The building blocks of smart autonomy

In Wärtsilä Voyage’s view there are three key areas that need to be considered to enable autonomous solutions

Very often these build on each other as a solution combines elements of all three areas, but solutions can also exist in just one of these areas, which we will look at later in this paper. To begin, let’s define what each of these areas entails in practice.

Situational awareness

Decision making and logic

Action and control

Situational awareness—what’s going on around the vessel?

As an enabler of autonomous operations, situational awareness refers, among many other things, to building a digital understanding of the vessel’s surroundings and own operating state. Situational awareness around the vessel is important for vessels of all sizes and is achieved by combining data from several different sources to digitise the environment and provide an accurate picture in order to detect and recognise hazards. These sources include radar, laser, cameras and environmental sensors (GNSS, wind sensors, MRUs, Echosounder etc.) to build up a complete and reliable picture of the environment around the vessel. Ensuring a complete overview of what’s going on also requires integration between all these sensors. This kind of integration requires action in three areas.

The services

The hardware stack

including a server with enough computing power to process huge amounts of data and predict/plot in real time, as well as peripheral hardware, power management and a well-balanced sensor suite.


the fusion and treatment of data is needed as every sensor generates both high volumes of data and different interpretataions of the ship’s environment.


a process is needed to patch and update the various applications within the entire suite to maintain cyber security without impacting/breaking the entire system.

Based on that understanding, the crew or an autonomous system receives the inputs needed to decide on a safe and optimal course of action, which is handled by the decision-making and logic capability. In many ways modern technology can already be used to augment the detection and tracking capabilities of crew and in some ways greatly surpass human capabilities. Human eyes and ears still play a central role in a mariner’s hazard detection but the advent of night vision, laser, LIDAR and short-range radar technology makes information accessible under conditions where humans are literally blind. In addition to knowing what is happening around the vessel, onboard situational awareness is also important for autonomous operations. This can include the state of the plant, cargo, ongoing maintenance activities on board and power management, which all have an impact on navigational decision making.

Decision making and logic capability— what needs to be done in a given situation?

At the heart of autonomous systems are decision-making algorithms that can apply machine learning to interpret a scenario based on data provided by situational awareness solutions and decide a safe and effective course of action. At its simplest, decision-making and logic capability is the ability to understand a vessel’s situation and make good decisions. In certain situations, such systems can help relieve some of the cognitive load on the crew and optimise operations onboard.

Decision-making algorithms have become more and more sophisticated, leading to smarter systems that have increased autonomous capabilities. For example, the goal of a future smart navigation system is to take a vessel from port to port in a safe manner while avoiding collisions or other hazards and maintaining compliance with COLREGS. Such systems would take advantage of digital modelling of the vessel and its environment.

Because of the risks of testing on real vessels at sea, simulation is a critical part of developing autonomous solutions. The computer models that underpin simulationbased training—capable of replicating almost infinite permutations of marine environments, vessel traffic situations and ship equipment—are the same that can be used to inform the decision-making capabilities of intelligent systems. Deployed in real time with real people in simulators, those same models can be used to test and validate autonomous solutions. These simulators can also be used to train crew to understand and use new systems.

Autonomous solutions can help humans make better decisions based on existing information and also make consistent, accurate calculations based on given and programmed data. At the same time while working alongside humans, this provides great feedback opportunities to ensure such algorithms are sufficiently robust and reliable when it comes time to switch from human control to a more autonomous control system.

Action and control—how do we safely and efficiently make the vessel take action?

When we talk about action and control, we’re talking about how a vessel executes the decisions made by algorithms to precisely take actions onboard. For the purpose of this paper, the focus is mainly on solutions for navigation as they can offer immediate benefits to operators in today’s markets. Smart vessel control and drive systems are needed to safely travel from port to port and to handle such tasks as docking, harbour entry and remote pilotage.

For example, an autonomous navigation system can dynamically plan a route in real time that avoids collisions based on situational awareness and decision-making and logic. These instructions are then sent to the action and control systems to execute. If manual control is required, the crew can intervene at any point.

The benefit of action and control in autonomous solutions is that, in applications ranging from ferries to long‑distance voyages, the ability to repeatedly and consistently deliver the same course of action greatly increases levels of efficiency and operational safety and helps reduce crew fatigue. In the future, action and control solutions will be essential to fill in for humans once they have left the control loop.

How smart autonomy fits in to current autonomy classification schemes

At Wärtsilä Voyage, we see that situational awareness, decision making, and action and control are the enablers for autonomous solutions. It’s up to each customer to decide how far they want to move in the direction of fully autonomous operations. How these levels are defined, and how useful they actually are, is still being discussed— here are two prominent examples.

The IMO’s view on autonomous shipping is based on four levels

  • Vessel with automated processes and decision support for onboard crew
  • Remote controlled vessel with humans onboard
  • Fully autonomous vessel that can make and execute all needed decisions

It’s important to note that Wärtsilä Voyage is not proposing an alternative or replacement level of autonomy scale in this paper—instead, we are focusing on clarifying the enablers and the opportunities that already exist in this area. This ensures that Wärtsilä solutions relate to customer needs while also being mappable to relevant class levels and regulations in the future.

Want to learn more?

Book a discovery call with our experts