The UK Government has committed to achieving net zero emissions by 2050, which will require significant changes for the energy industry, starting with wind power. Despite wind being a renewable source of energy, the process of monitoring and maintaining offshore assets such as turbines, subsea cables and pipelines, is not quite as environmentally friendly.

Conducting this type of work on offshore wind turbines usually requires energy companies to dispatch large vessels – carrying crews of remotely operated vehicles (ROVs) pilots and technical specialists, as well as support staff like cooks – that use vast quantities of fuel, with very high operating costs. Over its lifetime, one of these vessels can emit up to 275,000 tonnes of carbon emissions – a staggering amount that flies in the face of transitioning to a cleaner energy source.

With installed offshore wind capacity set to rise to 27.5 GW in 2026, up from 10.5 GW in 2020, it is imperative that energy companies take steps to address the emissions produced by these offshore operations.

The simple issue is that too many of these vessels are spending too much time at sea. Fortunately, as with many challenges in the industrial sector, innovative technologies have been developed to provide a solution.

The use of marine robots in the energy industry has dramatically increased since the introduction of unmanned vehicles in the 1970s, but operations are still dominated by ROVs. Assessing subsea assets is still typically conducted by ROVs, controlled by two pilots, spending hours at a time collecting video data to be reviewed manually by another team. The back and forth between teams has often contributed to idle time spent at sea, racking up costs and producing unnecessary emissions.

To reduce the risks to work on offshore assets, more and more companies are turning to technologies such as simultaneous localisation and mapping (SLAM), machine learning (ML), and autonomy.

Piloting an ROV requires skill and patience, but by adding autonomous capabilities, the process becomes significantly more efficient. For example, piloting an ROV along the route of an underwater cable connecting an offshore wind farm to the mainland, whilst capturing high-definition video, adjusting for the current, and avoiding obstacles, is a long and difficult task.

An autonomous ROV can complete the job with minimal human intervention. By enabling greater levels of autonomy, fewer pilots are needed, and they can be located onshore in a purely supervisory role, thereby reducing the need for bigger vessels.

With every hour the vessels they charter spend out at sea, energy companies are incurring significant costs. On top of this, some vessels are idle whilst video data collected by the ROVs is assessed.

The video data can be hundreds of hours long, covering dozens of miles of underwater cable. Assessing this data has often been completed by placing tens of people offshore on each vessel, manually trawling through every frame. To solve this issue, some companies are turning to ML analysis platforms to accelerate the process.

By cutting the video feeds up into discrete frames and running them through an ML analysis platform which can recognise and categorise key features and abnormalities, analysts can streamline the process significantly.

This works by identifying frames showing a feature such as pipeline nodes and instructing the platform to find frames showing similar features. It functions as a search engine for the human operators, allowing them to find exactly what it is they’re looking for, such as damage and potential threats, without the need to watch hours of video footage. By further reducing the time spent analysing the footage, the platform is cutting time spent at sea, reducing the cost to the energy company and the impact on the environment.

Those companies already working with the combination of autonomous robotics and ML analysis platforms are seeing significant cost benefits as well as realising the environmental upsides of reducing the size of the vessels they use and the time they spend offshore. Mass adoption of these technologies would enable the removal of many of the most problematic vessels in operation, decreasing the total emissions produced, and in doing so, improve the scalability of offshore wind power.

Technologies such as these have all been developed within the marine environment, which has proven itself as an ideal development arena for innovation. As the offshore energy industry has blossomed, so have the tools that serve it. In the future, we should expect to see the technologies developed here, such as autonomous robotics and SLAM, permeate into other sectors and realise their true potential. That’s a future I’m incredibly excited about.