Part I
Introduction
1 Introduction
Henrik Ringbom, Erik Røsæg, and Trond Solvang
Interest in autonomous ships has grown exponentially over the past few years. Whereas a few years ago, the prospect of unmanned and autonomous vessels sailing on the seas was considered unrealistic, the debate now centers on when and in what format and pace the development will take place.
Law has a key role to play in this development and legal obstacles are often singled out as principal barriers to the rapid introduction of new technologies in shipping. In this field, too, development has been very rapid. Within a few years, autonomous ships have turned from a non-issue to one of the main regulatory topics being addressed by the International Maritime Organization (IMO). However, the regulatory discussion is still in its infancy, and while many new questions have been raised, few answers have been provided to them to date.
Increased automation of tasks that have traditionally been undertaken by ships’ crews raises interesting legal questions across the whole spectrum of maritime law, which makes it particularly interesting to study the development from several different perspectives. Starting from the question of whether this type of shipping, can be introduced at all under the current law, more or less every aspect of maritime law is affected by the development. Even partial developments toward autonomy in shipping raise intricate questions relating to shipowners’ liability, collision rules, transport contracts, marine insurance, and many more.
The issues range from theoretical challenges of traditional legal concepts, such as the re-defined role of the ship’s master, or culpability of machine-based systems, to highly practical questions, such as the implementation of inspections, operations, and documentary requirements on ships with no crew on board.
However, the emerging interest in autonomous ship operations among legal researchers has not yet translated into widespread publication. While several legal academics have been involved in research projects in the past five years, legal publication through open academic channels remains relatively scarce and limited to a handful of authors. Until now, there has been no single book dedicated to the legal dimension of autonomous shipping.
The present book covers a broad range of topics linked to ship autonomy, spanning public and private law issues from international as well as national, mostly European, perspectives. The 17 substantive chapters analyze structures and frameworks linked to their specific topic, rather than linking the research questions to specific on-going initiatives or available projects. By focusing on the more conceptual implications of a shift toward autonomous operations in shipping, it is believed that several important but less urgent matters linked to the development, which have not yet featured in the discussions, could be highlighted and assessed, while at the same time extending the book’s temporal lifespan and relevance.
Aside from the thematic chapters analyzing different substantive topics, the book also seeks to assist the international research and regulatory communities in developing the necessary terminology and conceptual tools for addressing the matter. The first substantive chapter (chapter 2) is dedicated to this topic, which clearly is one of the key challenges facing international regulators in this field.
The substantive chapters are divided into four parts. The first part addresses issues belonging to the domain of public international law. The main convention in the field is the 1982 UN Convention on the Law of the Sea (UNCLOS). Not all aspects of this “Constitution for the Oceans” are addressed here. A particular focus is on law of the sea questions linked to at-sea enforcement by and against autonomous vessels. As the chapter by Anna Petrig demonstrates, the prospect of security-related crimes committed through the use of autonomous ships raises a broad variety of issues linked to the applicability of existing international instruments, including UNCLOS, as well as a renewed need to interpret key maritime terms, such as “ship,” “crew,” and “piracy.” The more technical standards for autonomous ships will be established through IMO instruments, which are addressed in two of the chapters. First, Henrik Ringbom reviews the developments within the organization to date, notably the work undertaken within the framework of the “Regulatory Scoping Exercise,” and asks whether it has been doing the right things to date. The answer is not entirely affirmative. Second, the prospect of national exemptions from various IMO requirements is addressed in the chapter by Frank Smeele, with a view to exploring the potential of individual (flag) states to take early measures on autonomous ships, e.g., in the form of trials, even before IMO rules have been agreed. The author concludes that the exemptions generally tend to be of limited scope, but that the recent IMO guidelines for trials with autonomous ships may provide an opening for states eager to go ahead with experimental solutions.
The second part concerns the liability of operators of autonomous ships. The five chapters in this part cover two main strands of issues: shipowners’ liability and product liability. Focus is either on an Nordic/European approach or the regulatory situation under English law. While a revision of existing liability rules may not be strictly necessary for the development of autonomous ships, it is clear that the development places significant strains on the current liability system, which is based on (human) fault or negligence as a foundation for the liability. If and when accidents involving autonomous ships start occurring, questions as to who is liable, and for what, will inevitably be raised. While it seems agreed that any liability rule should be designed in such a way as to not shift the risk of being left uncompensated onto innocent third parties, views diverge, also between the contributions by Felix Collin, Trond Solvang, and Erik Røsæg, as to whether the current liability regimes are adequate for autonomous ships and how they could be developed. In any case, it seems clear that higher dependency on equipment, technology, software, and algorithms, will bring along a whole new range of persons and scenarios affecting how ships are operated on the oceans. This will most likely result in new forms of liability rules, such as product liability rules, discussed by Vibe Ulfbeck and Robert Veal, receiving much more prominence as a complement to traditional maritime liability law.
The third part addresses marine insurance. Here, too, the contributions emphasize different aspects of the challenge linked to insuring autonomous ships. Trine-Lise Wilhelmsen and Hans Jacob Bull address the Nordic Insurance Plan, Meixian Song addresses the English insurance practice, while Tim Howse addresses the P&I system. They all argue that the challenges relating to autonomous ships are manageable. Mika Viljanen, on the other hand, questions whether the insurance market can handle autonomous ships from a risk perspective.
The final part covers a selection of other issues that are being affected, in different ways, by the developments toward autonomous ships. They cover fields as diverse as pilotage (Martin Davies), manning (Johan Schelin), and seaworthiness (Frank Stevens). Similar studies could have been equally relevant and interesting to undertake in respect of nearly any other classic discipline of maritime law (including transport contracts, salvage, and enforcement of maritime claims). While this was not feasible for reasons of space, the three selected topics already illustrate the variety of ways in which autonomy affects traditional areas of regulation, ranging from purely practical matters to financial and technological challenges, all of which are closely linked to the general legal framework.
The approach taken in this book is to study these issues at a somewhat higher level of abstraction than merely analyzing the legality of on-going ship automation projects and developments under consideration. By focusing on underlying frameworks and structures and on the more conceptual implications of a shift toward autonomous operations in shipping, several crucial, but perhaps less immediate, matters linked to the development, are highlighted and assessed. This somewhat more theory-oriented approach allows the authors to assess their topics, consisting of very tangible legal challenges, in a general manner that is of relevance to all jurisdictions, irrespective of the national legal system in place. This will somewhat alleviate the bias of the various chapters toward Nordic and – less so – English law, and will hopefully at the same time extend the book’s lifespan and relevance.
All website references have been verified and updated as per May 10, 2020.
2 Terminology and concepts
Henrik Ringbom and Felix Collin
1 Introduction
The development of new technologies aimed at increased automation in shipping has given rise to significant terminological confusion. Terms such as remote operation, automation, autonomy, intelligence, unmanned ships, autonomous ships, cyber-enabled ships, and smart ships are widely used but rarely defined.
Among the efforts to provide some conceptual consistency in this field,1 the work of the International Maritime Organization (IMO) is particularly important, in view of the organization’s key position as the international regulator of shipping. For the purposes of its on-going work in the regulatory scoping exercise, the IMO has defined a Maritime Autonomous Surface Ship (MASS) as being “a ship which, to a varying degree, can operate independently of human interaction.”2 The definition indicates, first, that it is the division of tasks between humans and technology which is of interest and, second, that it represents a sliding scale in which tasks may be performed by technologies “to a varying degree.”
The IMO’s working group on MASS has identified the following four “degrees of autonomy”:
However, the four degrees are of limited assistance outside the scope of the IMO’s own regulatory scoping exercise and are therefore not suitable as a basis for building a broader conceptual framework. Two of the four degrees of autonomy relate to remote control, which is not, strictly speaking, related to autonomy, but rather to the location from which human functions are performed. Relocating crew from the ship to shore raises very different legal issues to those arising from the replacement of crew functions by on-board technology and the two aspects should therefore be separated.
Another problem with including both crew location and autonomy in the four degrees of autonomy is that such a division leaves only two parameters for each aspect which, in turn, fails to appreciate their sliding scale in practice. In addition to either being fully manned or completely unmanned, a ship might very well operate by means of a reduced crew and thus be unmanned only periodically. Similarly, rather than having to choose between mere “decision support” and “fully autonomous,” as defined by the IMO earlier, it seems quite plausible that key navigation functions of MASS will be only partially performed autonomously, for example in a form where human intervention is available, but is only activated by alarms if pre-set parameters cannot be maintained by the system.
In view of such shortcomings, this chapter seeks to identify and clarify more broadly certain key concepts linked to autonomous shipping and describe how they relate to the associated regulatory challenges. The most immediate purpose of the chapter is to assist the contributors to this book by providing a conceptual framework for use in their chapters and thereby to contribute to a consistent use of terminology throughout the book. Apart from that, it is hoped that the framework and terminology outlined here could also serve a broader purpose, by contributing to a better understanding of the various elements involved in autonomy and to the development of a more uniform use of terms in the current national and international discussions on the legal challenges linked to autonomous ships. The focus on regulatory challenges places the attention on aspects that are relevant from a legal point of view, whether through public law issues, such as safety and security, or private law issues such as liability and insurance.
2 Three elements of ship automation
The starting point for the conceptual framework used here is the identification of three different elements of ship automation: technical capability, manning, and autonomy. The technical capability refers to the capacity of a ship to participate in the automation developments discussed here, i.e., remote operability and the introduction of more intelligent shipboard systems. Manning refers to the availability of competent persons to operat...