top of page

AI Case Study

Verv identifies home electricity usage by appliance and monitors for unusual usage patterns using machine learning

Verv is a consumer product which, when attached to an electricity meter, provides users with a breakdown of their appliance's electricity consumption as well as notification when one has been left on or appears to be faulty. This is achieved by using machine learning to recognize the devices and by taking data samples regularly.

Industry

Utilities

Electricity

Project Overview

From The Startup Network, Verv is "a home energy assistant that sits next to your electricity or smart meter and uses artificial intelligence to turn all of your appliances smart. Verv starts by identifying key appliances in your home so that it can tell you how much each one is costing you, in addition to their kWh consumption. As Verv gets to know your home, it will start to identify more of your appliances and continually provide you with a real-time feed so you can see what electricity you are using at any given moment." Verv claims its product can "sample energy data at hundreds of thousands of times per second meaning the insights we can give consumers into their homes extend far beyond cost. Not only can we detect what appliances are on by reading electricity signatures, we can tell if your fridge is faulty, if the washing machine drain is starting to block, in fact we can even tell you if you have overloaded the machine."

Reported Results

Results undisclosed

Technology

According to Verv's whitepaper it "uses machine learning algorithms to derive a real-time profile for key household appliances – providing homeowners with a view of their electrical appliances’ current status (i.e. on/off) and condition (e.g. operating at 95% efficiency)".

Function

Operations

Network Operations

Background

Verv aims to provide greater transparency for customers regarding their electricity usage by breaking it down into key appliances.

Benefits

Data

Appliance electricity signatures in energy data (kWh)

bottom of page