Interview with Mr Kevin Daffey

Dr Sanjay: Good day everyone, and thank you for joining us in this new episode of the SMI Horizon, a series of interviews and commentaries by distinguished academic and maritime industry thought leaders. 

In today’s episode, we have the distinguished pleasure of having with us Mr Kevin Daffey, who is SMI Distinguished Visitor for 2020. Kevin had recently participated in our SMI Forum and delivered a very enlightening keynote presentation. Kevin is the current president of IMarREST, and the Director of Marine Systems and Automation at Rolls Royce power systems. Prior to this, he was with Rolls Royce’s commercial marine business, where he was the Director of Ship Intelligence, Engineering and Technology.

For the SMI Horizon today, we hope to ask Kevin a couple of questions to get his perspectives on autonomous vessels and how we should prepare for it. So welcome, Kevin. How are you today?

Mr Daffey: Yeah, thank you Sanjay for talking to me today. I’m very well, I’m staying fit and healthy at the moment. Very difficult times as you know, around the world, but the work continues, and I’m certainly enjoying things at the moment. We’re doing all the work at the Institute of Marine Engineers, Science and Technology, to continue the great work we do advising the IMO and other stakeholders on matters related to marine, engineering, science and technology. So, really enjoying life at the moment, and happy to talk to you today. I’d love to be in Singapore today, because I think it’s a lot warmer than it is here in the United Kingdom where it’s only about eight degrees Celsius. 

Dr Sanjay: Yeah, we’re definitely warmer than eight degrees, but we did have a spat of rain, which kind of cool the whole island down, which was quite welcomed as well.

Maybe we just get into it. A lot of people are doing trials with autonomous vessels. You see it happening all over the world. The great shipping nations – Norway, Japan, they are all trying out these autonomous vessels. This seems like it is the future of the maritime industry, the way the investments are going. If it’s going to be the future, then what capabilities and technologies we need to start investing in today that will make sure that we have a safer journey in the future with autonomous vessels?

Mr Daffey: Okay, so let me unpack that question a little bit. Let’s just start – is it the future? I think definitely it is the future, and it has a role to play. I think, when I started the journey in 2016, there was an element of thinking potentially that this was going to be quite disruptive, and it would happen quite quickly. But I think it’s going to be more of a progressive change. As you know in the marine industry, in shipping, vessels are built and they generally last something like 20 to 30 years, and so it’s a huge investment. So therefore, you’d have to do something that’s really, really profound to basically get the ship owners around the world to renew their fleets. I don’t think that’s going to happen, so I think this is going to be a progressive change. 

We’ve got around about 80,000 vessels that are on the oceans today, and it’s unlikely that we’re going to see a profound move to this technology in the short term. The areas where it’s taken hold, and we’re beginning to see a lot more demonstrations, but also people investing in the technology is around coastal areas. There’s quite a lot of activity going on in the tug market to look at how autonomous tugs could operate. The reason that people are looking at that is because tugs are quite low utilized vessels. People on board those vessels are potentially put into harm’s way because they’re operating close to other vessels, they’re connected to other vessels. Therefore, there is a motivation to potentially move people out of harm’s way. Also, the cost of crew in these vessels is quite expensive as well. So, in terms of operating costs, it’s one of the bigger operating costs. So that’s one really good example of where I think autonomy will help that particular segment of the marine market. Then we can think of things like exploration or science exploration that has been on autonomous vessels in this market for a long time. And I think that will continue. Other areas could be the offshore industry, supply vessels and other vessels that could be operating around production wells or exploration. The reason there again is related to putting people into harm’s way and removing people away from those sorts of situations as well. 

And the other area, of course, where autonomous vessels are really beginning to be attractive is in the governmental area for the Navy. So, the US Navy invested an incredible amount of money into this technology. Other nations, like the United Kingdom and France, are also developing this technology, and particularly the Chinese as well. I think, in this initial sort of next five years, I think most of the push will be probably in the naval market, with a lot more of these demonstrations going on in the commercial marine market around coastal waters. 

One particular area or region of the world that I know is showing a lot of interest is Japan. Japan is an area of the world where it’s difficult to transport cargoes across land, and therefore coastal shipping is really important to Japan. I think they see this local geographic motivation to move towards using autonomy to reduce the cost of shipping, and to make it as safe as possible around their shores. And of course, finding the crew as well to man crewed vessels, it can be a challenge. So actually, going to a crew-less configuration is a benefit to them. That’s really the main motivations I think that are happening today. It’s going to be a progressive change.

I think probably the last area that we will move across to autonomous vessels will be ocean-going deep-sea vessels. The cost of crewing, although it’s an important parameter, is not the main parameter. Those vessels probably got the longest in terms of longevity, so renewing those fleets is going to take a long time. At the moment, most of the fleet renewals are focusing on decarbonisation, and reducing the effect of environmental issues like ballast water treatment, as well as reducing CO2 emissions. So, you know, a lot of the fleet renewals that are happening at the moment are really focused in that area. And we got to see what role autonomy does help in terms of the reduction of CO2 and environmental emissions, as well as the operating costs as well. 

I think the second part of your question was really around the technologies that can be developed to support this area. So, I think if we start on some of the technologies, I think are going to be important. First of all, there’s going to be areas around situational awareness where sensors and sensor fusion we can begin to build up a really good tactical picture of what’s going on around the vessel. So, this becomes the eyes and ears of the vessel owner, and the idea there is that we provide a really good tactical picture around the vessel. And that can be used with crewed vessels as well as un-crewed vessels. It can potentially help to make the maritime industry safer, even for all of the crewed vessels that we have today. So that’s going to be one of the key technologies, and there is a lot of activity going on to this area, and that has been for many years. Then the whole area of having navigation systems, again, is an area where there’s been quite a bit of activity to come up with navigational algorithms. Some of them are deterministic algorithms, and some of them are non-deterministic, and based upon machine learning or even artificial intelligence. I think the latter is going to take a little bit of time to mature because of course, the whole area of AI is an area where there’s a lot of work in terms of how would we certify, how would we make these technologies satisfactory to flag states and people to have trust in them.

Another area that we need to think about is around how do you make the ships autonomous ready. Because just having a layer of automation that allows the ship to operate and make decisions around where it goes, and how it gets somewhere, has to be backed up by how do you make the machinery more reliable. At the moment, that area has not really been addressed, because all of the machineries on board rely on humans to do maintenance for them. I think this is an area which has not really been addressed by anybody from what I can see, and that’s an area where we need to accelerate that type of activity. Maybe we can talk about that a bit a little bit later in the interview, Sanjay.

Dr Sanjay: Kevin, thank you so much for that. That really kind of painted the state of play today, and obviously how you see the state of play tomorrow. You did actually get into the question on technologies, and situational awareness is a major part of it. Obviously, a robotic chief engineer I guess you would call it in the future. On situational awareness, we’ve got a lot of technologies out there today, from the efforts on autonomous cars. How is that technology different for marine vessels, and what else needs to be done an up the ante to make it safer?

Mr Daffey: There’s quite a big difference between car vehicle technology, and that is required for marine. If you think about cars, you’re really looking at a few meters ahead of the car, maybe 100 meters ahead of the car if you’re on a highway. Whereas in the maritime, we’re looking at determining targets which could be several miles away. Because if you think of the largest vessels that we have today, their ability to react takes many, many 10s of minutes to effect course changes. Therefore, you really need to have a much wider field of view. Of course, the marine environment is a lot different with the salt water environment, the range of different weather conditions that we encounter as well. As well as the fact that ships are moving in more degrees of freedom. You’ve got the heave effect, as it were. And effectively in heavy seas, for smaller vessels, you can actually lose contact with the target where you would not necessarily have that with vehicle technology. So, you have to think about what happens when we’re in heavy seas. We temporarily lose contact with a target because there’s a wave in the way. How do we then continue to track that target where we might lose line of sight with it, and there may be no other recourse, apart from maybe an AIS signal, to know where that other target is, as it were. 

Cars don’t have to contend with half-submerged cargo containers, or mammals that might be in the way. This is really important, because actually there’s a great opportunity with autonomous vessels to sort of deal with the impact of actually hitting things like mammals. Because at the moment, there’s nothing really stopping ships just ploughing through the ocean, inadvertently hitting sea mammals. But actually, if we could detect well, then potentially we could take some corrective action or provide some sorts of warning. So, it’s an area where the IMarEST has been doing a little bit of thinking with its marine mammal special interest group and thinking, well, what could we do differently to actually protect our precious mammals in the sea, because again, these got to be protected. So, there’s a huge range of differences between vehicle and marine technology for situational awareness. 

There are some commonalities. So, first of all, the vehicle industry has been thinking a lot more about the sort of the ethics of what does it do in certain really difficult scenarios around. If it gets to a point where a collision is almost unavoidable as it were, what decisions does the car make? So, there’s been a huge amount of thinking in this area. The classic one is, does a car, when it is faced with the choice of running over a young family, or maybe people that are older. They’re trying to sort of conflict this type of really difficult ethical argument. So, they’re trying to understand that. 

On the sensor side, there is some commonality. Some of the LiDAR technology that’s been used in vehicles. People have been looking at using LiDAR, so this is the laser sort of equivalent to a radar effectively in the marine industry. But the vehicle LiDAR has a much shorter range, and probably not as robust as the ones that we need for marine use. The ones that we’ve been, that I’m aware of we’ve been using in the marine side, have come from the sort of naval LiDAR which are used on the back of naval vessels to help helicopters land on the back of naval ship decks, as it were. So, they seem to be more robust and built for the marine environment, and have got the longevity and availability that’s required. 

But I think there’s a lot that we can learn together. Some of the areas around what happen where we have a blend of automation, autonomy and operators, where we expect the operators to be in the loop. Now the car companies have been thinking long and hard about this. I mean, on some vehicles you have these sort of autopilot systems that take over the steering and the speed control, but they require the human to have their hands on the wheel at the same time. They’re looking at things like eye on the road and hand pressure on the wheel, to make sure that the human is there in the loop. I know that some of the AI systems that have been done by some of the American companies have been thinking long and hard about, how much do I have to be applying the human in the loop. If we’re on an open freeway where nothing’s going on, does it really matter if the human looks away from the front of the windscreen for a few seconds. But if they’re in a sort of metropolitan area near a school, then actually they want them to be really, really paying huge attention and to take control. So again, this type of thinking could be applied in the marine use. When we’re on the open seas, actually we could go to an unmanned bridge. But when we’re in harbour areas like Singapore, you’d want somebody to be there doing human watchkeeping, as well as automated watchkeeping. So, these are the sorts of things I think that we can learn from the vehicle industry, and apply it to the maritime.

Dr Sanjay: I think you brought up an important point because today, the vessel already interacts with the port quite a bit in terms of pilotage and things of that sort. So, the question now is, while the vessel is owned by the vessel owner, how does the port prepare for this? The port obviously wants to provide the services and would like the vessel to come to the port. So how does the port prepare, and what do they need to start thinking about? For example, remote control centres in the likes. Thoughts around the port’s role in this adventure?

Mr Daffey: That’s where the rubber hits the road, isn’t it? It’s in the port, because clearly loading cargo, unloading cargo. Let’s start with some practicalities. So, first of all, the most labour-intensive operations in port are docking and mooring. If you went back 200 years, we’re still mooring vessels in the same ways we did 200 years ago – with ropes. I say that labour intensity is huge. There are technologies out there today, using vacuum suction to assist the mooring operation, so actually reducing the time of mooring and the complexity of mooring, would immediately bring a benefit in terms of the amount of crew that’s required on board, the amount of people that are required alongside the vessel as well. Potentially to make that a safer operation, because mooring still inherently carries some risks, with ropes and chains being used to assist that sort of operation. So that’s some of the practical ways where autonomy could help and automation could help in that arena. 

Secondly, there is a lot of activity going on by some of the bigger shipping companies in the world like Maersk, to look at how they integrate together all of the disparate parts of the cargo infrastructure, where, a manufacturer, an end user, the freight forwarders are involved with this, the insurers, the ship brokers.  All of these areas are being addressed by using platform type technologies to try and make them such that they do integrate together. So, I think this digital connectivity, which will come from the freight forwarding side of the world, down through the charterers, and the ship operators, I think that’s going to be an important side of things. So, this is almost like the Amazon platform, which allows customers and suppliers to connect together. I know there are some efforts going on around the world to create these sorts of platforms. But of course, who should be the owner and arbiter of this platform because of course, value is going to be shifting away from the traditional parts of the maritime industry, maybe to some super platform operator, which could be somebody outside of our industry. It could be a Google. It could be an Amazon. Would people be happy if it was one of the ship owners that did it? Probably not. Because then of course value would be moving towards potentially a monopoly. But if we have lots of disparate systems, then this interconnectivity between different countries and different ports is going to be a real challenge as well. So, there’s a huge number of issues there that have to be solved. And that’s, you know, is it really a port and a ship thing? No, it’s actually more of an industry thing, because we have a fragmented industry. And it continues to want to be fragmented, because it doesn’t want anybody to steal a mark and steal a lead. 

The other thing with the marine industry is that countries have come together in the International Maritime Organization, specifically to facilitate the free trade across the open oceans and between countries. Anything that sort of creates monopolies within that sort of industry is going to be hugely resisted. So, the only way I think that we can do is to sort of create some sorts of open standards, and that will overcome it. So, in the aircraft industry, open standards are being created around air traffic control. A common language, English language, is used to do air traffic control. There’s a set of common standard terms which are used; the way that aircraft are dispatched and controlled around airports is common; the standards around how we number runways, etc. So, can we apply this sort of common standard way of thinking that’s been matured in the aircraft industry, and bring that into the marine industry. We’ve been very successful in the marine industry around environmental standards, the training of crews, standards around that one classification. So, we can do it in the maritime industry. It probably doesn’t answer your specific question about what happens between the ship and the port, but these are the bigger picture items I think that comes out of that question.

Dr Sanjay: I guess when we start thinking of this whole ecosystem, like you said it’s just not the port and the vessel, if we realize the complexity of this whole endeavour of going autonomous.  Singapore is kind of committed to want to be part of this technological wave. We’ve invested significant resources with CEAOPS, the Centre of Excellence for Autonomous and Remotely Operated Vessels at TCOMS. From the work that we’re trying to do there, how else can Singapore assert itself from a technological leadership perspective? There’s some uniqueness about Singapore’s port operations where there are 1000 ships within a very narrow port water. It must be a nightmare for autonomous vessels to navigate. In a sense if you can survive Singapore port waters, then you can survive anywhere, because I don’t think it gets any more crowded. What are your thoughts around this? How does Singapore maintain and uses its unique disposition to actually provide a technological advantage for the maritime sector?

Mr Daffey: I’ve been involved with working in research in Singapore since 2012, if not before. So, first of all, there are some unique characteristics of Singapore, which I think are unique and I think we need to bring to bear. So, first of all, you’ve got a great multicultural research community. Some of the best researchers in technology can be found in Singapore and you breed them from different nationalities, different nations. You bring them together, into the city, into the great research institutes that you have. By doing that, in bringing them into close quarters, you actually bring learning from other industries together. So, one of the areas that you’ve been doing a lot of work is around manufacturing, technology and robotics. So, let’s just take that example. Earlier we talked about the artificial chief engineer. Well, who’s going to take control in an autonomous ship’s machinery room? Do we have robots that can be installed into machinery rooms, a standard set of tools that the robot can access, all of the maintainable items on engines and compressors and water evaporators, etc?  You know that they always have the same sort of common standard fixtures and fittings so the robot can actually get access to. Is this something that Singapore can steal a margin because as I said before, this autonomous-ready is an area where I think people have left that behind. They’re sort of thinking that this is just about putting a layer of automation onto a ship that will navigate it, and it’s absolutely not that. How do we know an engine can spend three months at sea without any intervention, or the only intervention it gets is from a robot? Then there’s this sort of remote working. So, if the robot can do some standard tasks, but if it comes across a task it can’t do, how would you then give the person in the remote operating centre the ability to work remotely with maybe not a great data connection, to make that intervention? So, what can I learn from things like the space industry or other industries, where that sort of constraints on remote working have been overcome? 

Other areas to look at is around the human machine interface. I think there’s a huge amount of work that you’ve done around simulation of vessel operation, training of crews. So how do we bring that to bear in a remote operating centre, where at times the human is overseeing the operation of ships, to the point where the human is then taking a positive control of the vessel, but that vessel could be many kilometres, 1000s of kilometres away as well. 

One area that all countries have to think about is, well, how do we regulate this? So, the IMO is looking at how we regulate the Maritime Autonomous Surface Ship, MASS as they call it, for deep ocean, intercontinental, inter-country operations. But also, how are you going to regulate it, if you’re just looking at maybe tugs or crew boats that operate within the harbour, you got a lot of tugs that operate there. So, Singapore is going to think about, well, what are we going to do to regulate this industry in our ports to make sure that it’s both cost effective, and it’s safe as well? So, these are really important things that we need to think about. 

So, I think, yes, as you pointed out, Singapore is in a unique position, because maritime is really at the heart of what Singapore is all about. It’s all about free trade, a massive port infrastructure, a massive maritime history. So how does it bring that to bear against what it does really, really well, which is about technology development, and this sort of learning society that it has created, this knowledge society, and using that technology for the benefit of everybody around the world.

Dr Sanjay: Kevin, listening to you, what came to mind is, the future ship has to be designed differently if it’s going to be able to accommodate a robotic chief engineer, is going to accommodate or not have any accommodation for people, and start rethinking the way the vessel actually looks like, from a design perspective. What are your thoughts on the kind of areas where Singapore could consider coming up with the designs of the future that would lend itself to autonomy?

Mr Daffey: That’s a great question, Sanjay. This is where I think all of the design constraints that we find on existing vessels fall away. So, taking away habitable spaces on board a vessel because you don’t need them, it’s quite profound because all of a sudden, you’re now optimizing, rather than around the crew that are operating and stewarding the cargo on board, you can really design around the cargo and you can maximize cargo and freight carrying capabilities, whether it be passengers or cargo, cargo probably more. You’re getting rid of machinery, which adds to availability or lack of availability, water evaporators, and HVAC systems, all of these systems are either going to disappear or they will reduce. I mean, we’ll still need ventilation and cooling systems on board for the traditional machinery, but you won’t have to pump it into all of the habitable spaces. So, it allows naval architects to think really, really differently and if you see some of the concept designs, you can begin to see some of these designs beginning to take evolution where the focus is on what the vessel does, if it be a tug, which is there to take a towing line, then you can see the sort of the capture for the towing line and the winch is the key part of the vessel, rather than a large bridge with a winch at the front, that’s the key thing that’s on the vessel. 

And for the cargo carrying vessels, the concepts we start to see more focus around the cargo being at the centre of what’s going on. And then a small amount of ship deck space devoted to the sensory side of things to keep the vessel going. Some of the more mature designs are looking at even interchangeable propulsion systems where you can go from a mechanical to maybe an electric propulsion system with batteries, depending on the mission to create vessel flexibility. This is another thing that’s becoming more possible as well. So, I think that we’re really scratching the surface in terms of the degrees of freedom for naval architects. Again, I think there’s a whole range of activities of innovation that Singapore can participate in from its naval architectural teachings and research to begin to do this. You can use the facility at TCOMS, the big pool there, to maybe test some of these concepts out with real weather and real wave conditions, and test out what they do. 

The other thing is that you don’t need to create a vessel that has the same level of comfort onboard from passage through different sea ways. Potentially, you could send an autonomous vessel through quite dangerous seas, that we will consider dangerous, because you don’t have to think about seasickness of anybody on board. The artificial chief engineer I hope is not going to get seasick. We shouldn’t be replicating that feature, I don’t think. So again, internal standard is another area I think maybe we can touch on. If we are going to put robots on board, what standards do we create for machinery space so they use the same fixtures and fittings, and the same accessibility points to do the maintenance on engines and other pieces of machinery? So that’s going to be quite an interesting discussion with the machinery makers to see, well, how would we achieve that?

Dr Sanjay: And I guess when it comes to near shore, the comms technology, be it 5G or now China has come up with 6G, becomes far more critical in terms of the ability to have that shore-vessel communication, decision making, reactiveness and manoeuvrability, and things like that. But that’s an investment that the port needs to make, I guess, going forward?

Mr Daffey: Yeah, it is an area where the ports are going to have to make some changes and some investments. But again, I think if we get the standards right, then there maybe if we can minimize the investments that are made shoreside to maximize the investments which are made on the shipping side, maybe that will be a good thing. Or maybe it’s the other way around, minimize the investments on the ship side and maximize the port infrastructure. Ports are very expensive to change. Singapore is going through this move of its current port to a new location. You know how many billions that have been spent in doing that. Obviously, it’s creating a future port, one which is digitally connected, where autonomous actually plays a part on the land side effectively, to get the cargo from the ship to the hinterland. There’s a huge amount of autonomy into the cranage, the movement of containers which are going along. I mean that’s an exciting area where autonomy is playing a role, just around the operation of ports. That’s hugely exciting, and an area where we potentially could see some crossover from both port operation into ship operation, where we can learn from each other because it’s the same industry which is involved with developing that sort of technology. So yeah, exciting times.

Dr Sanjay: You know Kevin, I always enjoy talking to you, because the time just flies when we get into the discussion and the Q and As, and there’s just so much to talk about. We could go on, technically. Well, I think we’ve come to a point where we say thank you so much for sharing your thoughts on this very exciting area. It’s definitely at its early stages, and we’re going to see more developments in going forward all over the world. I think having you as a friend of Singapore in the maritime sector, we hope to have more conversations and pick your mind on how we should move forward with a bit more excitement, and technological safety as well because of our crowded ports. But again, thank you so much Kevin for your time and joining us on SMI Horizon.

Mr Daffey: Thank you very much, Sanjay, really good to talk to you today and will speak to you soon. Thank you.