Revolutioning the seas: Rolls-Royce and Intel to build fully autonomous ships

As global commerce pushes for a more efficient and effective supply chain, Rolls-Royce is promising to disrupt the main transport mode for long-distance trade: maritime shipping. The implications of Machine Learning developments to deploy autonomous unmanned ocean-going ships by 2035.

Christopher Columbus should appreciate that he is not living in the 21st. century, since the era of onboard captains is getting to its end. As global commerce pushes for a more efficient and effective supply chain, the reinvention of maritime shipping is a must. Rolls-Royce plans to offer a groundbreaking solution: autonomous ships to reduce crew expenditures, the most critical expense in vessel operating costs [1], and to optimize maintenance activities and fuel consumption.

Machine learning (ML) is the crucial element that would make this vision succeed. ML is applied to provide the vessel with the capability of simultaneously interpreting the inputs generated by multi-sensor technologies (radars, video-cameras, LIDAR and thermal imaging) and, based on these inputs, ML also allows the ship to make fast decisions using complex algorithms [2]. In addition to enabling a more efficient use of the crew and their skills, machine learning applied to vessels would be able to optimize maintenance activities and fuel consumption. On the first track, the ability to analyze historical data and to predict failures will significantly reduce the costs of unexpected maintenance stoppages or over-conservative policies for replacing parts. The reduction of unforeseen events is critical for a process that looks for minimizing human intervention. A similar concept is applied to maximize the fuel utilization, where the analysis of a wide range of data and the use of predictive modeling can unveil new optimal points for a different set of operating conditions. Finally, autonomous ships would also reduce critical risks in the industry, such as human-error risks (cause of between 80% and 90% of all accidents) [3] and piracy risks (1.7 billion USD yearly cost in the industry) [4].

By 2035, Rolls-Royce foresees a maritime shipping landscape dominated by autonomous crewless ocean-going ships. This year announced partnership with Intel aims to provide the vessels with the capabilities of collecting and processing daily terabytes of data [5]. Even when fully autonomous, ships would still be controlled by a small crew in Remote Shore Control Centers (RSCC). So, the connectivity and communication capabilities will be critical. The message to be communicated should be bi-directional, redundant (risk minimization) and comprehensive (all the relevant information elements). The channel should be reliable enough to support such highly dense data interaction and enable RSCC to take control of the ship when needed.


Rolls-Royce will disrupt the maritime shipping market by deploying a 3 phases strategy. The company launched this year the Intelligence Awareness system, which assists crew members by identifying hazardous obstacles even in low visibility conditions or by accurately measuring the distance from any obstacle while docking in narrow spaces [7]. In the next two years, Rolls-Royce plans to introduce smart ship equipment in existing vessels and, by 2020, it expects to launch the first remotely operated ship in local waters. In the following phase, the firm plans to expand remotely operated vessels to international waters (2025) and oceanic waters (2030) and deliver fully autonomous ships by 2035.

The main challenges for self-sailing vessels are both technological and regulatory. Regarding technology, one critical success factor is the scalability or cost-effectiveness of the self-sailing solution. Without the ability to scale, it would be tough to get the buy-in from an industry ruled by large and concentrated players. Another critical element for the project to success is that the technology must be highly reliable. Risk assessment and management is a crucial factor in the maritime shipping industry, so every disruption must be exhaustively assessed to identify the new risks associated with the implementation of the new technology.

Regarding the possibilities of overcoming these challenges, I think that the rapid developments of autonomous driving vehicles would drastically reduce today’s cost structure of these technologies and increase the reliability of autonomous solutions. Thus, both scalability and consistency won’t be a significant issue for Rolls-Royce, especially if the company develops different partnerships with other car manufacturers, autonomous driving technology providers or different players in the transportation industry to negotiate with big insurance companies.

Lastly, the regulatory landscape provides a political risk component that must be assessed together with the aforementioned technological risks. The primary challenge is to create a regulatory system at a national and international level, interacting with different organisms and stakeholders. In this sense, I believe that both the autonomous driving vehicles developments and the ability of the maritime shipping industry to form a dedicated committee to manage this aspect would be critical for the technology to succeed.

All factors seem to define autonomous vessels as the top trend for the shipping industry. However, in a global context where human-less solutions would be the next standard of transportation, what would be the competitive advantages of Rolls-Royce to prevail over the entrance of new players in the maritime shipping market, such as car manufacturers or big tech companies?

(800 words)

[1] Greiner, Richard “Ship operating costs are set to increase for 2017 and 2018”, Moore Stephens UK, Oct. 26, 2017. [], accessed November 2018.

[2] Rolls-Royce, “Autonomous Ships – The next step” (PDF file), downloaded from Rolls-Royce website, [], accessed November 2018.

[3] “Human error the cause for most coastal vessels accidents in harbours”, Safety4Sea Journal, March 28, 2018, [], accessed November 2018.

[4] Graham, Luke. “Somaly piracy is back 1.7 billion dollar problem after shipping firms lower vigilance.” CNBC news, May 3, 2017.[], accessed November 2018.

[5] “Rolls-Royce and Intel announce autonomous ship collaboration”, October 15, 2018, on Rolls-Royce website, [], accessed November 2018.

[6] Rolls-Royce, “Future remote shore control centers”, YouTube, published March 26, 2016, [], accessed November 2018

[7] “Rolls-Royce offers ship navigators a bird’s-eye view with Intelligent Awareness game-changer”, June 3, 2018, on Rolls-Royce website, [], accessed November 2018.



How traditional automotive OEM, Volvo Cars, embraces autonomous driving


Duolingo: Machine Learning Our Forgetfulness

3 thoughts on “Revolutioning the seas: Rolls-Royce and Intel to build fully autonomous ships

  1. The Rolls Royce – Intel collaboration on autonomous has had lots of press coverage. The picture is impressive; a futuristic vision of humans controlling robot-filled seas (and, by extension, world). People who actually are on board of these ships spend months away from home, so -even though it pays very well- this might be a move that will encounter little resistance on the worker-replacement issue.
    However, as you mention, Rolls Royce might have moved first, yet it won’t be the last. I see little competitive advantage in their capabilities, yet there are huge investment capitals necessary to make this work. Not only on the technology side (developing the software), but also -more importantly- in the ship-building side of this project. In the context of squeezing profits in the global shipping industry [1], I think that this partnership will not be long-lived.


    1. I think the partnership will be successful and their position defensible for one reason: the data set. Rolls Royce is also partnering with Maersk, the industry leader in shipping, which will allow them to continue to develop their models with additional data added from Maersk ships. Through this additional data, they will have better inputs and therefore better operational outcomes. As you say, there will be large CAPEX necessary to launch similar projects, making it prohibitively expensive for another company to achieve a similar level of quality and gain the industry’s trust.

  2. Thanks for this very interesting post. One of the challenges for machine learning and autonomous vehicles is what is the level of computer error that is acceptable. If a self-driving car gets into an accident, are we willing to allow the algorithm to continue to function? What is an autonomous ship crashes and spills oil? Autonomous ships are undoubtedly becoming safer, but I still think there is human bias against autonomous vehicles when they are involved in safety issues – even if they are on average safer than humans.

Leave a comment