Researchers at the University of Western Australia have developed autonomous underwater vehicles specifically designed to assist with decommissioning offshore oil and gas platforms, a challenge that will cost tens of billions of dollars over the next two decades as aging infrastructure reaches end of life.

Australia has over 150 offshore petroleum installations, many approaching 40-50 years of operation. Operators are required to remove these structures when production ceases, but the process is expensive, dangerous, and technically demanding.

The UWA robots can inspect underwater structures, cut through corroded steel, and monitor structural stability during removal operations. That’s work currently done by human divers or remotely operated vehicles controlled by surface operators, both expensive options.

Professor Steve Dunn, who leads UWA’s Centre for Offshore Foundation Systems, said autonomous systems could reduce decommissioning costs by 30-40% while improving safety. “You’re working in deep water, often in harsh conditions, dealing with corroded structures that might contain residual hydrocarbons. Taking humans out of that environment is desirable.”

The robots use a combination of sonar, cameras, and laser scanners to build detailed 3D maps of underwater structures. Machine learning algorithms trained on hundreds of hours of inspection footage can identify corrosion, structural damage, and other issues that might complicate removal.

One interesting capability is the ability to autonomously plan cutting sequences for removing platform legs and other structural elements. The system analyses the structure’s geometry and condition, then generates a cutting plan that maintains stability throughout the removal process.

That planning capability draws on research in structural engineering, materials science, and robotics. The team collaborated with decommissioning contractors to understand what actually happens during platform removals and where automation could add value.

The robots aren’t replacing human decision-making entirely. An operator onshore or on a support vessel reviews the robot’s plans and monitors operations, but doesn’t need to manually control every movement. That substantially reduces operator workload and allows one person to supervise multiple robots simultaneously.

The research received $8 million in funding from the National Energy Resources Australia (NERA) industry growth centre and several oil and gas companies including Woodside and Santos. Those companies have significant decommissioning liabilities and strong incentives to reduce costs.

Australia’s offshore decommissioning challenge is small compared to the North Sea, where hundreds of platforms installed in the 1970s and 1980s are now reaching end of life. The global decommissioning market is estimated at $150-200 billion over the next 20 years.

If Australian technology can capture even a small share of that market, it represents a significant export opportunity. Several international decommissioning contractors have expressed interest in the UWA technology.

The project builds on decades of marine robotics research at UWA. The university’s Australian Centre for Field Robotics has developed autonomous systems for everything from underwater pipeline inspection to seafloor mapping.

One technical challenge is operating autonomously at depths of 100-200 metres where water pressure is high and visibility is often poor. The robots need reliable sensing systems that work in murky water, plus robust power and communication systems.

The team developed new acoustic communication protocols that allow multiple robots to coordinate their activities and share sensor data underwater. Radio communications don’t work underwater, so acoustic methods are necessary, but they’re much slower and more error-prone than radio.

Battery life is another constraint. The robots can operate for 8-12 hours before needing to surface for battery swap or recharging. That’s sufficient for many inspection and cutting tasks but limits how far from a support vessel they can work.

The research team is exploring inductive charging systems that would allow robots to recharge from subsea docking stations, extending mission duration. That’s particularly relevant for decommissioning projects that might take months or years to complete.

Environmental monitoring is another application for the robots. When platforms are removed, operators must verify that no hydrocarbons or other contaminants remain on the seafloor. The robots can conduct detailed environmental surveys using water sampling and seafloor imaging.

Some environmental groups argue that certain offshore structures should be left in place because they’ve become artificial reefs supporting marine ecosystems. Decommissioning regulations in Australia generally require complete removal, but there’s ongoing debate about whether that’s always the best environmental outcome.

The UWA robots could potentially help inform those decisions by providing better data about what marine life has colonised a structure and what environmental impacts removal would cause.

Field trials of the decommissioning robots are scheduled for early 2026 at a decommissioned platform in Bass Strait. If those trials succeed, the technology could be commercially available by 2027.

Several startups are working to commercialise UWA’s marine robotics technologies. That includes companies focusing on pipeline inspection, underwater construction, and environmental monitoring in addition to decommissioning applications.

Whether these ventures succeed depends partly on how quickly the offshore industry adopts autonomous systems. Oil and gas operators tend to be conservative about new technologies, particularly for critical operations like decommissioning where mistakes can be extremely costly.

But with decommissioning costs mounting and a global shortage of experienced divers and ROV pilots, the economic pressure to automate is increasing.