Beyond the Hype: Five Lessons from Deploying Autonomous Robots on Unmanned Offshore Platforms

Today, the question is rather: how do we use robotics to fundamentally improve the way offshore assets are managed?

Our experience suggests the answer means going beyond simply deploying robots. It requires using autonomous systems to collect consistent, reliable operational data.

Our autonomous ExR-2.5 inspection robots are now carrying out routine inspection missions at facilities operated by companies including Shell, BP and Repsol.

They collect critical operational data multiple times per day, while reducing the need for personnel to enter hazardous areas.

Working with Shell and our project partners, we deployed what is believed to be the world’s first operational inspection robot on a normally unmanned offshore platform.

Six months later, it was still completing two autonomous inspection missions every day without physical intervention.

This experience led us to develop five key lessons every operator should know when considering autonomous inspection for their assets.

Lesson 1: Start with the Operational Challenge

The most successful robotics projects don't begin with technology, they begin with identifying the problem and the operational outcomes.

For Shell, the objective wasn't simply to deploy a robot. It was to improve the reliability of a newly commissioned multiphase pump on a normally unmanned installation. This was done by collecting consistent inspection data while reducing the need for offshore visits.

The robot's daily missions were designed to monitor pressures, temperatures, vibrations and any signs of leakage, providing engineers with the information needed to support predictive maintenance.

That clarity of purpose influenced every subsequent decision, from inspection routes and sensor selection to the required communications infrastructure and docking location.

Lesson 2: Preparation Determines Success

One of the biggest surprises wasn't how much work went into deploying the robot offshore. Rather it was how much happened beforehand. Robotics projects are often judged on day one, but their success is determined months earlier.

Our deployment team in the Netherlands recreated the platform's grid mesh flooring to confirm safe manoeuvring. Engineers used Shell's digital twin to simulate inspection routes and verify inspection points could be reached. Wireless connectivity, docking locations and crew training were all assessed before installation.

By the time the robot reached the platform, most technical risks had already been addressed.

The remaining challenges were practical integration issues. They included Wi-Fi bandwidth, navigation tuning and obstacle mapping, which were resolved as the deployment matured.

Robotics is every bit as much about preparation and operational readiness as it is about engineering.

Lesson 3: Reliability Matters

Whenever I speak at conferences, people understandably ask about artificial intelligence, sensors or computing power. Very few ask the question that really matters most.

Will the robot still be carrying out its inspections six months from now?

That is what operators ultimately invest in. Operators need confidence that inspection data will continue arriving regardless of location or operating conditions.

On Shell's normally unmanned offshore platform, the robot continued operating in an isolated environment exposed to tropical rain, corrosive sea salt spray and hazardous conditions. All while completing scheduled inspection missions without physical intervention.

This is why certification matters. Operators need confidence that autonomous systems can operate safely in hazardous environments under recognised industry standards. The ExR-2.5 became the first robot certified under UL Solutions’ UL 6260 standard for hazardous-location inspection robots in North America.

Reliability is not simply another specification; it is the foundation on which trust is built.

Lesson 4: Autonomous Doesn't Mean Unattended

One of the biggest misconceptions about robotics is that autonomy removes people from the equation. In reality, it changes the role people play.

During the Shell deployment, inspection missions were programmed remotely and monitored by a Human Robot Supervisor more than 1,000 kilometres away. As experience grew, around 80% of missions were completed without intervention, with the remainder successfully supported remotely.

Rather than spending hours collecting routine inspection data, engineers can focus on interpreting information, diagnosing issues and making better maintenance decisions. This is increasingly important as the industry faces a growing skills challenge, with experienced personnel approaching retirement and taking valuable knowledge with them.

Autonomous workflows can help preserve that expertise by embedding inspection routines, alarm thresholds and operational judgement into future operations. Technology does not replace experience; it amplifies it.

Lesson 5: Robots & the Bigger Digital Ecosystem

The biggest mistake operators can make is to think of robotics as a standalone technology. It isn't. The future belongs to connected operations, where robots, drones, digital twins, fixed sensors and artificial intelligence work together to create a continuously updated picture of asset health. We're already seeing this happen.

We see other customers deploying underwater and ground inspection robots to carry out routine inspections during live operations, while using integrated digital twins of all nine offshore platforms to simulate scenarios and plan maintenance remotely.

On Shell’s normally unmanned platform, the value was not simply that a robot completed an inspection route; it was that inspection data was automatically collected and made available to engineers without requiring offshore visits.

As autonomous inspection becomes more widely adopted, robots will increasingly become part of a connected operational ecosystem, working alongside drones, fixed sensors and digital twins to provide a more complete picture of asset health.

Looking Ahead

The offshore industry has always embraced technologies that improve safety, reduce risk and increase operational efficiency. Autonomous inspection is following the same path.

Within the next decade, the question will not be whether operators should deploy inspection robots, but how effectively they integrate autonomous systems into everyday operations.

The industry does not need robots for the sake of automation. It needs better information.

Autonomous inspections give operators the timely, consistent and reliable insights needed to make safer, faster and better-informed decisions.