AI for utilities

How to unlock the true value of data in the energy transformation

Data is powerful. But how to fulfil its business potential? Utilities, like many large organisations, typically create a ‘data lake’, but often lack the in-house skills to unlock its value. Even when companies achieve a successful initial Proof of Concept, scaling up and offering a machine learning based service to millions of customers requires very different capabilities. New tools and a new mindset are required.

Many utilities offer their customers an app that displays categorised energy insights, such as whitegood usage. These disaggregated insights are one thing, but how do you use them to create user relevance? Here's where we jumped ahead to provide the service that consumers always expected, but never got: cue Artificial Intelligence, for utilities.

By using A(rtificial) I(ntelligence), we take user data, compare and contrast it, in order to get more detailed and personalised insights. Based on those insights, we suggest actions. It's good to know when your energy usage changes, but even better to know why and what you can do about it.

What can you get out of AI?

Data is a powerful tool for utilities seeking transformation in three ways:

  • Bringing new data driven services to your customers;
  • Upselling non commodity services, and
  • Reducing internal costs.

AI for utilities can be used to create new services and better, more precise user apps. An out of the box example: Eneco also uses our real-time data insights to allow for more effective energy trading on intraday markets.

Finding a great partner in data driven transformation

Naturally, it’s essential to have a sound data platform. Yet there are other aspects to a data partner, that can help you reach success.

  1. The bigger, the better? You bet!

    The key to learning is labelled data. The more data the algorithm can learn from, the better and faster it can learn. Which is why a bigger data lake enables better data services.

  2. High and low resolution data. What’s the difference?

    Typically, high res electricity data reads electricity every 1 to 10 seconds, while low res data is read every 15 minutes to an hour. High res data requires additional hardware to read from the meter, whereas low res data can be read directly from existing meters. A data partner with access to both, has an advantage. It’s similar to having the world atlas (low res) and the A-Z street map (high res) together - providing the full solution to data challenges. 

    ! Don't be fooled by solutions that claim true disaggregation services for monthly-read meters. Monthly-read meters never provide an accurate, personalised or full picture of a customer's energy usage.

  3. Tailor made

    Many organisations provide branded or ready-made disaggregation services. But every (utility) company has specific needs, that need to be tailored to their market and user base. That’s why, with our in-house expertise, we deliver tailor-made data services and products, and bring the company's vision to market, from start to finish.

At Quby, this AI technology is real, built and deployed. Our AI tech operates at an unprecedented scale. Since the release of Toon, with our smart thermostat and energy technology in 2012, Quby has expanded to over 350,000 users, which continuously fuel our lake with both high and low resolution data.

Want to explore the possibilities of data with us? Do you have a killer idea, but need the AI capabilities to make it happen? Can you, for example, imagine the role of data in a world where customers generate their own energy? Contact us at

All about Qubys data.png
Quby Brand and Marketing