We're launching a new data science competition

We have become the first distribution network operator (DNO) to launch a series of data science competitions open to everyone.

The Data Science Challenge is being run in collaboration with Energy Systems Catapult (ESC) and builds on the success of a science challenge that formed part of our NIA-funded innovation project Presumed Open Data.

WPD Data and Digitalisation Manager Jonathan Berry said: “The level of interest in our first data science challenge was so incredible we knew we had to keep them going and make them a regular feature as part of our data and digitalisation programme.

“Making our data open is all about using it to add value and benefit to our customers and stakeholders and we know this is a great way to get people involved from all over the world to test the ‘art of the possible’ with our data at the centre.”

The Data Science Challenge series will identify industry-wide problems and use collaboration to deliver tangible insights and solutions to support the net zero transition.

This first challenge will investigate the loss of estimation accuracy when reducing the temporal resolution of collected demand data. The problem it is trying to resolve is that collecting and analysing high resolution data means larger storage needs and more computational processing requirements, which increase costs.

The challenge asks whether certain high-resolution features – in this case extremes of demand – can be estimated using lower resolution data, weather data, and limited historical high-resolution (one minute) data. This supports the view of using multiple data streams, such as smart meter data, and machine learning to minimise the level of physical monitoring on the network to drive the same level of insight, thereby reducing costs to customers.

Jonathan added: “This and all our challenges will be open to everyone, whether an experienced data scientist, a student, someone who is just interested in picking up data science and using the challenge to target their development – we’re keen to create an enthusiastic and inclusive energy data community.

“We have already made lots of our data open, through our Connected Data Portal, and we will continue to increase this. We’re looking forward to seeing how a collaborative community of data scientists and energy enthusiasts will use our data to answer some of the toughest questions the industry faces as we move to delivering net zero effectively.”

The challenge begins on November 11 and ends on December 1, with the winners being announced on December 8. The Digitalisation and Data team will be holding a kick-off webinar on November 11 to give further information.

Register here