What's new in v22: Philips Hue


You should learn a little bit more about machine learning training. The “training” you describe is not particularly useful for machine learning. Data scientists want to see whether the actual device at the outlet or socket is on or off (ground truth) at every sample point (with Sense, every microsecond). Better yet, add in a second piece of data, the rough power that the device is consuming. The on/off data can be offset in time by a second or so, since the machine learning network will have some “short-term memory” built into it.

The good news is that there are lots of readily network accessible sources of power usage ground truth data to feed machine learning. Hue is one good example. Other sources include smart thermostats (Ecobee, Nest), smart plugs (provide power information as well), smart chargers for/in EVs, smart appliances, etc. And, not every device needs to report in for Sense to improve with this additional supervision / feedback. Other devices are likely to be better defined if Sense more accurately pegs devices with ground truth input. So ground truth feedback should be good for everybody. The bad news it that you need a machine to provide training input, not yourself.


Good question. And no, they do not charge at the same time typically, because usually I use the same charger for both of them and it only has one plug!


This is a point well taken. I don’t happen to have that much lighting going on, but my case is not everyone’s case!


I can highly recommend the Coursera course on machine learning. It’s taught by a Stanford Prof. If you remember enough from algebra, you’ll do OK. If you remember linear algebra, you’ll do great. If you remember calculus, you make the rest of the class look bad; pretend you don’t remember.

But you’ll learn a ton that will put this in context.


Yes, the Coursera series on Machine Learning by Andrew Ng is really cool… Gives hands on insight into some of magic of machine learning plus how it can go off the rails (overfitting, underfitting, drift, wrong neural network structure)


Ryan is the prioritized list of next integrations a secret? Which? and when can we expect more? Every IoT device is theoretically “seeable” using this approach, correct? With my single vote I’d recommend high-draw devices like pool pumps (Pentair app), HVAC (via NEST or other theomstats), EVs (like my I3 that know exactly when they are on/off via BMWConnected app).

Firmware Update: Version 1.12.1966

Got a big patch update (iOS v22.1, Android v22.2).

This release greatly improves the syncing between Sense and Hue. Now, as you change your Hue setup (add new lights, make new rooms, move lights between rooms, etc) Sense will better detect those changes. In addition, you will now be able disconnect Hue from your Sense monitor (under Settings > My Home > Connected Devices).