What is Sense?
The Sense Home app helps homeowners track and manage their electricity usage by providing real-time information that empowers users to see which devices are currently in use and how much electricity they are consuming. Sense returns information that helps pinpoint energy hogs, allowing users to take steps to reduce both electricity bills and carbon footprint.
Sense Labs, Inc. (“Sense Labs”) provides the Sense system for this analysis of home electrical use, often referred to as "Sense" and composed of:
- A Sense Product, which is hardware (Sense Home Energy Monitor) and software embedded in hardware (Sense-enabled electric meter), and
- A Sense Service, which is a user account that provides applications and services through mobile devices and/or our website (https://sense.com). Together the Product and Service are called “Sense”.
Once a user connects the Sense Home app to their home's electricity source, Sense begins to automatically detect and differentiate specific appliances and devices in the home by analyzing their electrical signatures. As this data is processed, small bits are also sent through the home’s Wi-Fi to the Sense cloud servers, which enable app-based reporting on the home's electrical use from a variety of perspectives.
The Sense system can also send notifications when it detects anomalies or issues in an electrical system, such as a malfunctioning appliance or unusual energy consumption. And it offers historical energy usage data, tracking a home’s consumption over time and helping users make informed decisions about energy habits.
How does automatic device detection work?
As the bits of collected information are processed in the cloud, machine learning algorithms developed by the Sense data science team look for individual devices among a home’s mix of electrical signals. This is a complex process that is rooted in some of the same technology used for speech recognition. Sense tries to separate the “voices” of each individual device in your home from the “noise” of all the other activity in your home, and then recognizes patterns in these “voices” to determine what they might be.
Once the algorithms identify a signal as an individual device, a model is automatically developed to track that device within the home. That model is sent back to the home’s Sense system so that device monitoring can be handled locally while the servers continue their analysis. Working together, the home’s system and the cloud servers identify devices and deliver real-time energy data to the “client” side of Sense software, which is the iOS, Android, and web apps that users can interact with every day.
This graph represents changes in whole-home power consumption as monitored by Sense, and some of the distinctive patterns that Sense uses to identify individual devices.
Within a split second, a home’s energy use data is ready to be viewed in the Sense apps for iOS, Android, and the web. It can take a while for your Sense monitor and our cloud services to determine patterns in your devices and start recognizing devices automatically, but in the meantime, you can learn a lot using the Sense app just by looking at the Power Meter and switching devices in your home on and off. How much energy do they use? Is a device's energy use constant, or does it change over time?
Device detection occurs over time for several reasons
- A signature is present only when the device is in use.
- Signatures for similar devices can be quite similar to each other or may be masked by an inverter.
- Some devices actually have multiple motors or heat sources, and Sense may see each one as an individual device (think of a clothes dryer or a refrigerator).
In some cases, the Sense system can be optimized by users proactively helping it merge elements within a complex device or differentiate similar devices in the home.