Challenges to the adoption of smart autonomy

Currently, there are several potential blockers that need to be considered when rolling out autonomy solutions, including technical, organisational and regulatory. These key obstacles can be summarised as follows.

An evolving regulatory environment

This is perhaps the biggest obstacle to rolling out fully autonomous operations. The IMO is looking into a regulatory framework for autonomous shipping and is currently undergoing a scoping exercise. In the meantime, the IMO’s Maritime Safety Committee (MSC) has released its Interim Guidelines for Maritime Autonomous Surface Ships (MASS). These guidelines have been used to guide recent trials such as Wärtsilä and PSA Marine’s IntelliTug trial conducted in collaboration with the Maritime and Port Authority of Singapore. In practice, rolling out autonomous solutions means operating in a fragmented and changing environment based on regulations set out by flag states and local authorities.

Regulations that impact remotely operated and unmanned vessels

  • The International Convention on Standards of Training, Certification, and Watchkeeping for Seafarers (STCW Convention)
  • The United Nations Convention on the Law of the Sea (UNCLOS)
  • The International Convention for the Safety of Life at Sea (SOLAS Convention)
  • The Convention on International Regulations for the Prevention of Collisions at Sea (COLREG Convention)

Human and organisational inertia

Whenever rolling out a new solution, it is critical to ensure the people who are supposed to use it not only understand it and are properly trained, but also see its benefits. Such changes are not always easy as they require adjustments to ways of working, challenge old methods and routines and also raise fears about job security. Added to this is the fact that the shipping industry has relatively long innovation cycles, due partly to its operating nature, but also linked to expensive assets and asset lifetime. To Wärtsilä Voyage, autonomous technologies are foremost about helping existing crew operate more efficiently and safely. Autonomous operations should not be just equated with unmanned vessels, but seen as technologies to help operate vessels better, today.

Moving from more traditional operations to the smart autonomy operations model requires a shift in mindset and the adoption of new processes and competencies across the entire organisation. Rolling out new solutions requires bespoke training in order to ensure buy-in from all stakeholders. Commitment from leadership is key—it will be up to management to help set out the vision and guide the process of both choosing a solution and evolving roles, processes and partnerships to ensure successful uptake of autonomous solutions.

Technical challenges and lack of interoperability and system integrators

While massive improvements have been made in machine learning and AI, these technologies are still developing and will need to develop further to ensure that fully autonomous operations are both realistic and safe. Development of integration between sensors into a single system needs to continue to ensure there are no blind spots and that small-object detection is reliable. One of the challenges in this area is that some of the more mature sensors (such as GNSS and AIS) have well established data standards. Other sensor types are still maturing and there are new areas of innovation (such as LIDAR and video) where the data standards are not as established—this will mean customisation is required. Furthermore, it’s worth noting that most of the available current sensors are not built to support autonomy; their data frequency is fine for situtional awareness but it isn’t high enough for the real-time decision-making needed by an autonomous system.


Multiple players make single systems that do one thing (and may do it very well) but expanding from there is difficult if not impossible. As these solutions are not mutually compatible it leads to multiple data silos as more onboard systems are added, increasing rather than reducing complexity. The solutions are often not linked to real‑life operations or a specific business goal, which makes it hard to see any concrete benefits and hence define a compelling business case for smart autonomy.

One of the ways to improve systems is through simulation before deployment to test that they function as expected. These systems can then be further improved through gathering real-life operational data that filters back to continuously improve the system with the help of machine learning.

Cyber security concerns

As autonomous solutions rely on digitalisation, integration and automation, marine cyber security risk management is a growing and justified concern. There is also a regulatory aspect to this. As the IMO has recognised that cyber security is becoming critical for data protection and reliable and safe marine operations, from January 2021 it requires that cyber security be addressed in safety management systems.

In 2017, the IMO adopted resolution MSC.428(98) on Maritime Cyber Risk Management in Safety Management Systems (SMS). Any digital systems should have a robust cyber security framework based on best practices and guidelines to ensure the security of operations.

Want to learn more?

Book a discovery call with our experts