International shipping has the potential to undergo an evolution with developments in autonomy—these developments present opportunities to both increase safety and reduce risk to vessel operations. Zulu Associates, a Belgian company which describes itself as an innovator in marine logistics and focusses on smaller vessels, expects to put small autonomous container ships into the English Channel or Southern North Sea by 2026. In an interview with TradeWinds, the CEO of Zulu Associates, Antoon Van Coillie, indicated that shipping insurance markets are cognizant of autonomous systems and ships. He asserted that financing would not be an unsurmountable barrier, since financial institutions are especially interested in vessel sustainability.

In 2021, the global autonomous ships market had a revenue share of over 89 million USD, and is projected to grow at a compound annual growth rate of 6.81% through 20311.


According to the International Maritime Organization (“IMO”), a Maritime Autonomous Surface Ship (“MASS”) is defined as a ship which, to a varying degree, can operate independently of human interaction. This would include the use of technology to carry out various ship functions such as navigation, propulsion, steering, and control of machinery.

Autonomous shipping can support shipowners and charterers with economic, environmental and risk-related benefits. These benefits can include decreased voyage times, reduced fuel consumption, and reduced risk to crew. Voyage efficiency can benefit many aspects of the industry: in addition to saving expenses, it can help parties with binding and nonbinding efforts to reduce emissions and adhere to international standards on carbon reduction. Of course, shipowners and charterers will have to balance the potential benefits of autonomy with costs and risks, especially related to incorporating new technologies during manufacturing, reworking crew duties and concepts of liability.

Levels of automation

Fundamentally, autonomous devices learn from their surroundings and complete tasks without continued human input. This can include a simple automated task on a vessel, or a vessel that conducts an entire voyage without human interaction. This spectrum of autonomous functions can be characterized using different degrees or “levels” of automation (“LOA”). There are many similarities in the LOA used across maritime organization, which generally include four categories. The lowest level still involves human navigation and operation, and the highest level employs full (unmanned) autonomy. The details of these levels may differ across organizations such as the IMO, the American Bureau of Shipping, Det Norske Veritas (DNV) and Lloyd’s. The IMO, for example, used the following classification in its regulatory scoping work:

  • Degree one [lowest level]: Ship with automated processes and decision support. Seafarers are on board to operate and control shipboard systems and functions. Some operations may be automated and at times be unsupervised but with seafarers on board ready to take control.
  • Degree two: Remotely controlled ship with seafarers on board. The ship is controlled and operated from another location. Seafarers are available on board to take control and to operate the shipboard systems and functions.
  • Degree three: Remotely controlled ship without seafarers on board: The ship is controlled and operated from another location. There are no seafarers on board.
  • Degree four [highest level]: Fully autonomous ship: The operating system of the ship is able to make decisions and determine actions by itself.

Regulation of MASS Ships – developments in the pipeline

The successful proliferation of MASS requires maritime safety levels commensurate or higher than conventional ships.  This will require the adoption of regulatory standards across the industry; an issue rife with complicating factors like rapidly advancing technology and existing maritime conventions.  To address these challenges, standards-setting organizations are working to address functional requirements for MASS operations and reviewing existing maritime conventions for compatibility. 

On the international stage, the IMO is conducting a regulatory exercise that includes the development of a MASS code.  The goal of the IMO is to adopt a non-mandatory goal-based MASS Code to take effect in 2025.  This will form the basis for a mandatory goal-based MASS code, which IMO expects to enter into force in 2028. 

The technical components of a final MASS code may contain requirements related to:

  • Function—Addresses what the automated components of the vessel do;
  • Performance—Addresses how well the automation functions accomplish tasks, including response speed and execution time;
  • Interface – Addresses the link between hardware, software and other components of the automated functions;
  • Communications –Addresses the automated elements’ communication, including with remote crew and other vessels;
  • Safety;
  • Cyber-security.

The IMO working group on the MASS code is expected to report at the 108th session of the organization’s Maritime Safety Committee in May 2024.

As noted above, standards-setting organizations will also have to address the effect of any new regulations relating to MASS. As part of IMO’s regulatory work, it has analyzed the potential effect of a MASS code on existing maritime conventions such as SOLAS, STWC, and COLREGs. Among other findings, IMO amendments to existing conventions may need to occur, including the role and responsibility of the master, liability questions, and causation issues.

While mandatory MASS standards are still under development, the IMO approved in the interim nonbinding guidelines for MASS trials development. Similarly, the EU has also developed guidelines for its member states. During these trial phases, most autonomy testing has occurred in inland waterways. There, the vessel is closer to shore for potential intervention, and individual nations have more autonomy relating to navigational rules.

Current status

While the development of a MASS code will enable advanced automation technologies, it is worth noting that automation is not an entirely new concept in shipping. In fact, automation is already used to some extent on traditional cargo ships. For example, in an effort to conserve fuel, shore-based ship routing services have used satellite and vessel-based readings for decades. And, as it relates to navigation, some vessels already have autonomous navigation-related systems on board. For example, the vessel PYXIS OCEAN set sail in August 2023 fitted with a “Windwings” system of sails, which are set and trimmed automatically based on camera, sensor and weather data, rather than by the vessel’s crew2.

Beyond vessel routing and navigation, additional automation technologies are being tested and deployed. The Yara Birkeland is an electric autonomous 80 m 120 TEU Norwegian-flagged containership launched in Spring 2022. It carries out shuttle voyages between Heroya and Eindangerfjorden in Norway, and has been tested carrying out container loading and unloading operations autonomously3.

New legal frontiers

  • (a) Responsibility and liability

Despite the fact that automation is currently used in certain elements of voyage, fully autonomous vessels will demand new legal analyses in contracting and litigation. For example, the concept of autonomy in shipping brings about questions relating to responsibility and liability. Consider the notion of “seaworthiness”, which is generally met when a vessel is properly constructed, prepared, manned, and equipped for the voyage intended. In the case of autonomous vessels, the query turns to how a shipowner can guarantee that the vessel is properly manned for the voyage intended. One consideration is whether responsibility for seaworthiness would then shift to the manufacturer of the automated functions or vessel.

The concept of “seaworthiness” is one of many guarantees that is regularly incorporated into contracts for the use of ships and cargo of carriage—which is when questions of responsibility can morph into issues of liability. Sticking with the example of seaworthiness, a shipowner may be liable for resultant damage if he fails to use due diligence to make the ship seaworthy before and at the beginning of the voyage4. This is true even if there has been an intervening nautical “fault” on board. The legal issues of responsibility of the shipowner, and subsequent applicable liability for vessel seaworthiness, will require significant legal inquiry.

  • (b) Delegating responsibility for seaworthiness

The concept of seaworthiness is also relevant to marine insurance coverage: a shipowner typically has no cover in connection with a knowingly unseaworthy vessel. However, the option to delegate seaworthiness to another entity (such as a contract worker or presumably, a remote operator) is limited.

As far back as the 1961 case of The Muncaster Castle in 1961, the English courts have found the shipowner responsible for a failure to exercise due diligence, even when another actor was responsible for faulty work rendering a ship unseaworthy. In that case, failure by a shipyard worker to tighten screws before a voyage caused cargo to be wet by seawater. In Munster, this was nevertheless held to be a failure by the shipowners to exercise due diligence to make the vessel seaworthy, notwithstanding that the failure was caused by a contracted shipyard worker.

But there are exceptions to the limitation on a shipowner’s ability to delegate seaworthiness. These exceptions are illustrative when analyzing concepts of liability in the autonomous context. Under one such exception, a shipowner or carrier might not be responsible for faults made by independent contractors before the ship came into their possession or “orbit” – involving, e.g., faults made by the shipbuilding yard during the course of building the ship, or faults during a previous owner’s time of ownership5. Perhaps a shipowner or carrier could claim that a defect in the vessel’s autonomous systems came about before the vessel came into their “orbit”—and therefore, they are not responsible for the lack of seaworthiness.

Another option to limit liability would be for shipowners to seek to categorize errors by software or remote operators as “navigational faults,” for which they should not be responsible. The possible removal of “human error” on board ship, could be replaced by factors such as an autonomous system failing to “understand” or properly categorize what its sensors have detected. That is, the relevant “fault” may lie in the underlying software, errors in data processing by that software, a mechanical breakdown to the sensors, or possibly a mistake by a remote human operator.


As noted above, various international bodies working to set forth binding guidelines relating to MASS vessels. The proliferation of agreed-upon standards will enhance the legal analyses of challenges resulting in the autonomous shipping context. Nevertheless, parties to a dispute in this area will require bespoke expert technical help from both software and maritime experts.


[1] Straits Research,



[4] Art.III r.1 of the Hague-Visby Rules.

[5] See The Muncaster Castle