The transformation of maritime operations

2024 proved to be the breakthrough year for artificial intelligence (AI), seeing rapid adoption across a wide variety of industries. The maritime industry is not isolated from such developments as it stands at the threshold of a profound transformation, perhaps the most significant since the adoption of containerisation or transition away from sail. The integration of digital technologies, artificial intelligence, and autonomous navigation systems (ANS) promises improved efficiency, cost savings, and enhanced safety. However, the adoption of these technologies is not without challenges and the transition could prove a choppy crossing. 

A complex and vital industry

Digital technologies have already revolutionised several industries and many potential benefits have been identified for maritime including: enhanced route planning, optimised operations, and financial savings through lean crewing. However, with these incentives comes the critical need to ensure the safety and reliability of AI-driven decision-making. Moreover, the maritime sector has been slow to adopt these advancements owing to lack of assurance frameworks and regulatory uncertainty and a lack of compelling business cases.

The shipping industry is inherently complex with a wide array of participants interacting in diverse and sophisticated ways. It plays a vital role in the economic stability of nations, acting as the arteries of trade and commerce. Consequently, the risk of emergent behaviour in AI-driven maritime systems is a significant concern. Existing systems and business models acknowledge that human operators will continue to play a pivotal role in overseeing and ensuring the safety of these evolving systems. However, the complex nature and unintuitive behaviour of such systems necessitate the adoption of a more human-centred and through-life approach in maritime digital transformation and adoption of autonomous systems.

Challenges in maritime autonomy

Despite the potential benefits there exist considerable gaps in how digital transformation and AI development are being approached in the maritime sector. Current regulatory frameworks and design methodologies do not place sufficient emphasis on enabling human oversight and meaningful agency, while simultaneously expecting human operators to provide a crucial layer of protection from inadequate or unintended behaviours. Furthermore, as digitalization extends into shore-based systems, frontier AI models, such as large language models (LLMs), are beginning to impact ship masters' decision-making autonomy, raising concerns about liability, crew safety, and operational resilience.

Additionally, naïve approaches to achieving efficiencies and the unpredictability of AI behaviours risks creating brittle systems that lack robustness in handling disruptions—the effects are multiplied in highly connected systems. As seen in other industries, approaches that prioritize raw value extraction over resilience can introduce unforeseen risks such as stock market crashes like the Flash Crash seen in 2010 making regulatory oversight and rigorous testing essential.

Plotting a course forward

For these transformations to succeed on a global scale, reliable communication, and other maritime autonomous infrastructure (MAI) e.g. Remote Operating Centres, VHF Data Exchange System, collaborative innovation, and structured regulatory evolution are required. The Organisation for Economic Co-operation and Development (OECD) has reported that a 1% increase in a nation's GDP can result in a 4% rise in trade. This highlights the potential for maritime innovation to be an enabler of equitable economic development in a period of stagnation and feared retrenchment. However, developing the right infrastructure requires new and sustainable business models for telecommunications companies and national investment. Moreover, different nations will face unique challenges in implementing these technologies depending on climate, geography, regulator development, and available skills. Achieving this will require locally tailored and inclusive approaches to realise the benefits.

Addressing these concerns now, collaboratively, in a pre-competitive phase will bring significant benefits but it requires industry experts, regulatory bodies, and academic institutions to:

1. Develop Equitable Regulatory and Standards Frameworks Ensuring that policy is evidence driven and regulatory evolution is not dictated by a few powerful entities but rather reflects the diverse realities of global maritime stakeholders.
2. Bridge the Knowledge Gap and Upskill Workforces Through the provision of access to professional development of mariners, supporting knowledge transfer between sectors and establishing skills programme tailored to the needs of different nations and maritime operators.
3. Facilitate Safe and Secure AI Integration To ensure that AI-driven maritime technologies are developed and deployed in a way that prioritizes safety, security, reliability, and resilience.

Ensuring the safe implementation of autonomous systems

A key challenge in maritime autonomy is the assurance of AI and Autonomous Navigation Systems (ANS)within complex, dynamic environments. Traditional regulatory approaches are often prescriptive and unsuitable for emergent AI behaviour and the significant heterogeneity of current approaches. In response there is increasing adoption of Goal Based Standards (GBS) such as in the IMO MASS Code to facilitate adoption of diverse solutions. Compliance with GBS is a new departure for many communities in the maritime sector, and the lack of familiarity and guidance has led to delays in gaining certification and approval even for trials. One approach for complying with GBS is to adopt Safety/Assurance Cases—widely utilized in safety critical industries like aerospace — which offers a promising solution in combining structured argumentation with evidence from rigorous testing and simulation.

However, to achieve this requires software developers, system integrators, regulators, and operators to work together to:

• Clarify Safety Evaluation Processes: Establishing clear safety evaluation standards and validation methods for AI in maritime applications.
• Define Functional Performance Requirements for ANS Components: Ensuring that system components meet the necessary safety and reliability benchmarks for their deployment domains and throughout their useable lifetime.
• Establish adequate and proportionate testing approaches: The generation of evidence is essential for meeting GBS. AI driven technologies require a shift to testing methodologies that are demonstrate requirements are meet, comprehensively exercise the algorithms and are proportionate to the risk.
• Leverage Cross-Domain Expertise: Drawing from best practices in aerospace, automotive, and other safety-critical industries to inform maritime standards development.
• Develop Comprehensive Risk Assessments and Regulatory Frameworks: Using structured methodologies, to assess the maturity of current technologies ahead of their deployment and identify significant safety concerns where present.

Collaboration for a safer, smarter future

The maritime industry's transition toward autonomy cannot succeed as a simply technological shift—it requires a systemic transformation that recognises the socio-technical complexity of the maritime sector and take a holistic approach to identifying affect stakeholders. By combining the expertise of maritime operators, AI developers, regulatory bodies, and research & education institutions, as well as the affected coastal communities, infrastructure providers and third parties such as logistic hubs, we can create a future where autonomous maritime systems operate safely, efficiently, and equitably.

Through the development of robust regulatory frameworks and assurance methodologies, the maritime industry can realise the full potential of these transformational technologies while maintaining the highest safety and operational standards. Now is the time to bridge the gap between innovation and regulation and ensure a future where autonomous maritime systems are not only technologically innovative but also safe, resilient, and globally accepted.

CfAA’s expert, peer-reviewed guidance on assuring the safety of autonomous systems in the maritime sector and other domains is publicly available and free to download.

References: 

1. CfAA, “Navigating Towards Autonomy: Trust in Maritime Safety Assurance Grows” (2024)
2. McKinsey & Company, “The state of AI in early 2024: Gen AI adoption spikes and starts to generate value” (2024)
3. The Economist, “Flash Boy, The curious case of the Hounslow day-trader” (2015)
4. IMO, “Autonomous shipping” (2024)