Abstract

     Autonomous ships would require higher cyber-physical interaction in comparison with traditional shipping operations, thus increasing the vulnerabilities associated with cyber security. The increasing complexity surrounding the innate characteristics of the shipping industry makes it challenging to build a resilient framework for ensuring cyber security. This study proposes a multi-criteria decision-making (MCDM) framework for assessing cyber security risk in the autonomous shipping context. The research was validated through surveying subject matter experts, system designers and seafarers. Different types of equipment and systems are ranked based on their perceived vulnerability to cyber threats. Survey data from 28 subject matter experts are collected and analysed through the Bayesian best–worst method (BWM). At system level, the results indicate that navigational systems are the most vulnerable to potential cyber threats, while propulsion systems are the least vulnerable element in the context of future autonomous shipping operations. On a sub-system level, the three most vulnerable parts are Global Navigation Satellite System (GNSS), Electronic Chart Display and Information System (ECDIS) and the communication devices on shore control centres (SCC), while the least vulnerable parts are engine controls, SCC integration platforms and cargo handling at ports.

Introduction

     Merchant ships have been an integral part of international trade, carrying about 90% of the goods transported globally (OECD 2021). Since its inception, eforts have been made by the stakeholders to make this mode of transport safer, more efcient and more cost efective. The introduction of advanced technologies along with the skills development of the workforce over the years has reportedly increased workplace safety in the maritime domain (Chauvin et al. 2013; Allianz 2020).

     Consequently, modern technologies such as Automatic Identifcation Systems (AIS), Global Navigation Satellite Systems (GNSS) and Electronic Chart Display and Information Systems (ECDIS) have contributed to safer navigation. The increased reliability and efciency of these technologies has resulted in their broader adoption in the shipping industry such that large ships are now being efciently operated by fewer crews (Lee and Sanquist 1996). The notion that it is possible to maintain navigational safety in the future even without any crew onboard has led to the discussion towards the adoption of autonomous shipping technology. Wróbel et al. (2017) stated that the introduction of Maritime Autonomous Surface Ships (MASS) could further reduce the likelihood of navigational accidents, such as collisions and groundings, as human error played a role in the majority of navigational accidents. However, the probability and consequences of non-navigational accidents such as fre and structural failure are likely to increase if no crew members are onboard (Wróbel et al. 2017).

Conclusions

     This study has presented an integrated approach to elicit cyber-security vulnerabilities in the context of future autonomous shipping utilizing a MCDM method. Our results suggest that navigational systems are the most vulnerable to cyber-threats, whereas propulsion control systems are the least. This necessitates communication devices and protocols used for the navigation of autonomous ships be redundant and cyber-safe. Other emerging new systems in the context of autonomous ships, such as the shore control centre and associated equipment, although opaque for the time being, are potentially vulnerable to multidimensional cyber-threats as described in this study. The unpredictable nature of cyber-security threats, along with inherent repercussions, calls for a multifaceted approach to safeguard the future of shipping. A pre-emptive approach should include not only technical guidelines but also policy and industrial ones and best practices in risk management, and in the behavioural and cultural aspects of shipping. In addition, experts concur that only compliancebased actions would not be sufcient to safeguard against cyber threats. Rather, an all-out comprehensive approach, considering the above aspects, would ensure a safe and secure adoption of autonomous shipping.

The fifth assessment report published by the Intergovernmental Panel on Climate Change indicates that international and coastal shipping accounts for nearly 10% of global transport sector CO2 emissions [1] or 2% of CO2 emissions from fuel combustion [2,3]. According to the third International Maritime Organization (IMO) study on carbon emissions [4], maritime transport emits around 1000 million tonnes of CO2 annually and is responsible for about 2.5% of global CO2 emissions. Depending on future economic and energy developments, maritime transport carbon emissions are predicted to increase between 50% and 250% by 2050. A combination of operational and technological measures can help reduce ship’s energy and hence carbon emissions by up to 75% [5]. Among the list of ten most effective measures in reducing energy intensity, and hence carbon emissions, speed reduction has ranked top of the list [6]. Studies in the literature generally employs life cycle analysis (LCA) or other analytical methods to explore the decarbonization potential of maritime transport or shipping. Existing studies on shipping tend to focus either exclusively on just the marine vessel or to provide a system view of the maritime freight over the vessel’s lifetime. Those focusing on the marine vessel are further divided into studies on the complete life cycle of the vessel from manufacturing to end-of-life scrapping [7], shipbuilding [8], vessel operation [9,10], docking at ports [11], and ship scrapping and recycling [12–16]. These studies can provide insights on specific stages of a larger maritime transport system, but they are unable to provide a holistic understanding of the life cycle system of shipping. Those examining the larger system of maritime freight are further divided by focus areas and methods. Some studies focus on examining technical and operational measures for reducing carbon emissions such as [17–20]. Other studies focus on the development and/or use of LCA software to study the environmental impacts of shipping such as [21,22]. Relatedly, some studies tend to focus on the issues related to LCA methods, such as adaptation of LCA methodology to suit maritime transport, system boundaries selection, and life cycle inventories (LCIs) [23]. Although the existing literature has somewhat covered most aspects of LCA studies on shipping, the number of LCA studies on maritime transport is far smaller as compared to the number of LCA studies on energy systems. That has led to even smaller number of LCA studies addressing the critical issues related to system and boundary formulations to ensure consistent and unbiased results. The scarce number of LCA studies is likely caused by the fact that all stages of shipping have already been studied in quite some detail. Furthermore, it is conceptually intuitive and has been demonstrated by existing studies that the shipping stage accounts for the majority of the life cycle carbon emissions. Since fuel typically accounts for around 40% of the total cost in shipping, freight business operators have a strong incentive to reduce transport speed [24]. However, the metrics used for evaluating the decarbonizing potential of speed reduction have not been thoroughly evaluated from a life cycle standpoint. Due to the lack of studies focusing on system, boundary and input-output definitions, many other similar issues as “speed reduction” require further independent assessment using alternative LCA methods. More importantly, the scarce number of LCA studies on shipping also highlights the need for further developments in LCA methodologies for analysis of systems providing services.

Global offshoring has increased both the international movement of end products and the need for the transport of half-finished goods and components (Gereffi and Lee 2012). Undoubtedly, the majority of low-tier suppliers have become heavily dependent on remote manufacturers and suppliers in the globalisation process; specifically, the auto-parts industry has entered the global market as car manufacturers participate actively in global operations (Humphrey 2003; Sturgeon, Van Biesebroeck, and Gereffi 2008). Auto-parts suppliers can also benefit from involvement in global operations through economies of scale that can be realised by both aggregation and possible market growth. Several challenges, however, emerge at the same time. Auto-parts suppliers usually depend heavily on global auto manufacturers for their operations and logistics management, which makes them vulnerable to market changes or macro-policy decisions taken by large manufacturers. Taking advantage of the power they have over suppliers in their relationships, vehicle manufacturers explicitly or implicitly force suppliers to provide auto-parts to meet the time frame of production planning. Because delay is unacceptable despite the length of the supply chain, autoparts suppliers tend to accumulate extra buffer inventory or use expensive logistics and transportation options (Nieuwenhuis, Beresford, and Choi 2012). In addition, suppliers are exposed to even fiercer competition in the global market than in the domestic market. They are expected to improve the quality of their products while maintaining a price level that is reasonably competitive in the global context. Whereas a long-term strategy to stay competitive in the market could be based on technological advancement or investment in facilities, the short-term option tends to be logistics cost reductions. Therefore, having a clear and effective logistics strategy is a primary prerequisite for auto-parts suppliers in the global market.