Powering a data-driven energy sector

• The energy sector is undergoing big changes with the adoption of AI and digitalisation, shifting from traditional practices to real-time data and analysis models.

• Integrating IT and Operational Technologies (OT) is crucial for enhancing the energy sector, enabling proactive and predictive maintenance.

• Generative AI offers promising opportunities for the energy sector, but its implementation requires careful consideration of risks and constant review for accuracy.

For the energy sector, change is in the air. With the advent of AI and a rapid push for digitalisation, the traditional reliance on experience and established practices is evolving towards a real-time data and analysis model.

Although this swing from knowledge to hard data is not exclusive to the energy sector, innovation within the industry will be felt by electricity customers far and wide. Because being data-driven means developing our use of data to inform vital business decisions which will benefit our operations and the end user.

What’s more, moving from manual to digital data collection is key to making real-time decisions. Not only is the data gathered more sophisticated and granular than ever before, but also the time it takes to gather said information is significantly shorter.

So, what can energy providers really do with all the right data in hand at the right time?

Less revolution. More evolution.

Backing the energy sector with better IT tools and procedures can do more than transform the industry; it can change how we use Operational Technologies (OT) today. Because data can power a move from reactive to proactive and, ultimately, predictive maintenance, it is key to futureproofing our systems.

Matt Webb, CIO of UK Power Networks, believes that IT and OT should converge. In fact, if IT and OT function separately, they are less effective, or as Matt puts it: "If you have a divergence of the two, then there's a problem". This interdependence means that IT is essential for supporting and enhancing OT. In other words, for the hardware – the OT – to work optimally, the IT – the data – must work harder to optimise, analyse and utilise data more intelligently.

By identifying high-value, low-risk use cases – such as testing predictive maintenance, enhancing grid management, or optimising renewable energy integration – AI capabilities can be securely tested before being rolled out at scale.

At their core, generative AI outputs need constant, careful review for accuracy since they depend heavily on the quality of the underlying data. For example, some large language models are known to make plausible but erroneous logical leaps, making human oversight non-negotiable. This is especially true when considering that machines should not have the final say on critical human infrastructure environments – even if they are based on reliable data.

Putting data into action

Planning to use data-led insights and familiarising yourself with hypotheses is a start, but how can the energy sector affect, today, with the data and tools at their disposal? 

Service interruption responses are a real-life use case for UK Power Networks, for example. Currently, the organisation uses data to improve communication and response times during emergencies, particularly in storm scenarios, identifying affected customers as well as alerting and updating them with SMS messages on restoration efforts, even before an issue has been reported.