early adopters” who should have known better…like me. I have been patient since December
@mstraka606, Sense is building a smart home energy monitor with the goal of giving people greater awareness of what goes on in their home. The machine learning driven device detection is how we are going about achieving that goal.