Data quality issues and the AI gap

Even when systems connect, quality of data is also an issue that can slow momentum. “Technology is racing ahead but it is bottlenecked by the quality of data,” Steven Steer explains. There is a lot of variety in the way data is stored and understood. “Data tends to be in local languages from proprietary providers who record it in a certain way. And each network holds that data in its own fashion. The languages, the lexicon, the meaning of the words that get used, and the metadata that wraps around them, are very localised.

“It might work well in a walled garden, but we have not coordinated that data at a sectoral level. And until you do that you are not going to get sectoral level insights on that data, or train AI models effectively.”

Artificial intelligence will inevitably play a role in the smart grid but practical examples of it enabling the grid to self-heal (as envisaged by the ‘Father of the Smart Grid’, Professor Massoud Amin, back in 2001) are thin on the ground. “You’ve got the brochureware answer,” says Elliman, “and you’ve got the reality.”

He says there are examples, particularly in the US, of engineers implementing some small-scale solutions of switchable on-demand routing of power grids based on algorithms. “They will tell you they are becoming the brain of the grid.” He adds there are some cases where equipment availability has been boosted because of predictive maintenance and utilities have reduced unplanned outages, which they can attribute to AI.

But the rapid pace of AI innovation contrasts sharply with digitalisation of the network. “I think the concern is the technology is racing ahead on the one hand, and physical networks are struggling to catch up.”

The prize when physical networks fully exploit the potential of digital technology is a big one: an energy system based on renewable power and not beholden to international gas prices. For struggling consumers, this is a true measure of resiliency, says Steer. “Resiliency is not just a system that works, but a service that doesn’t bankrupt me. We need a resilient energy service, not a resilient energy system; these two things are not the same.”

A backlash is taking place against net zero, he notes, but decarbonisation makes sense from the point of view of needing to upgrade our ageing energy infrastructure. “Whether you are targeting net zero or whether you just want electricity in the future, many of the decisions you would take are the same.

“If you are frustrated with net zero, relying on gas is the implied alternative. But what you are then doing is trading off investment in renewables at the cost of exposure to international gas markets. In other words, the choice isn’t between net zero or money in your pocket – that’s a falsehood. The choice is between spending to futureproof the grid or keeping the money in your pocket today, but losing your energy service tomorrow.”

The smart grid represents one of the most complex engineering and digital transformation challenges of our time, Steer notes. Bridging the gap between vision and reality will require bold, coordinated action, and it must start now.

According to Zühlke, the foundation of grid resilience isn’t just technology and data – it’s people. It’s the engineers, analysts, and policymakers who choose to act, experiment, and collaborate across boundaries. The future grid won’t be built by those who already have all the answers, but by those who keep asking better questions and have the courage to learn from every challenge.

The transformation to a smarter, stronger, and more sustainable grid is already underway. But delivering on its promise will depend on one thing above all: connecting the human and digital intelligence that powers it.