A leading zero emissions and autonomous ferry project in Norway presents DNV GL with a golden opportunity to research how to shape the future of digital assurance.
Driverless cars are currently the main focus of public attention and debate on critical issues of safety and trust. But it turns out that a more viable and immediate case can be made for autonomous ferries. They operate in a more controlled and simpler operational design domain than cars, and as such provide a unique opportunity for learning about the broader context of autonomy. Impressive technology for autonomy already exists, but it cannot be deployed until the risk of significant failure is proven to be acceptably low. That, however, is only the starting point: public acceptance is also key.
The case for autonomous, zero emission ferries
From a user perspective, the case for autonomous, zero-emission ferries is very appealing. Such vessels could solve both the transport and environmental needs of cities and local communities, while helping to decongest heavy traffic in urban centres. New housing areas and sustainable infrastructure can be opened without costly and ecologically impactful roads and bridges. By connecting people to services while facilitating sustainable living and enhancing wellbeing, autonomous ferries provide win-win solutions for mitigating climate change and adapting to rising sea levels.
There are notable advances in ferry transport around the world, for example, the Water-Go-Round hydrogen fuel cell ferry prototype in San Francisco. But it is in the Nordics, and Norway in particular, that the idea of autonomous passenger ferries is being advanced. Ferries are a vital part of Norway’s highway infrastructure, and link communities across fjords and between islands, conveying over 6 million passengers in 2019. Major advances in electrification of ferries are being made, utilizing the country’s green hydroelectricity, and Norway possesses a cutting-edge cluster of ship design, construction and digital technologies needed for autonomy. Substantial investment capital is also available for the development of electric and autonomous ferries, including the Zeabuz (see below).
The major hurdle…
What is lacking, however, is the trust that these autonomous and zero emissions ferries can operate safely and reliably. They will also need to gain social acceptance as well as recognition for their potentially substantial contribution to the Sustainable Development Goals.
Building trust in autonomy
The safety and security of autonomous vehicles is not simply a matter of avoiding objects in their trajectory. Rather, they must be seen as elements in a wider context, interacting with the natural environment as well as a part of society.
How can we build trust in complex and intelligent systems that enable autonomous mobility? This is the key research question behind DNV GL’s research project, “TRUSST” – Assuring Trustworthy, Safe and Sustainable Transport for All.
The primary objective of TRUSST is to innovate an integrated assurance framework that takes as point of departure the insight that autonomous transport systems are formed by a complex and interdependent system of people, technology, organisations, regulators and the natural environment. The safety and security of autonomous vehicles is not simply a matter of avoiding objects in their trajectory. Rather, they must be seen as elements in a wider context, interacting with the natural environment as well as a part of society. The project’s integrated assurance framework aims to transform this complex system into a trust ecosystem.
At heart, an autonomous vehicle is a complex AI-governed cyber physical system. Traditional assurance methods and tools are not well-equipped to build trust in assets that are embedded with multiple digital technologies, and where AI gives the system a new kind of agency and responsibility. Moreover, since these systems will be changing throughout their life cycle, they will require effective means to deliver continuous assurance. This triggers the question of what can be considered “safe enough”, and the associated challenges of so-called corner cases or the long tail of the possible scenarios that may occur. Add to this some unknown unknowns and the complexities of human-machine interplay, and we have a rather complex assurance case to solve.
The assurance approach must also attend to ethical dilemmas and engage with citizens to provide societal trust and ensure the actual deployment of these new technologies are not only responsive to societal needs, but also leads to enhancing people’s lives and contribute to a more sustainable future.
Our hypothesis is that such an assurance process must take a holistic view, one that enables the framing and analysis of the resulting ecosystem of technology, people and their social and natural environment from a systems perspective. We cannot solve the deployment of digital technologies in society by looking only into the technologies – although that is, of course, critical.
This point is a central concern for Erik Dykoren, CEO of Zeabuz: “My prediction is that building trust in our electric, autonomous ferry system is going to be just as challenging as designing and constructing the system itself. And, consequently, will require a similar amount of resources to succeed. We need to prove the system trustworthiness towards a wide range of stakeholders, including our passengers, the general public, marine traffic, regulators, insurance companies, municipalities, transport system operators and owners, and finally our suppliers and partners.”
Providing trust to society
Our proposed approach places a central focus on the stakeholders involved and their diverse interests and concerns. We build from our existing expertise in the qualification of new technologies and give a central role to maturing an in-depth assurance case, which is an efficient strategy based on argumentation and logic in cases when governance and industry standards are not yet fully matured.
Assurance serves, ultimately, to provide trust to society, including attention to ethical dilemmas, engaging with citizens in democratic deliberation, and in co-design with other key social actors, to envision optimal ways of deploying autonomous technologies in society. The TRUSST project aims to identify the worries and concerns of citizens regarding autonomous transport systems. For example, are people worried about the loss of jobs for seafarers and shore workers? Or do citizens rather see the potential for new employment opportunities in the new services these integrated, zero emissions and autonomous transport systems will require (for example, a remote-control centre) and increased quality of life?
Zeabuz as a reference case
The TRUSST project uses the pilot ferry milliAmpere 2 (mA2) owned by NTNU and the Zeabuz mobility system as a reference case for iterative development and application of a new Assurance of Digital Assets framework. As shown by the dotted line in the infographic, the mobility system includes not only the autonomous ferries, but also the docking/charging stations, the communication network, the remote support centre, and the integrated digital twin. The project looks also at the stakeholders and their requirements outside the dotted line, including people, the natural environment and wildlife.
In sum, the project innovates digital assurance concepts that integrate processes, tools, simulation-based testing, use of digital twins, and full-scale testing. This innovation is complemented with the use of citizen engagement methodologies and environmental impact assessment to transparently provide trustworthiness to all stakeholders, and in turn unlock the potential for scaling up autonomous and sustainable transport solutions.
“I do not doubt that autonomous, zero emission ferries will eventually enter service, and that they may solve many logistical and environmental challenges,” says Remi Eriksen, President & CEO of DNV GL. “The question is, how soon can we get to these benefits safely? The assurance challenge requires not only lowering technical risk to acceptably low levels but ensuring that the technology is welcomed by society. That is why we are working on shaping a new approach to assurance that is founded on building a trust ecosystem.”
The Assuring Trustworthy, Safe and Sustainable Transport for All (TRUSST) project is an industrial research project led by the DNV GL Digital Assurance Research team, in partnership with Zeabuz, a spin off start up emerging from R&D activities on autonomous ships at the Norwegian University of Science and Technology (NTNU), and Marine Technologies as industry partners and NTNU as research partner, and will run through Q2 2023. It is partially funded by the Research Council of Norway under grant number 313921. The project also has a reference group consisting of the Norwegian Maritime Authorities, Torghatten and Trondheim Kommune. Sub-suppliers Missions Publiques and Terravera will assist with citizen engagement and assessment of impacts on the UN Sustainable Development Goals, respectively.