Opening the discussion, moderator Edwin Lampert, Executive Editor & Head of Business Relations, Riviera Maritime Media, asked ABB Marine & Ports Head of Regulatory Affairs and resident autonomous expert Eero Lehtovaara to elaborate on the position outlined in their recent Marine & Ports White Paper on autonomous shipping.
“We are at a phase where we need a shared definition of autonomy,” Lehtovaara stated. “If we do this individually, we will end up with at least ten different answers.” The discussion going forward will need to focus on the level of automation, he said: “In principle it has nothing to do with crewing, or where you make decisions. Any discussion should refer to the level of human attention or attendance.”
Autonomy in the maritime industry is being enabled by digitalization, he continued. “More data is being used to make coherent decisions onboard, but coexistence of machines and humans onboard is the foreseeable future for shipping. Humans in the loop will be a part of the solution. I have never heard any shipowner wanting to get rid of people. We are talking about rearranging working conditions on board.”
The rise of technology should also provide an opportunity to get young people back on board, he added. “The ships of the future will become hugely interesting and exciting as workplaces.” But progress of this magnitude will require a dialog between original equipment manufacturers (OEMs) and regulators, he maintained: “We risk an unsafe future if these two are not on the same page. This is something ABB has been promoting actively for quite some time.”
Next up, Kenneth Lim, Chief of Technology at the Maritime and Port Authority of Singapore (MPA), related that Singapore is working to support the entire autonomous ecosystem. “We are looking at a network of ports using multiple port authority resources to provide support on port rules and requirements for safety, and criteria for operations. When vessels go from port to port, we need to start working at the port authority level.”
Singapore is funding local pilot projects to this end, he noted. “We currently have three tugs, one Ro-Ro and one other vessel in the program. We are exploring autonomy on levels that cover more than just the technology. To start with, the crew also need to become familiar with the technology and a new way of operating.”
To this end, Lim reported that Singapore is banking on simulation to familiarize crew with their new reality. “The human element is very important. We are not working to replace crew, but to assist them. We want to ease their burden. They are working long hours, and autonomous technology can help to reduce fatigue.”
Getting authorities and suppliers working together is another essential factor in reaching the goal of autonomous operations, he maintained. “We want to ensure that information on autonomous trials is being shared between players. ABB’s contribution on autonomy is important for us in this respect. Our own efforts will be focused on promoting autonomous operations in the R&D space.”
Rens Groot, Innovation and Performance Manager at Berge Bulk, noted that many owners are running lean organizations, relying on a network of suppliers to meet their needs: “Not long ago, the crew could repair physical equipment onboard. Now with advanced electric solutions, they need remote support. We have seen this even on engines that have experienced blackout. We cannot always move service engineers to the ships, especially with Covid-19, so we have to learn how to control operations remotely.”
A cyber secure environment that allows safe access to key equipment is key to reaching full autonomy, he stated. “It is a real challenge to get all the equipment safely connected in a central data system, when each piece is essentially stand-alone. Right now we do not have an overview of the entire ship ecosystem. We are not 100 percent sure how to proceed on systems integration.”
In a digital world, who owns the data?
Mahesh Krishnappa, Singapore Local Business Line Manager at ABB Marine & Ports, got the first call to reply: “Normally the customer owns the data,” he affirmed. “We can use it, we can process it, but they own it.”
Lehtovaara elaborated: “Data ownership is just one part of the equation. Before we can see the ship as a system, we must resolve the issue of data sharing between components and use of data.” For that to happen, he claimed, the industry needs to define data: “We are dealing with both operational and proprietary data, and not everything is shareable. But if we do not make it possible to see ship as system, the ship could become unsafe by design. We could end up with no insight or influence over equipment, and even less possibility of making relevant changes if needed.”
Imran Ibrahim, Principal Consultant, Head of Research and Development at DNV, agreed with Lehtovaara that a definition of autonomy is needed in order to support different concepts of operations: “Which functions do owners want to see as autonomous, and at which levels? The key is to define these functions.”
Interoperability, he explained, is behind DNV’s work with standards for the contextualization of data. “Class is showing how data can be contextualized in a uniform fashion. There are platforms that allow data sharing across multiple segments and between equipment. These are the stepping stones to data sharing.”
Brijesh Tewari, Marine Client Manager at Lloyd’s Register, concurred that Lloyd’s sees digitalization as the key to autonomy. “Let humans do what they do best, and design machines to do the routine tasks. This will reduce fatigue and add to efficiency.”
Looking ahead, he noted that the next generation is ready to step up to a higher level of digitalization: “Eventually everything in the industry will have to be moved up to this level.” He added that while the International Maritime Organization (IMO) is actively looking into autonomy, the ongoing coronavirus pandemic is demanding much of their attention: “We will all need to give more attention to autonomy when things return to normal.”
The critical elements of autonomy
Sanjay Kuttan, Executive Director at the Singapore Maritime Institute (SMI), named collision and foundering as the key issues to be resolved within autonomy. “Our main focus is considering the impact of autonomy on the Safety of Life at Sea (SOLAS), from design and construction, to maneuvering and navigation. We are using cyber physical modelling to improve situational awareness and the speed of decision-making, focusing on the three main factors influencing these: current, wave and wind.”
Triangulating these elements is critical to achieving greater situational awareness and decision-making, he explained: “Advanced cyber physical modelling and algorithms allow us to check the robustness of calculations by challenging them with potential real life situations. With climate change presenting us with new weather situations, how will an algorithm adjust to a scenario it has never been exposed to?”
At present SMI is developing a methodology to assist in piloting of autonomous vessels in Singapore, he said. “We are working on a pre-trial of the algorithm, how it will respond to wave conditions. We are also aware of similar projects in the EU. This will allow us to compare with EU results and test the effectiveness of our system, in order to get as close to the truth as possible.”
Anis Hussain, General Manager at KMDTech, noted that future autonomous systems will be made up of different sub-systems, including the bridge, the engine room, and others. “But an autonomous ship is not necessarily unmanned. The ship will not sail alone, but be guided by a control room on shore,” he maintained. “This will make workplaces more attractive as well. Smaller vessels are not crewed by local resources today. But if they can go to work in a control room rather than onboard the ship, this will attract a new generation of maritime workers.”
He noted that KMDTech is using digital twinning to learn more about collision avoidance. “Model test basins are also being used for testing of local ship designs. This allows for a proliferation of situational awareness. Using cyber physical modelling, complex surroundings can be controlled and tested.”
Alf Kåre Ådnanes, AMEA Regional Division Manager, ABB Marine & Ports, concurred on the complex nature of a complete autonomous system: “When we consider electric and autonomous systems in traffic, autonomous is not just what happens on the bridge, but also with propulsion systems, the engine room, and electrical systems. There is not really enough focus on this in the industry.”
He noted that autonomy involves more than just navigation, particularly in heavy traffic: “We need highly reliable or reconfigurable systems that are both robust and redundant, with a high degree of recovery. As has been said, the chief engineer and electrical engineer cannot make repairs on the spot. The systems will need to be able to resolve issues themselves.”
Addressing the legal framework
Lehtovaara noted that the IMO and flag states are currently considering the conditionally uncrewed bridge on open waters. “The lookout is the main legal challenge here. Within highly autonomous solutions, we have ferries and tugs, but these are very different from global international traffic with large vessels. Here we need to go stepwise forward and look at individual solutions and how they fit into specific contexts before the regulators can get involved in decision-making.”
He informed that ABB has proposed an autonomous lookout function. “Detection, identification and verification are all performed on the bridge. Detection can be performed by machines, while humans need to evaluate and verify. This is the way forward for open seas. But in areas of dense traffic, the question is how to use technology to assist the crew in all the functions. Uncrewed navigation in crowded straits will be one of the last things we do.”
Lack of regulations to either support or forbid autonomous operations present a risk, Lehtovaara maintained: “Legally the captain is responsible because they make the decisions, but how do we evaluate the correctness of the information on which decisions are based? This is a grey area. Digitalization is bringing forward information that has a huge impact, but it falls outside the regulatory framework. If class has no rules to relate to regarding information, then information falls outside the evaluation process.”
Kenneth Lim, Maritime and Port Authority of Singapore (MPA) Chief Technology Officer and Senior Director, Innovation, Technology & Talent Development, concurred on the many safety concerns that need to be addressed in the march to autonomy, emphasizing that they need to be identifiable and manageable: “This will require further investigation from class, ports, and operators. We need to figure out what is doable and what is not. The roles of the captain and others need to be ironed out before we can have truly autonomous solutions.”
Tewari highlighted the potential benefits of autonomy and the challenges it currently faces. “We can already implement certain levels of autonomy. This approach can be used to enhance safety when the capabilities of automation and autonomy are combined with the qualities of highly trained seafarers and excite the younger generation to consider careers in shipping. We could go in that direction first, and class is prepared to meet these requirements. Engine room automation is already allowed in SOLAS. It is only on the bridge, where SOLAS requires a lookout, that there is no immediate path to automation.”
Ibrahim agreed: “Class typically develops hazard identification documents and proposes solutions based on these. This level of operations can be addressed by unmanned and autonomous solutions today. The space is continually evolving as well, due to developments like 5G, and new contributions will be made as this evolution continues.”
The cost of communications
Addressing the moderator’s question of whether the cost of communications is a significant factor in achieving autonomy, Groot replied that most ships are already equipped with basic communications infrastructure: “But what we have today covers the data sharing side. We also have to consider cyber security. The connectivity is there, but security is lacking,” he maintained.
“We would like to see increased situational awareness to enhance safety, more like we have seen in autonomous vehicles,” he added. “Better situational awareness would allow us to provide better predictions, but we don’t have systems that can provide a secure connection between information systems and operations yet.”
Autonomous operations: revolution or evolution?
To round out the discussion, moderator Lampert challenged participants to take their place on the ‘revolution or evolution’ spectrum of autonomous development.
Mahesh Krishnappa voted in favor of evolution: “The journey started 20 years ago with discrete sensors. Now sensors continuously monitor operations and provide information on what is happening with the entire ship. We have plenty of data. Now comes the issue of what to do with it. We are taking the steps one at a time, looking at how digitalization can increase safety and productivity along the way.”
Liza Chua, Associate R&D Engineer, Electric Solutions in ABB Marine & Ports, offered that digitalization is gradually changing the way crews work: “The collection and processing of information has increased, and this is helping us to achieve autonomy at different levels. The initial stage of remote control will come first. Using remote control, one operator could operate multiple vessels as well. We are taking small steps forward, and this is changing the working environment.”
Sanjay Kuttan opted to straddle the issue: “I have to say both. There are three major trends influencing developments: decarbonization, digitalization and automation.
Telecommunications and connectivity are also evolving. When these factors mature and converge, there will be a revolution. Today ship design is under-optimizing all of them. Future design will do the reverse, and maximize them.”
Kenneth Lim agreed that many factors are moving forward in parallel: “Artificial intelligence is advancing, and supporting technologies are being developed independently. This allows the innovators to put things together in a clever way and come up with something we might not have expected.”
Brijesh Tewari hedged in favor of evolution: “But suddenly, one day, we may wake up to a revolution!”
Anis Hussain foresees evolution within technology and automation, but revolution in ship design: “The yards building today’s ships will not necessarily be those building tomorrow’s, much less the ships of a more distant future. There will be a completely new model driving design and construction, and this is where I believe there will be a revolution.”
Imran Ibrahim noted that the day’s hybrid event constituted a good example of a revolution: “In the same way as we have embraced digital communication, it depends on whether the industry chooses to exploit autonomous technologies at a sustained pace. Right now I would say it appears to be a mixed bag.”
Rens Groot again related to the major changes taking place on ships: “We have already seen a revolution on board: people do not have insight or control over the equipment. Crews have traditionally been taught that they have to be able to do everything alone. The revolution is that they cannot do it alone any longer. They need help from shore. Dependence on shore assistance is the new reality.”
Alf Kåre Ådnanes took the long view: “Our grandchildren will call it a revolution, but we will see it as evolution. We need to make the right steps going forward, and each step has its value. All the elements that make sense by themselves coming together will create a revolution, but it will happen slowly enough that we will experience it as evolution.”
Eero Lehtovaara cast the day’s final vote for evolution, and a slow one at that: “Autonomy will be achieved step by step. The changes will not be big, nor is there a real need for dramatic changes in the short term. We are definitely talking about evolution.”