Energy Sector: Just 23 Percent of Utilities Have a Clearly Defined Strategy for Artificial Intelligence

Munich, March 2018: Utility companies are experiencing a rapid transformation. They are witnessing their industry shifting from a regulation-driven, reliable environment to a technology-driven, uncertain and altogether more sophisticated marketplace where real-time data on energy supply and demand and energy prices are increasingly crucial. Analyzing such data fast and responding accordingly are therefore becoming more important to the survival of companies operating in this market. Artificial intelligence (AI) is going to play a key role here, given its ability to help utilities position their offering on the market optimally and to remain competitive. However, the latest Roland Berger study, Artificial intelligence: A smart move for utilities (Download study here), found that even though 83 percent of Europe's utilities are aware of this transformation, only 23 percent actually have a strategy in place for how to use AI. This is despite the fact that the consistent application of artificial intelligence can dramatically improve utilities' ability to do business successfully.

"Energy sector players have realized that the good old days are drawing to a close," says Torsten Henzelmann, Partner at Roland Berger. The industry is shifting away from the previous low-risk environment driven by regulation, where investment cycles were long, business was predictable and returns were safe. Now, utilities face the challenge of a technology-driven marketplace characterized by demand and supply volatility and the kind of fluctuating prices that inevitably come with that. "This disruption will necessitate change, as well as new business models," says Henzelmann. "Companies will need to exploit the possibilities of new technologies, such as AI. Failure to do so could cause them to fall behind in the race to remain relevant in the fiercely contested energy market."

Utilities have plenty of homework to do

The overwhelming majority (83%) of European utilities polled agree that AI will play a significant role in their business in the future, but only 23 percent of them have so far developed a strategy to take advantage of the technology. And as many as 40 percents admit to having no notion of how to use artificial intelligence for their business. "Utilities firms that deploy AI solutions could see efficiency gains in the order of 20 percent over the coming five years," says Torsten Henzelmann. "So I am really surprised to see so many companies failing to turn this important competitive factor to their advantage."

Get AI right – step by step

Artificial intelligence is finding its way into narrower, more focused business applications, where machines and operating systems are pre-programmed to use data and algorithms to not just undertake a specific job, but learn to do it better. In the energy sector, such applications have the capacity to automatically control supply and demand by looking at the amount of power being generated and the amount being consumed by customers in real time. "With this knowledge, utilities will be better able to plan and manage their decentralized power generation," says Henzelmann.

The Roland Berger experts advise utilities to take a step-by-step approach to adopting artificial intelligence. To begin with, off-the-shelf AI products can help to optimize existing models, for example via predictive maintenance – when machines learn to monitor and maintain themselves, negating costly emergency repairs – or the introduction of automated energy trading, which can improve efficiency and cut costs for utility companies. These bottom-line savings can then be used to invest in new AI-based business models.

"Utilities firms are waking up to this potential, but their approach is still very cautious and risk-averse," says Torsten Henzelmann in summary. "That said, companies should not wait too long before exploiting the unmistakable benefits of AI technology and identifying the implications for their business model. AI algorithms, in particular, constitute an area where utilities do not need to be implementing proprietary solutions or bringing new people on board – because there are so many options already available on the market."