Executive summary
The PLATINA3 project
The Horizon 2020 PLATINA3 project provides a platform for the implementation of the NAIADES III Action Plan. PLATINA3 is structured around four fields (Market, Fleet, Jobs & Skills, Infrastructure) of which Work Package (WP) 2 deals with various aspects of the inland navigation fleet, such as 1) zero-emission strategy; 2) climate-resilient vessels; 3) automated vessels; 4) fleet data; and 5) funding the energy transition for the fleet; 6) energy label index for vessels; 7) regulations for zero-emission vessels.
This report presents the conclusions from PLATINA3’s Task 2.3 which aims to facilitate the development of regulatory frameworks at European level for onboard systems allowing automation of inland navigation vessels. This deliverable is based on desk research building upon existing studies and analyses and is further substantiated by expert interviews. The outcomes of the 5th PLATINA3 Stage Event (19-20 October 2022), where experts made presentations on this topic, a draft deliverable was showcased, and an interactive workshop was held, are also integrated into this report.
Scope of the report, core definitions, and methodology
The scope of this report is limited to systems allowing automation as well as remote-control of navigational tasks and the corresponding European regulatory frameworks for vessels. Issues related to professional training and qualifications, police requirements, economic and market-related implications, infrastructure, liability and insurance, dangerous goods, and fuel, emissions, and sustainability aspects, are outside the scope of this report.
The core definitions used in this report are based upon the CCNR’s levels of automation, the first international definitions of automation tailor-made for the IWT sector. The levels of automation range from steering assistance and partial automation (levels 1-2) to progressive delegation of tasks without intervention of the boatmaster (levels 3-4). Fully autonomous vessels correspond to level 5 (independent command with no human involvement), the most advanced stage of automation.
The report distinguishes whether the vessel is remotely controlled or not, and the degree to which it is automated, i.e. which tasks are completely delegated to the computer, which remain in the human domain, and which are handled by both. Although automation and remote-control are not completely independent from each other, they are functionally different concepts.
This report employed a step-by-step approach based upon an incremental and primarily inductive research design. An analysis of European pilot projects was carried out to determine the TRL and evaluate the RD&I needs of the various systems allowing automation of inland navigation vessels. The report then identified the main functions allowing automation of navigation-related tasks as well as their associated safety concerns. Considering these functions, a gap analysis of relevant European legislation was carried out to identify the possible regulatory barriers/gaps and thereby propose new requirements via a technologically neutral approach. Ultimately, the outcome of this analysis was translated into recommendations accompanied by a Roadmap.
Current state of play and policy context
On 17 October 2018, the inland navigation Ministers of the five CCNR Member States (Belgium, France, Germany, the Netherlands, Switzerland) adopted the Mannheim Declaration and called, inter alia, for the “development of digitalisation, automation and other modern technologies in order to contribute to the competitiveness, safety and sustainable development of inland navigation”. In 2022, the CCNR published a vision to support the harmonised development of automated navigation via a holistic and technologically neutral approach.
In response to the Paris Agreement, the European Commission adopted the European Green Deal (EGD) in December 2019, which aims to shift a substantial portion of the freight transported by road (currently accounting for circa 76% of EU inland freight) to inland navigation (circa 6%) and rail (circa 18%). As transport accounts for a quarter of the EU’s greenhouse gas emissions and is growing, achieving climate neutrality implies that a 90% reduction in transport emissions is required by 2050. All transport modes must contribute to make climate-neutral, resilient, and intelligent synchro-modal automated transport by 2050 a reality. Through this, the EU will unleash the full potential of data, integrate electronic ticketing facilities for seamless multimodal transport, and deploy automated mobility.
As the transport arm of the EGD, the Sustainable and Smart Mobility Strategy (SSMS) lays the foundation for how the EU transport system can achieve its green and digital transformation ambitions and become more resilient to future crises. It underlines the need to increase the use of more sustainable transport modes and indicates that IWT and short-sea shipping should each increase by 25% by 2030 and by 50% by 2050. The SSMS envisions that automated mobility will be deployed at large scale by 2030 to increase the efficiency and reliability of transport, logistics and supply chains. Automation is also identified as a driver of smart mobility in achieving seamless, safe, and efficient connectivity.
In June 2021, the EC launched the NAIADES-III initiative, which sets a 35-point “Inland Navigation Action Plan 2021-2027” aligned with the Multi-Annual Financial Framework to meet the objectives of the EGD and SSMS.Its two core objectives are to shift more cargo to Europe’s rivers and canals and facilitate the transition to zero-emission barges by 2050 to boost the role of IWT in environmentally sustainable mobility and logistics systems. One of the eight NAIADES-III policy flagships aims to support the development, demonstration, and deployment of holistic, smart, and automated shipping concepts with a focus on the most promising applications in terms of feasibility and commercialisation, as well as in terms of environmental benefits.
Analysis of European pilot and research projects
The analysis of pilot projects revealed that conducting pilot tests allows regulators and innovators alike to gather critical knowledge, data, and real-life experience to adapt the relevant regulations to achieve automation. Most of the systems needed for automated vessels are already available but some technologies must be further developed and tested before becoming fully operational. Furthermore, questions related to the interoperability between software and onboard systems remain unanswered, both onboard the automated vessel (human-machine interface) and in relation with other vessels (communications, signalling, intent sharing) in a mixed navigation environment.
There is a need for both overall pilot projects on long stretches or the entirety of a given waterway to test the operational feasibility of automated navigation, as well as very localised projects to test specific operations, such as entering locks, passing infrastructure and chokepoints, or making challenging turns. To avoid fragmentation, there is a need to develop industry-wide standards or guidelines. The IWT transport sector would benefit from standardization activities at European level for the scaling-up of automation. Where possible, there would be a clear added value to have similar standards across the waterborne sector.
Systems and functions allowing automation of inland navigation vessels
Six main functions allowing the automation of inland navigation vessels were identified. These are situational awareness, collision avoidance, communications, navigation control, safety, and fall-back capability. Each of these main functions is further broken down into sub-functions, including several ‘reliability guarantees’. Each sub-function is underpinned by several system families in varying combinations, such as sensors (RADAR, LIDAR, cameras), positioning systems (GNSS, inland ECDIS), communication systems (4/5G internet, AIS, VHF), computer components (IMU/GPU, centralized PLC system, AI/machine learning), and more. For each subfunction, minimal requirements are identified, both in terms of technical regulations, safety prescriptions, systems’ interactions, and human involvement. Some proposed solutions to address common safety concerns are provided to guide the work of regulators.
In terms of Technological Readiness Levels (TRL) and outstanding RD&I needs, it appears that most of the systems needed for low level automated navigation (levels 1-2) are already in a relatively high state of market readiness. This includes the core systems allowing automation (RADAR, LIDAR, cameras, GNSS, communications, global internet, track pilots etc.), which are considered to have reached a high TRL level. On the other hand, techniques and systems for high automation and autonomy (levels 3-5) have comparatively low TRL levels. Indeed, the most advanced systems (collision avoidance, AI, neural networks, sensor fusion and integration, etc.) still need additional technical improvements to move from TRL 5-6 to TRL 9. Furthermore, on some small sections of the Rhine and on most of the Danube, high speed internet connectivity (4G/5G) remains unavailable, which is a virtual precondition for operating a significant share of automated vessels, especially remote-controlled vessels. Finally, encryption, data integrity, and cybersecurity systems and protocols still need additional testing and improvements to become fully mature. This remains critical for the safe deployment of remote-controlled vessels and other higher automation applications.
Some systems allowing high levels of automation are currently in use, although there is always a human as a supervisor and backup – either onboard or in an RCC. More testing locations for automation levels 3 and above are needed to gather as much data as possible. This data is necessary for developers to improve the performance of their systems and for regulators to make informed decisions.
Results of the regulatory gap analysis and recommendations
The identified regulatory obstacles to the uptake of automated inland navigation vessels in ES-TRIN fall into two main categories. The first category regards provisions that constitute regulatory barriers and therefore do not allow or contradict the aims of automation. These typically refer, explicitly or implicitly, to the presence of a boatmaster and/or crew members onboard, either to perform an action or to interact with equipment designed for manned operations (e.g. doors to be passed, signs to be read, etc.). These provisions should, broadly speaking, be amended to account for the specificities of automated inland navigation vessels. The second category regards the absence of regulations pertaining to specific functions identified as necessary for the safe automation of inland navigation vessels – i.e. regulatory gaps. This absence could generate a legal vacuum leading to a proliferation of patchwork solutions and possible low safety standards. At the very least, these functions should be incorporated into the regulatory framework.
Regulators (EU, CCNR, CESNI), standardisation bodies (CEN, ETSI) and classification societies should work together to fund, support, and learn from pilot projects to gather the necessary data and experience to better regulate the IWT policy area to allow the automation of inland navigation vessels. Regulatory work should be complemented by targeted interventions to bridge the financial gap and the respond to the outstanding RD&I needs for automation-enabling systems to reach technological maturity in the short to medium term.