The offshore wind farm support sector is shaping up to be a fertile breeding ground for new AI and drone solutions, attracting tech developers and skunkworks projects united by the goal of making turbine supply and maintenance tasks safer and more efficient.
It’s a goal that could be achieved via multiple applications – from delivering critical spare parts and tools directly to turbine technicians, to tightening bolts on these structures and keeping them clean. During a webinar hosted by ORE Catapult in February 2020, Dr Iain Wallace, chief scientific officer at subsea ROV services firm Rovco, opined: “We look at some [offshore wind farm support-related] tasks where people are involved, and think – ‘people really shouldn’t be there’ – but the technology to help them isn’t always available.”
His view was echoed by Sara Bernardini, professor of AI at Royal Holloway University of London, UK, though, she cautioned, there is some resistance to widespread exploitation of AI in the offshore wind farm sector. “A lot of skilled people take pride in doing dangerous jobs,” she said. Persuading such people to embrace AI could perhaps be achieved by offering operators more realistic visual content, to create greater engagement. “I think the way forward will be joint deliberation between human and robot, so they are both in charge of the mission,” Bernardini offered.
A few schemes are underway. For example, SOV operator Esvagt has joined forces with Siemens Gamesa and energy company Ørsted to devise a way to transport small electrical components and tools to offshore wind farms by aerial drone. The “Operative package deliveries by drones” project will investigate the viability of using drones to fly packages weighing up to 3-4kg directly to wind turbine nacelles, to eliminate the hassle of using boats to transport spare parts to the wind farm. This could also cut back on work stoppages and reduce opportunities for slips or falls.
Flemming Hjorth, head of new services business development at Esvagt, has commented: “Our ambition is to get a drone to transport spare parts between two variable points, following a route that will be adjusted along the way. It is complex, even when using drone pilots, and becomes even more demanding once you add the changing weight of the cargo, the wind’s impact, the use of magnetic compasses in offshore wind farms with lots of steel, and so on.” For instance, guiding the drone to a moving nacelle and past rotating blades will certainly pose a challenge, as will launching the drone from an SOV that’s constantly moving around various sections of the wind farm.
Another notable AI venture has been the CleanWinTur project, an effort to develop an ultrasonic method of safeguarding the integrity of wind turbine substructures and splash zones. Typically, human dive squads and ROVs travel out to the turbines to conduct inspections of the superstructures and apply anti-foulings/paint treatments, and so on. However, as more powerful turbines crop up further and further from shore, this task becomes time-consuming, tiring and dangerous.
Instead, CleanWinTur proposes an ultrasonic system, permanently installed at the turbine, which would fulfil two key roles. Firstly, it would use ultrasound to create acoustic cavitation on the outer surface of the turbine substructure, to prevent the growth of micro-organisms. Secondly, the system could be used as an ‘always-on’ condition monitoring tool – detecting any changes in the thickness of the turbine substructure wall, and/or spotting the earliest signs of corrosion before it has the chance to do damage.