OVO Tech Blog

Energy Insights: What we can do

Introduction

Thomas Reid

Thomas Reid

Data Scientist at OVO Energy


data science Energy Insights analytics smart meters

Energy Insights: What we can do

Posted by Thomas Reid on .
Featured

data science Energy Insights analytics smart meters

Energy Insights: What we can do

Posted by Thomas Reid on .

First, introductions - my name is Tom and I work as a Data Scientist in the Engagement Team. What do we do, you ask? Well, as it happens, it’s pretty engaging stuff...

Helping members cut their carbon with science
Like most folks at OVO over the past 3 months, my team has been working hard on our new energy upgrade, OVO Beyond. If you’ve been out of the country, Beyond gives our members the green energy and tools they need to start cutting their carbon footprints. It’s wicked.

The tools bit is where my team comes in – we call them Energy Insights/Spotlights.

But what are Energy Insights, exactly?
To our members, they’re a monthly email and a dashboard – both are home to personalised insights and clever energy-saving actions that help them to shrink their bills and lighten their carbon footprint.

To my team, they’re the result of us crunching members’ energy usage data from their smart meter, with the power of Data Science, Machine Learning and Artificial Intelligence. But don’t let those buzzwords put you off. All of our insights have been designed from the ground up to be easy for everyone to understand.

Turning up the heat
Take heating for example. A standard energy bill doesn’t break down your energy usage by the individual things you’ve used it on. But we already know that heating and hot water typically make up the biggest chunk of it.

To us, it made sense to start by telling members how much they’re spending on heating. But we realised that knowing how much you spend isn’t really that actionable. Is dishing out £20 on heating in October good? Or is it bad?

Castles and one-bed flats
To give members more context, we started on simple comparisons like ‘your energy usage is above average’. It turns out this isn’t that fair. What if one lives in a giant castle, when most live in flats? There’s a big variety of houses, and we don’t expect them to use the same amount of gas for heating.

The answer was personalised targets. If we know a bit about members’ homes and the area they live in, we can use some of that data science and machine learning to build a model to estimate how much gas we’d expect them to use. We can then compare this to what they actually use – and see if they’re above that expectation.

Action stations, insights incoming
So, we’ve given members a more useful comparison of their gas use. But what happens if they’re using more than expected? And how can they make sure they’re only using what they need to?

It’s all about ‘actions’. With them, our members can master their energy usage. Our first actions include showing how much money could be saved by turning the thermostat down by 1 degree. Or by delaying when the heating is switched on properly for winter. Or even by switching the heating off, if it’s being left on when nobody is home. We can also use this analysis to observe trends across the country - for example the outside temperature at which members turn on their heating can be seen below for England and Wales. It is interesting to see that cities seem to turn on their heating at higher temperatures.
Screenshot-2019-10-28-at-09.01.51
We believe these insight-action combos will have the biggest impact in helping our members cut their carbon and their bills. By tailoring their insights with their energy usage from smart meters, we can estimate the potential savings that result from taking a specific action. It’s powerful stuff.

It’s just the start
There are loads more actions and insights on the way. And we’ll need help from researchers, software engineers and our members, too. The climate crisis is a vast, complex problem – only time and testing will tell us if we’re on the right track – but I’m really proud to be helping to tackle it with energy insights. One team, zero-carbon dream!

I hope this post has been insightful – and I hope you’re feeling energised by the fact we’re moving in the right direction. If you’d like to go into more detail about the potential savings from these insights check out the ‘Measuring Energy Savings’ post on the OVO Tech Blog. This work is so much more than just data science – so if you have any ideas or expertise to offer, we’re all ears over on the OVO Forum.

Thanks for your time – let’s kick that carbon!

Thomas Reid

Thomas Reid

Data Scientist at OVO Energy

View Comments...