Machine learning.

Sense is designed for quick installation and constant improvement. The monitor uses sensors that read the electrical current over 1 million times each second, while machine learning uses that data to determine what devices are on and off. It’s a hard problem to solve, but worthwhile, and improving all the time.

Just like speech consists of words and sounds, each device's signature is made up of distinct elements.


Sense gradually learns about your home as things turn on and off.

As Sense tracks your home’s energy, it begins to recognize most appliances and other devices that use more than 60 watts. Typically the Sense monitor detects 12 devices in the first month after it’s installed and 25-30 devices after 12 months. Each home is different, however, so the number of detected devices may be higher or lower.