How AI is helping make offshore wind power more sustainable
Image credit: Michal Bednarek/Dreamstime
Automation could massively reduce the carbon emissions associated with the time-consuming process of monitoring and maintaining renewable energy assets.
The global offshore wind market grew by nearly 30 per cent each year between 2010 and 2018, according to the International Energy Agency (IEA). Countries such as Denmark, China and the UK are leading the way in offshore wind power, with Prime Minister Boris Johnson even stating that he wants Britain to be the “Saudi Arabia of wind power”. The industry promises to keep growing with investment in renewable energy hitting record highs in the first half of 2021, according to BloombergNEF.
However, while wind farms harness a sustainable source of energy, the energy transition itself needs to be appropriately managed to reduce both carbon dioxide emissions and the impact on the environment from the businesses that make up this burgeoning industry.
The cables and foundations that support turbines and carry power from wind farms back to the mainland need constant monitoring and maintenance. Conducting this work on offshore wind turbines usually requires energy companies to send out large vessels that use vast quantities of fuel, with very high operating costs, and are often crewed by up to 60 people from engineers and submersible pilots to cooks and cleaners. Over its lifetime, one of these vessels can be responsible for up to 275,000 tonnes of carbon emissions.
However, new and innovative techniques are helping to minimise the negative impact on the environment from this otherwise planet-preserving process. The use of more advanced technologies, such as simultaneous localisation and mapping (SLAM), machine learning, and autonomous ROVs (remotely operated underwater vehicles), presents an opportunity for offshore asset-monitoring and maintenance processes to evolve and revolutionise this industry sector – making it easier, cheaper, and cleaner to move to a more sustainable energy generation mix.
Today, many offshore energy companies are using manually operated ROVs to collect video data for asset inspection. Each typically needs at least two pilots to operate it, and then the data collected is reviewed manually by an additional team who also live onboard the vessel.
While an ROV would normally be manually controlled by an operator onboard the ship, autonomy technology enables the submersible to take the SLAM information, work out where it is and analyse ‘on the go’ in real-time, presenting choices to a land-based human supervisor. The machine then executes its mission while navigating obstacles, or course correcting to mitigate the effects of currents. The supervisor can then make informed decisions with a single touch.
By enabling greater levels of autonomy, fewer pilots are needed, and they can be located on shore, in a purely supervisory role, thereby reducing the need for bigger vessels offshore.
Another challenge companies face in this space is that the asset data collected from ROVs is too large to be transmitted via satellites. Tens or even hundreds of hours of 4K video files are typically analysed manually onboard the vessels to determine issues and potential threats, such as damage or marine growth. This is painstakingly slow and time consuming. Time during which the vessel has to remain at sea.
By compressing video feeds into a 3D point cloud with discrete images and running them through a low-bandwidth cloud platform, surveyors can accelerate the process significantly. Employing machine-learning analysis on these frames further speeds the process, with the technology able to recognise key features and abnormalities and categorise them – enabling human operators to simply check the work and confirm the findings, automating a long and tedious process. This technology makes it possible to reduce not only the time needed to carry out this task but the need to take these crew members on the vessels at all, enabling the work to be done on-shore – significantly lowering the cost and preventing needless carbon emissions.
By incorporating this technology into existing processes, companies are making strides towards sustainable and clean energy transition. By the year 2026, my own company Vaarst expects its technology to have eliminated the need for 75 of the large vessels currently used for offshore maintenance work. This will effectively reduce carbon dioxide pollution in the industry by 825,000 tonnes per year.
It's clear that this technology has huge potential in subsea applications. The marine environment offers a complex but valuable test bed to develop and train autonomous robotics, however, the technology is not limited to this use-case. There is significant opportunity for the benefits to be leveraged across numerous industries.
Autonomy technology, along with SLAM systems and machine-learning analysis platforms, can be applied to almost any robot, not just undersea ROVs. It can be fitted to drones, to be used to inspect skyscrapers. It can be used in nuclear wastelands, completing work in areas too hazardous to humans. We even see the potential for it to be used in orbit, conducting inspections on satellites. With the possibility of almost limitless application, expect to see more and more businesses reaping the benefits this technology has to offer.
Brian Allen is CEO of Vaarst.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.