The regulatory challenges relating to Marine Autonomous Surface Ships (“MASS”) involve two main areas of interest. First, there are the public law rules, covering safety, manning, technical and operational standards etc. for ships. These are the rules that determine whether it is lawful to operate MASS in the first place. Second, there are the liability rules which enter the scene when things have gone wrong. A key question here is how the increased automation in shipping may impact the distribution of liability among the parties involved, including entirely new players, such as software developers and providers of situational awareness (lookout), navigation and communication technology for MASS.
Uncertainty surrounding liability rules does not necessarily prevent the introduction of MASS. The backbone of the existing maritime liability regime, i.e. the owner/operator of the ship, does not disappear with increased automation, nor does the flag state. The novel liability issues that MASS introduce, such as the attribution of fault to computer systems or the scope liability rules governing errors committed by technology providers with only remote links to the ship, need to be settled over time, notably to ensure that the new developments do not represent unreasonable risks to third parties. However, it is neither necessary nor realistic to achieve clarity and consensus on these matters before such ships are in operation. This means that such matters will probably be accommodated gradually, once there is more experience with the practical operations of MASS, including through national standards and/or case-law-based interpretation. Gradual development, in turn, implies national and regional variations, which may necessitate discussion on whether a new international instrument is needed.
It could well be that the difference between MASS and other ships in this regard has been overemphasised. Are failures of MASS technology really that different from other technical failures on board ships (which are routinely handled by the existing legal system in all parts of the world)? Most analysts appear to answer that question affirmatively and call for some form of strict liability regime for the ship operator to fill the legal void
s. A related question that has received less attention is whether the identified differences are limited to MASS. Should one not ask instead whether the increased reliance on technology on board ships justifies a closer look at the general liability rules for ship operations, and the range of persons included, rather than the creation of a (necessarily random) legal differentiation between MASS and other ships?
By contrast, there must be a good level of certainty and consensus on the applicable safety requirements before MASS can be brought into operation. This is particularly so for MASS engaged in international trade, but some of the key international rules apply in any sea area and it is not obvious that the territorial sovereignty of the state concerned will always trump such international commitments. The place to resolve questions related to safety requirements is the IMO, as international technical rules will make it significantly easier for other regulatory lawyers to follow suit. This applies ‘upwards’, to the jurisdictional regime governing flag and coastal states’ rights and obligations laid down in the 1982 UN Convention on the Law of the Sea, as well as ‘downwards’ in relation to national legal systems.
There is already a considerable amount of research and literature highlighting the legal difficulties involved in accommodating MASS into the existing legal framework. Maritime lawyers from all quarters, including academics, civil servants, and the IMO itself through its recently completed ‘regulatory scoping exercise’, highlight the differences of MASS as compared to traditional ships and the difficulties that arise from the wording of the existing conventions. A common assessment appears to be that a few amendments, some adjustments, and several general clarifications are needed, but that MASS generally give rise to few direct conflicts with the existing rules (the main exception being the watchkeeping provisions of the STCW Convention). Despite the decidedly human-oriented focus and background of today’s international safety rules, authorization of MASS would accordingly not require a major overhaul of the existing rules.
Oddly, little attention has been given to the question of how the numerous regulatory voids that MASS have unveiled should be filled, which arguably represents a much greater regulatory challenge than the adjustment of existing rules. Current rules on lookout, for example, are not easily translated into purely technology-based situational awareness information, which will be needed at any degree of autonomy or manning reduction. Machine-based decision-making in navigation is another regulatory blind-spot. For these and other new areas for regulation, standards will be needed to modify the current human performance-oriented rules into a new quantifiable format, but also to develop common rules and principles for data handling, communication, redundancy systems and emergency procedures, to name but a few.
The goals and content of such rules will have to be specified internationally in due course. However, the existing IMO framework already offers a range of mechanisms and procedures that could be used for advancing MASS developments. Most of them operate through flexibility offered to the ship’s flag state administration, and all are based on the principle that the overall safety level must not be compromised.
There is, for example, already an established systems for exceptions, alternative designs, and equivalent solutions for technical standards. This may prove an important avenue for flag state administrations to approve variations to technical standards, notably those laid down in the first three chapters of SOLAS and the Load Lines Convention.
Another agreed procedure currently in place relates to the approval of alternative designs, where accepted, through a complex dialogue between the owner and the administration, as outlined in IMO Documents MSC.1/Circ. 1455 and MSC.1/Circ.1212. This procedure might usefully form a basis for the approval of goal-oriented technical construction and equipment requirement for MASS too.
For operational requirements there are fewer exemptions or alternatives available. Generally the rules that regulate human activities on board ships, including the COLREGs, cover all ships and include limited scope for exceptions. However, the assessment of the safe manning of individual ships laid down in SOLAS Regulation V/14 and associated guidelines, includes a general appraisal of all functions required to be performed on the ship in question. This procedure could serve as the centrepiece for the assessment of the operational aspects of MASS, probably without much modification. Yet, the (national) criteria on which the assessment is based, in particular with respect to how technology solutions may assume duties traditionally performed by humans, need to be adjusted in the case of MASS.
At company level, the existing safety management procedure, as regulated in the SOLAS Chapter IX and the associated ISM Code, will no doubt also retain a crucial role in the management of safety for MASS. The current regulatory system appears well-suited for MASS, but the manner in which the procedures operate, e.g. on the communication between ship and shore, will have to undergo important alterations in the process.
The acceptance of MASS also requires considerations by relevant coastal states regarding whether a particular area is suitable for MASS operations, whether the safety of other ships and the environment is ensured, and regarding the availability of various supporting infrastructure etc. There is no precedent for handling such issues within IMO, but the 2019 Interim Guidelines for MASS Trials (MSC.1/Circ. 1604) represent a first step in this direction. The Guidelines may be interpreted as a cautious endorsement by the IMO of experiments with MASS in international areas, even if not all technical requirements are strictly met. In this sense, para. 2.2.1 of the Guidelines provides that “compliance with the intent of mandatory instruments should be ensured” (emphasis added). This endorsement, in turn, may have implications on whether MASS trials undertaken along the lines foreseen by the Guidelines (which are not limited in time or to a certain sea area) are perceived as being undertaken in accordance with “generally accepted rules and standards”, as required by the law of the sea. As far as watchkeeping is concerned, the Guidelines may have a more direct regulatory impact if they were considered to enable (constrained) exemptions from the watchkeeping rules under Regulation I/13 of the STCW Convention, which refers to trials conducted “in accordance with guidelines adopted by the (IMO)”.
Overall, these examples highlight that, even if there is a long way to go before regulation of MASS is anywhere near robust, the existing regulatory regime offers a variety of possibilities for individual states to support the development of MASS without having to await a full review of the IMO rules, and without being in conflict with the existing international rules.
Clarity on liability rules is desirable, but not a conditio sine qua non for operating MASS. This is different with respect to safety requirements, where the legality of operating MASS needs to be confirmed beforehand. In both areas, states may already prepare for MASS and have a number of mechanisms and instruments at their disposal to accommodate MASS into their legal framework, and to start specifying the key requirements that will be needed for such ships. In view of this, it is surprising that the regulatory debate on MASS to date has been so ‘backward-looking’ focusing on the impact on existing rules. The bigger challenge is arguably to agree on what rules and standards are missing, and how the gaps should be filled, in the shorter and longer term, nationally and internationally.
Professor Henrik Ringbom is a Professor II at the Scandinavian Institute of Maritime Law in Oslo, Norway, and Head of Research at the Department of Law, Åbo Akademi University in Turku/Åbo, Finland. He recently co-edited the book Autonomous Ships and the Law (Routledge, 2020).