Hello! I just bought Sense and got it set up a few days ago. I’m in that stage where I’m figuring out all sorts of cool things to do with the data. I feel like IFTTT is just scratching the surface so I’ve begun to look into the unofficial API and tinker a bit. I mean… what else am I supposed to do while I’m waiting for it to detect device signatures?
I first heard of Sense on Ask This Old House season 15 episode 14, and I was immediately intrigued. In that episode, they took a short tour to the Sense offices, and I noticed an interesting piece of software that Ghinwa Choueiter was using to show 2-phase data. It’s a little fuzzy from the screenshots, but it appears to read “Wattsurfer.” I cannot seem to find any information about this online.
This got me to thinking, could any of the Sense Engineers or Data Scientists be able/willing to share some information about this and any other tools they may have on-hand?
I second that question… I’m guessing that Sense’s data science environment and machine learning environment is custom built and proprietary, but built on top of a standard data science framework like R or Julia (lots of MIT grads), and a machine learning framework like Theano, Caffe or Keras. More glimpses and info on what they do internally in this excepted webinar.
For example, clustering by features at 1:20 into the video - you can see they use at least 17 features, likely more. Power and phase0 as visible axes for clustering in this screenshot.
And here are a couple of activities to consider doing while you wait:
Export your hourly data and compare against your power companies hourly measurements, assuming your utility enables that kind of access (mine does). That will give you a feel for Sense’s overall accuracy, plus hints at where it might be off from your revenue meter (hint - there can be places and times where your data might get lost communicating the Sense mothership)
Buy some TP-Link HS110s on Black Friday / Cyber Monday sales for use on outlet strips that power your router, servers (under 15A), switches, bridges, access point, AV equipment and other stuff that Sense won’t recognize for a long time to come. You’ll get quick identification and better accounting for some of your power usage right away.
Thanks for the webinar link Kevin! I have been spelunking in the forums, but hadn’t stumbled upon that webinar yet. Good information in there!
My power company offers the ability to, but I have not had them come out to install the smart meter so they can provide such hourly measurements. This is something I will be doing soon.
I have read quite a bit about the smart plugs. I’m hesitant to use the HS-110s (or the Belkins for that matter) on my server equipment as I have them on a rack-mount UPS, which, once I get the new circuits installed, will be plugged into an ATS, and that ATS plugged into both circuits. I’ll definitely need to try a few out to see if it works properly with my setup, but it is definitely something I’m interested in trying. Would you put one between the device and the UPS? One between the UPS and the wall? Both?
For your server, I would try 2 HS-110s, but only monitor the one at the wall with Sense, to avoid double counting. That’s the location that will align with Sense accounting.
But once your put the ATS in your circuit box, things become trickier, but only if you expect Sense to measure usage correctly during power outages. Having energy sources (batteries) that time-shift incoming metered energy (charging) to energy delivered to later to a device (battery backup to your server) is going to force differences between your meter and your device usage.
It is our internal tool to visualize signals and data. As you can imagine, such a tool has to be smooth, efficient, and capable of handling large amounts of data.
