News

Ladies and Gentlemen,

1.     Good morning.

Artificial Intelligence in the Maritime Sector

2.     The focus of this year’s SMI Forum on digitalisation and Artificial Intelligence (AI) is timely. AI, Big Data, Machine Learning – these are fields that have the potential to achieve the following positive outcomes: Uncover deeper insights about our operating environment; Improve the efficiency and environmental sustainability of the maritime sector; and augment the capabilities of our workers.

3.     We need to combine technical expertise in AI and machine learning with a practical understanding of the challenges and needs of the maritime industry. Some of the experts in AI and machine learning may not be the exact same group of individuals as the people who know the maritime industry well. Similarly, maritime industry veterans may not all be AI experts. The key is how we can bring together our skills and knowledge across different fields by encouraging collaboration and partnerships.

4.     There are many possibilities to apply AI and machine learning in the maritime domain. For example, can we use ship incidents and vessel data to train AI models to classify vessels according to their risk profiles, so that classification societies, ship operators, insurers and other stakeholders can make better commercial decisions about their vessels? Another example, can we use AI models to analyse data already collected today on vessels’ engine conditions, fuel consumption and voyage information, so that regulators can validate vessels’ declared emissions without installing additional sensors or devices on vessels. Or, can we develop AI and machine learning models to help ship owners to compare and identify the most effective energy-saving technologies that they can install on their vessels, based on their vessel specifications and operational needs? These are just some of the examples of the many possibilities where we can apply AI.

5.     In April this year, SMI launched the Maritime AI R&D Grant Call with support from the National Research Foundation, to develop new AI tools and applications for the maritime industry. SMI will be awarding grants totalling $4.1 million to three project teams from NUS and SIT to develop AI solutions for the challenges that I spoke about earlier. SMI will also be establishing the new Maritime AI Programme in partnership with A*STAR’s Institute of High Performance Computing (IHPC). The programme will serve as the central R&D node to coordinate the development of AI capabilities and facilitate industry adoption.

6.     SMI has awarded S$4.8 million for Phase 1 of the programme, which will focus on building capabilities in maritime data and AI modelling. The programme will develop AI toolkits to tackle issues such as maritime traffic safety management, predictive maintenance, and monitoring and reduction of carbon emissions. I would like to thank SMI for its efforts in advancing Singapore’s R&D initiatives in maritime AI.

The Growing Necessity of Cybersecurity

7.     As we digitalise, we must also ensure that our technology, assets and data are well-protected, including from malicious cyberattacks. Reported incidents of cyberattacks on the maritime industry’s Operational Technology (OT) systems increased nine-fold between 2017 and 2020. Such attacks can target not only land-based digital systems, but important control systems on board ships as well.

8.     In 2017, cyber criminals took full control of the navigation systems of an 8,200 TEU container vessel for 10 hours. In 2019, hackers disabled a cargo vessel’s computer network as it headed into the Port of New York and New Jersey. I am glad that SMI has stepped up R&D efforts in maritime cybersecurity.

9.     SMI funded a study by the iTrust Centre for Research in Cyber Security into the strengthening of shipboard OT systems. This culminated in the publication of iTrust’s guidelines for cyber risk management for such systems in February 2022. The guidelines provide a comprehensive and in-depth study of the possible threats to shipboard OT systems, and includes a checklist of actionable measures to mitigate the cyber risks from such threats. 

10.    We know how these things evolve – it is a bit of a cat-and-mouse game. As we develop new security measures, actors with malicious intent will develop counter-capabilities to attack, hence we need to continue to pump in additional resources to stay ahead of the technology. 

11.    I am pleased to announce that SMI will be awarding iTrust with S$4.8 million to develop MariOT, a maritime testbed for shipboard OT. Currently, there are two main ways to test a new cybersecurity product for shipboard systems. The first is to install the product on an actual vessel for testing, but this could be highly disruptive for the vessel’s operations. The second way is to test the product through virtual simulation, but this may not capture the realistic shipboard environment in which such products will need to operate.

12.    As the world’s first cyber-physical facility for maritime cybersecurity capabilities, MariOT will provide a third option that offers the best of both worlds. MariOT will replicate the physical and digital components of shipboard operating systems, while also using simulators to simulate cyberattacks on such systems. MariOT will thus offer a realistic, safe, and controlled testing environment for cybersecurity products, without disrupting the operations of actual vessels.

Deepening Capabilities in Established Areas of Interest

13.    Even as SMI moves into new research areas like maritime AI and cybersecurity, SMI has continued to deepen R&D capabilities in established areas of interest which continue to be of concern to the maritime industry. Two of SMI’s Centres of Excellence are approaching the end of the first five-year phase of their research.

14.    The Maritime Energy and Sustainable Development Centre of Excellence (MESD) has completed 14 projects in the areas of energy management, emissions management, and sustainable maritime operations. This includes projects on developing ammonia bunkering operations and infrastructure in Singapore, exploring the feasibility of methanol-fuelled vessels in China and Singapore, and studying the use of bio-LNG in the shipping industry.

15.    The Centre of Excellence in Modelling and Simulation for Next-Generation Ports (C4NGP) has advanced its research into next-generation port systems and platforms through the completion of 9 projects. 

16.    This includes the development of a maritime simulation platform for optimising algorithms for port operations, projects on vessel collision avoidance in Singapore’s port waters, and the optimisation of automated guided vehicle operations in Tuas Port, that we are currently constructing and will be ready only in the 2040s, but we want to start researching on this capability now, so that we can be ready for the future.

17.     SMI will build on this momentum by renewing its support for MESD and C4NGP for the next 5-year phase of their research. SMI will award S$12 million to MESD to further its research into the feasibility, compatibility and safety of future and alternative fuels, such as ammonia, biofuels, hydrogen, and electrification.

18.    C4NGP will be awarded S$10 million to work on projects such as upgrading the digital twin for Tuas Port, to improve the operational efficiency and real-time management of port operations.

Conclusion

19.    While I have touched on the importance of technology being at the forefront of maritime R&D, in applying advanced technology like AI to improve our operations, one thing that I would like to bring up is that these various technologies are ultimately designed by humans, for humans. This man-machine interactions are important – we are not trying to replace human beings with artificial intelligence, or machine learning, we are trying to augment our workforce, with additional tools. So we can think of AI as added inventions, something which will help us to develop stronger capabilities, for our workers to do their jobs even better. Now, this man-machine combination has been found to out-perform humans or machines working separately. 

20.    IBM’s Deep Blue, a chess-playing supercomputer using artificial intelligence and machine learning, became so good at the game over time that it was able to defeat human chess masters. However, it lost out to a team of scientists, who were good chess players who, while were good chess players but not at the grandmaster level, combined their knowledge of chess with a computer, to help them in analysing their chess game. The man-machine combination made use of the computer’s strengths to perform complex analysis, while retaining the human player’s ability to think out of the box and come up with new, unexpected moves, eventually helping them to beat IBM’s supercomputer. I wanted to use this example to illustrate the importance for us, as we navigate the way forward looking at technology and exploring various options, to always remain human-centric. 

21.    Another dimension of how technology can remain human-centric, is also making sure that the tools and systems we develop remain user-friendly. A good example of this would be companies like Apple, or Dyson, where they still retained the user-friendliness of their products even as they applied advanced technology. In fact, they have designed it such that one does not need to be a scientist or an engineer, in order to use those tools effectively. Hence, the beauty of combining advanced technology and engineering, with very human-centric design and ergonomics.

22.    R&D will continue to be crucial in expanding the frontier of possibilities, empowering us to push the envelope to achieve more with our finite resources, to discover new solutions and better ways of working, and to turn possibilities into reality.

23.    With that, I would like to conclude by thanking SMI and the maritime R&D community for your important contributions to Maritime Singapore. Thank you very much.