Artificial intelligence is starting to reshape wind despite warnings from leaders in the technology community about an AI-driven stock market bubble. We look at how companies can use AI to make operations more efficient and discuss why competition among manufacturers potentially limits AI’s usefulness.
Stock markets may be on course to repeat the dotcom boom and bust cycle of the late 1990s. That was the warning from Google chief executive Sundar Pichai last week.
Pichai said soaring share prices for companies dependent on artificial intelligence (AI) technologies contained “elements of irrationality”. This year, the combined value of eight leading AI stocks including Nvidia, Microsoft and Google parent company Alphabet have quadrupled, but there are fears that this is now coming to an end. Pichai said that every company would face financial knock-on effects if the AI bubble bursts.
This arguably irrational exuberance for AI stocks comes after growing evidence of how it is reshaping sectors. The so-called ‘big four’ consultants – Deloitte, EY, KPMG and PwC – and other management consultants including McKinsey are among those that need to adapt as the automation of research tasks is leading to changes in the workforce. This work is being increasingly carried out by AI systems rather than junior staff and is also having a direct impact on the value of the work these consultants carry out for clients.
This year, McKinsey has reported that 78% of businesses have used AI for one or more business function, which is up from 55% in 2023. This is in addition to the 90% of firms that plan to step up their investment in AI over the next three years.
But businesses in the wind industry need to get to grips with different types of AI, from large language models such as ChatGPT to more specialised AI-based wind analysis systems. A report published last week by wind technology firm Shoreline Wind, called ‘The Wind CFOs’ Guide to Artificial Intelligence’, shared some ways that businesses in the wind industry are looking to use AI to deliver projects and wider business functions more efficiently. It also sought to address misconceptions and discuss the limits of AI.
Optimising the project life cycle
The report argued that AI could assist companies to develop wind farms more efficiently by enabling them to optimise the positions of turbines for maximum energy production, for example. It also looked at how AI could help develop construction plans taking into account actual or predicted pressures in the supply chain; identify projects facing wind farms already in operation; and support management teams with wider market analysis.
Interviewees in the report also highlighted how AI could help automate data-heavy tasks such as energy yield assessments, and help with asset management, customer support and regulatory compliance. In an era where the wind industry is likely to face shortages of skilled professionals, advances in these areas could prove to be invaluable.
Matthew Geraghty, founder and chief executive of ReWind Energy, argued that there is “a lot of scope for modernisation” of working practices in wind because some are “not very efficient by today’s standards”. This could include removing menial work of energy yield assessments, weather data assessments and wind measurement campaigns.
Other interviewees saw potential in AI in key tasks such as diverse as planning offshore construction schedules and pulling together summaries of different regulatory regimes.
But the report also highlighted some limitations of AI. For instance, AI systems can only produce reliable predictions if they draw on robust data, and will likely struggle if turbine and vessel manufacturers are secretive about the specifications of new machines.
Anders Frekderiksen, general manager at Head Energy Denmark, said this competitive pressure made it difficult to secure sensitive data from manufacturers in the market at present and the growth of AI would not change that: “We work for a lot of companies so we cannot share data between them. We cannot disclose this and AI would have the same problem,” he argued.
Shoreline Wind reported that AI systems could help reduce the operational expenditure of wind farms by 10% with greater efficiency. But Ole-Erik Endrerud, founder and chief product officer at Shoreline Wind, said there are still misunderstandings in the industry about what AI and machine learning are, as the terms can be used interchangeably.
AI is a broad term for a wide range of systems that mimic human intelligence, including machine learning systems and large language models like ChatGPT. Machine learning is a branch of AI that uses a neural network to detect patterns in large datasets based on known issues and patterns. This opens the way for AI software specialists to create industry-specific solutions, although they are only as reliable as the data that is put in.
Endrerud added that Shoreline Wind’s systems are based on AI-powered solutions that can “provide the CFO with information so they can make decisions that lower the risk of a company”. AI systems can produce analysis “a lot faster than an analyst did” and with greater accuracy, which means that “CFOs can sleep better” at night.
AI is on the rise but companies in the wind industry are barely scratching the surface in how they use it. The impact on analytical and logistical tasks could be transformational but companies will also need to understand the likely limitations of AI.
Six years ago at our annual conference in London, the then-CEO of Siemens Gamesa, Markus Tacke, spoke about the potentially transformational impact of AI on efficiency in the wind industry. We are only now starting to see this happen, and the potential of AI is set to remain even if the sky-high share prices of AI-dependent companies do not.
Strategic communications, insight and events for the global energy transition.