Unmanned Surface Vessels (USVs) are perhaps the most visible smart shipping applications to date in the Netherlands. Several developers of USVs and USV technology are commercially successful in the Netherlands, and various dredging and offshore instruction companies in the Netherlands are already starting to use USVs in their daily operation.
In the context of this roadmap, we define USVs as small vessels (<20 m for inland vessels, <24 m for seagoing vessels). Unmanned Surface Vessels are, as the name implies, not equipped for the carriage of people, so small autonomous ferries are not meant in this sense. Ferries have their own challenges and different market drivers, which are described in the inland ferry use case.
Although a lot of USV technology is developed and used in a military context, this roadmap focuses on civilian applications of USVs.
Functional requirements of navigation systems
Journey and route planning systems
Situational awareness and obstacle detection
Collision avoidance systems
Development of multi-vessel operational techniques for efficient use of USVs
Functional requirements of navigation systems
Determine functional requirements for safe smart shipping systems (offshore: situational awareness and visibility are of additional importance for this segment)
To establish safe autonomous navigation and vessel route guidance systems, the functional requirements and performance standards of these systems need to be established to an objective and consistent standard. The behaviour of these systems will be compared against the behaviour of human operators and against collision regulations, which are currently not framed in a suitable manner for use by machines.
The performance standards for equipment used for gaining situational awareness and situational understanding are currently unavailable. It is currently not known how good is “good enough”, and quantified safety standards for navigational safety are lacking. Development and testing of such standards is essential input for the development of new regulatory instruments, both in IMO and for national legislation.
Since requirements for the robustness of autonomous navigation functionality also depends on the availability of reliable communications to revert to shore control, as well as varying techniques of dealing with system failures makes a flexible yet robust set of requirements challenging to formulate.
Development of these standards will require a combination of testing, analysis and (simulated) behavioural studies of autonomous navigation systems and standards development.
Visibility of the USV
Especially in offshore operations, visibility is of additional importance. But also for inland waters the detection of small USVs by other vessels can be challenging. No uniform guidelines for how this visibility is to be improved are available.
Journey and route planning systems
While various journey planning and route planning systems are being trialled, these need to be developed and tested against functional and performance requirements as a prerequisite to regulatory and market acceptance.
Development of these systems will have to be done in parallel with the development of these requirements in an iterative manner. This will require co-operation between the developers of these software and hardware systems and (independent) knowledge institutions.
Situational awareness and obstacle detection
Autonomous navigation systems rely on new sensors to build up a situational awareness of the environment around them. These sensors may be on the ship, on other ships or even on shore. Development of these sensors is ongoing, but their performance needs to be measured against standards which are currently not available for their purpose as supporting autonomous navigation. Iterative development of performance standards and validation of actual performance is likely to be required through extensive development programmes.
Furthermore, when sensors are not installed on the ship itself, but as part of the infrastructure or on other ships, the data from these sensors needs to be uniformly interpretable by the navigation systems. Common data standards for these sensors are currently lacking.
Collision avoidance systems
Machine-learning tech nology which is used in collision avoidance systems are under development with various companies and being tested in real environments, but it has not been determined how the performance of these systems is to be evaluated or proven.
To ensure market and regulatory acceptance, the functioning of these systems needs to be validated in novel ways, and this cannot be done without the sensors from which the system builds up its situational awareness.
Development of multi-vessel operational techniques for efficient use of USVs
Many operational profiles rely on the use of multiple USVs in “swarm” operation for efficient use of the assets. Control and supervision of multiple USVs in close proximity poses different challenges and requirements compared to single shi p navigation.