AI, AI Captain: Autonomous ROVs

By Brian Allen, CEO at Vaarst
Brian Allen, CEO at Vaarst discusses the evolution of remotely-operated underwater vehicles (ROVs) and their role in enabling global communications

Fleets of robots, thinking, seeing and operating as an autonomous hive, working together to maintain critical infrastructure…all underwater. Not long ago this would have been a great intro to a Sci-Fi novel. 

Now, it’s nearing reality. Forget shipwrecks and bullion: there’s a far more modern booty that can be found on the beds of the world’s oceans. Millions of miles of oil and gas pipelines carry the fuel we use to generate our electricity, the structures that hold in place the wind turbines that will make up an increasing proportion of our energy generation infrastructure, and around 1.3 million kilometres of subsea cables, carrying an estimated 97% of the world's communications.

The treasures of the deep

It’s almost a year ago that Google, for instance, in partnership with SubCom, set a new record for data transmission capacity across a subsea cable network. The long-haul system, called Dunant, crosses the Atlantic from the US to France and has the capacity to, “transmit the entire digitized Library of Congress three times every second.”

This wasn’t the one-time search engine’s first foray into subsea cables. Between 2016 and 2018, Google invested $47 billion to improve its cloud infrastructure, which included 14 subsea cable investments around the world. 

Of course, Google isn’t the only one. Facebook is among the many others investing in subsea cables, with five currently under construction – including the 37,000km 2Africa cable project, which it’s claimed will provide nearly three times the total network capacity of all the submarine cables serving the continent today.

And this is even before you consider the vast sums of money being invested in the energy transition – particularly in offshore wind as we look to realise a net-zero economy, and prevent global warming beyond 1.5C of pre-industrial levels.

Offshore wind capacity today is only 2% of what the world needs to get to net-zero by 2050. However, the Global Wind Energy Council (GWEC) predict that 235 GW of new offshore wind capacity will be installed over the next decade under current policies. That capacity is seven times bigger than the current market size and is a 15 per cent increase on last year’s forecasts.

Building, monitoring, and maintaining all of this subsea infrastructure is big business – but it’s a business wrought with inefficiencies and emissions issues of its own.

“It’s not about building it, it’s how to protect it”

A significant level of time, money and human effort is – understandably – invested in the construction and maintenance of subsea infrastructure. Fortunately, there have been few recent incidents of major disruption. 

However, in 2008, a ship attempting to moor off the coast of Egypt in bad weather led to an internet blackout, limiting access to services for 75 million people. At the time, Mustafa Alani, head of security and terrorism at the Gulf Research Centre in Dubai, was quoted as saying, "This shows how easy it would be to attack. When it comes to great technology, it's not about building it, it's how to protect it."

Whether it’s due to shipping accidents or sand shifting with the currents – increasing the stresses and strains on sections of pipeline, keeping a close eye on these assets is critical to our energy and even national security.

Deploying specialised ships and highly trained crews to inspect and maintain these assets costs tens of thousands of dollars per day, added to which is the cost of replacing components of the infrastructure.

Enter the robots

What’s the alternative? As in many other sectors touched by artificial intelligence and machine learning technologies, it’s robots. 

Remotely-operated underwater vehicles (ROVs) have been used for years to inspect cables, pipelines and mooring chains, maintain and model offshore wind farm assets, and provide measurements and data on marine infrastructure critical to the oil and gas sector.

However, much ROV technology is not capable of sending data from a subsea site to teams working on-shore. As a result, teams of specialists are required to operate the technology from ships in close proximity to the assets being inspected, with each ROV requiring individual operators (often working in shifts). 

Teams of analysts needed to interpret the data collected and advise on action needed are often also stationed on these survey vessels, so crews can swell to up to 60 people. In addition to the financial impact of this approach – both in terms of the salary costs and vessel charters, which can be eye-watering – there are significant environmental concerns: larger vessels can generate up to 275,000 tonnes of carbon emissions in their lifetime.  

ROVs will continue to be critical to maintaining and analysing subsea infrastructure for quite some time. Yet with new installations increasing to match capacity demand, and asset owners working to extend the useable life of existing assets, the cost – to the environment and to those deploying the solutions – is unsustainable.

ROVs go indie

Recent advancements in technology have led to the development of semi-autonomous ROVs. Managed via a platform that uses SLAM technology to help ROVs understand their location, and position relative to subsea assets within a marine environment, these robots significantly streamline the marine survey practice. 

These ROVs are now starting to run on predefined strategies, with minimal supervision from operators (who can now ‘pilot’ the vehicles from onshore).

With new navigation and visualisation technologies enabling the ROVs to ‘see’ their surroundings and react in real-time, vehicles are completing their predetermined strategies – from pipeline inspection to chain corrosion analysis – whilst navigating obstacles, or course-correcting for currents. They do this with very little input from pilots, occasionally prompting them for one-touch decisions to confirm actions and complete the job.

This is now. But in the near future, we can expect to see the deployment of fully autonomous, coordinated fleets of robotic vehicles – both above and below the surface. Collaborating together, these fleets will be able to provide faster underwater data collection and analysis and can be supervised by a handful of users (or a single individual) from remote locations anywhere in the world.

ROVs get smart

This autonomy, coupled with new machine learning (ML) capabilities are now able to significantly reduce the time and costs of carrying out what was once manual, laborious work.

Video feeds from the ROV can be run through an ML platform, which can determine what it’s looking at, whether it’s supposed to look like that, and identify any anomalies in the infrastructure the vehicle is surveying in a fraction of the time it would any human. 

By classifying and intelligently grouping images that share similar properties and features, the technology enables analysts and surveyors to process images much quicker than current approaches which can involve watching hundreds of hours of video footage and identifying trends and potentially recurrent issues.

Again, none of this requires crew members to be on vessels at all, enabling the work to be done from on-shore – lowering the costs significantly and preventing needless carbon emissions.

Autonomy, AI and ML are all technologies that are being deployed to help many industries and businesses operate more efficiently and to deliver more benefits, to more people. 

We believe there is huge potential for these technologies to support the energy transition – by reducing the costs of a subsea survey by up to 80 per cent, they can dramatically improve the financial case for investing in more renewable energy infrastructure.

If we can increase the adoption of autonomy and automation across the marine and energy sectors we can save the time, resources, and costs, operators need so that they can continue to invest in renewable capacity, move to more environmentally sound monitoring and maintenance practices, and achieve net-zero before it’s too late.


Featured Articles

GPT-4 Turbo: OpenAI Enhances ChatGPT AI Model for Developers

OpenAI announces updates for its GPT-4 Turbo model to improve efficiencies for AI developers and to remain competitive in a changing business landscape

Meta Launches AI Tools to Protect Against Online Image Abuse

Tech giant Meta has unveiled a range of new AI tools to filter out unwanted images via its Instagram platform and is working to thwart threat actors

Microsoft in Japan: Investing in AI Skills to Boost Future

Tech giant Microsoft is investing US$2.9bn into AI and cloud infrastructure in Japan to upscale the country’s skills, research and cybersecurity efforts

Microsoft to Open New Hub to Advance State-of-the-Art AI

AI Strategy

SAP Continues to Develop its Enterprise AI Cloud Strategy

AI Applications

AI Consortium: Leading Tech Companies to Improve Upskilling

AI Strategy