How Data is Changing Arctic Navigation with Railotech and Memorial Univ.

Evolving sea ice conditions haven’t made sailing in the Arctic any simpler. If anything, increasingly complex ice regimes, year-round operations and new trade patterns are raising the stakes for shipowners who need to keep people, assets and schedules safe in some of the harshest waters on the planet.

In this episode of Setting Course, an ABS Podcast, Rob Hindley of Railotech (formerly Aker Arctic), Dr. Oscar De Silva of Memorial University of Newfoundland, and ABS Senior Engineer Ed Moakler join host Brad Cox to explore how new data and tools are reshaping Arctic navigation.

They discuss ICESIGHTS, an ABS-led initiative that gathers and interprets sea ice information in real time, what that could mean for bridge teams, and how operational data can loop back into future ice-class ship design and life cycle decisions.

Share this episode on social media, leave a review on your favorite podcast platform or send feedback to podcast@eagle.org. Learn more about how ABS is supporting the maritime industry at www.eagle.org.

Takeaways

• The Arctic shipping landscape is evolving as conditions in the region continue to change.
• Year-round transportation in the Arctic is now feasible for some ship types.
• Data availability is increasing, enhancing operational decision-making.
• The ICESIGHTS system aims to provide real-time guidance for navigation.
• AI can complement traditional physics models in ice navigation.

Guests

Rob Hindley is Head of Consulting and Technology Development at Railotech. His role involves leading the practical application of arctic technology to new, and often novel, ice-going ship designs, transportation systems and offshore structures. Previously Rob worked for Lloyd’s Register, where he held overall technical authority for ice class, winterization and implementation of the Polar Code. This included an assignment representing IACS at the IMO during the development of the Polar Code and coordinating the development of POLARIS, an ice risk evaluation system used to set operational limitations for ships in polar waters. He is a Chartered Engineer, a fellow of the Royal Institution of Naval Architects, and holds a master’s degree in Naval Architecture from Newcastle University. He is currently undertaking postgraduate study at Aalto University with a focus on arctic shipping risks and regulations.

Dr. Oscar De Silva, PhD, P.Eng., SMIEEE, is an Associate Professor in the Department of Mechanical and Mechatronics Engineering at Memorial University of Newfoundland, Canada. He received his PhD from Memorial University of Newfoundland. His expertise is in sensing and navigation system development for platform autonomy using model-based and certifiable data-driven approaches. Prior to joining academia, he worked as a research fellow on computer vision systems with the American Bureau of Shipping Harsh Environment Technology Center in St. John’s. At Memorial University, he leads an NSERC-funded research program on resilient navigation for autonomous platforms. His group collaborates with ABS to develop AI-assisted multi-sensor ice navigation system integration and software for Arctic vessel field trials. 

Ed Moakler, P.Eng, is a Senior Engineer with the ABS Harsh Environment Technology and Digital Research Center (HET&DRC), located in St. John’s, Newfoundland and Labrador, Canada. Ed specializes in ice mechanics, ice load monitoring system design and operation, vessel capability in ice assessments, advanced structural analysis, and the application of the IMO Polar Code. Ed is responsible for the upkeep of the ABS Ice Class Rules and Ice Class specific tools, both in house and external.

Transcript

Brad Cox (00:07)
Welcome to Setting Course, an ABS Podcast, where we're charting the future of the marine and offshore industries. I'm your host, Brad Cox, and today we're talking about how technology is changing the way ships operate in some of the harshest waters on the planet: the Arctic. 

We'll be exploring ICESIGHTS, an ABS-led initiative that's developing a shipboard system to detect and characterize sea ice in real time, giving bridge teams tactical guidance on how and where to proceed. Joining me are three guests who sit at the heart of this work. 

First, we have Rob Hindley, Head of Consultancy and Technology Development at Railotech. Formerly known as Aker Arctic, Railotech is a Finish specialist of icebreakers and ice capable vessel design. Rob, thanks for joining us.

Rob (00:44)
Great to be here. Thanks, Brad.

Brad Cox (00:45)
And from Memorial University of Newfoundland in St. John's, Canada, I'm pleased to welcome Dr. Oscar De Silva, Associate Professor of Mechanical Engineering. An interesting fact about Oscar is he previously worked as a research fellow with the ABS Harsh Environment Technology and Digital Research Center, where he developed a computer vision system for ice detection, which has some relevance to what we're talking about today. So, thanks for joining us, Oscar.

Oscar De Silva (01:07)
Thanks for the intro, Brad. Glad to be here.

Brad Cox (01:09)
And we're also joined by Ed Moakler, a Senior Engineer at the ABS HET&DRC. Some of our listeners may know Ed from the recent ABS LinkedIn series that followed his voyage to Antarctica to put ICESIGHTS to the test. Ed, welcome to the show.

Ed Moakler (01:22)
Thanks, Brad. Happy to be here.

Brad Cox (01:23)
So, let's go ahead and jump right in here. To get us started, I'd like to set the stage a little bit. Rob, from a ship design standpoint, what's the biggest shift you've seen in ice-class shipping over the last decade?

Rob (01:34)
From an overall perspective, it's really the last 20 years where we see a substantial change in how Arctic shipping operates. And that comes all the way back to Norilsk Nickel in 2006, which was the first Arctic going containership that actually operates year-round in the Arctic. And since then, actually, we've had year-round transportation in the Arctic. Moving to larger projects like the Yamal LNG export from Siberia where year-round export of LNG has been going on since 2017. And that really is a result of the technology, the double acting ship concept that Railotech kind of pioneered, but also technological advances in sensing, so ice sensing and being able to understand what the environment is. So, I think that's really the shift. The shift is year-round transportation is now possible.

Of course, there is relatively few ships operating in the Arctic and a lot of the routing outside of these large projects is seasonal. We expect that to grow as the ice retreats and I think what we do see is an increase in interest from various different shipping segments actually in the last decade.

Brad Cox (02:32)
Thanks, Rob. So, Ed, from what you're seeing from the industry, how is decision making around ice operations evolving? What are operators most worried about in the next five to ten years?

Ed Moakler (02:42)
This actually follows on perfectly with what Rob was saying and how there's still very few ships that are operating year-around, but we're seeing a huge shift in the ice conditions. It's becoming more accessible. We're seeing lots of non-ice class ships and low ice class ships getting in there. So, the number one concern always has been, and will still continue to be over the next five to 10 years, is that safety aspect. 

Especially as we're seeing newer operators entering the space, because there's lots of economic opportunity up there. There's lots going on, shorter shipping routes, everything. There's lots there. But they don't have the experience. They don't know how to identify the ice and categorize it. And maybe they don't understand the risks and the remoteness and the hazards they're operating in. 

So, we've seen the IMO introduce the Polar Code back in 2015, 2017. And that's been a big shift. It's difficult to obtain data, especially up north. It's getting a lot better every year. Lots of new satellites and more passes every day. Ice charts are still good. But it's just a matter of getting enough data for them to operate safely and efficiently in such a remote environment. And that's where we're seeing the most from our clients is just that inexperience. And then, OK, how do we get the next step? How do we make it safer and collect the right data and plan accordingly?

Brad Cox (03:48)
So, I think this feeds into the next question I have for Oscar really well. So, Oscar, from a research perspective, how is the Arctic as a data environment changing? Are you seeing more interest in that measured ice and ice ship interaction data?

Oscar De Silva (04:00)
Yeah, so definitely more interest because different projects that we've been doing at our lab, at least, thus far has continuously shifted towards the Arctic domain because there's more interest from ABS and its partners and also the government of Canada and other Scandinavian research institutes. So, there's many partners around the world converging on this aspect because there's a lot to be done and it's a challenging environment and challenging amount of data to deal with. 

So, what changed is the amount of data that's available. So, I think everybody talked about the amount of data and also a way to handle that amount of data because before the advent of AI, we didn't have a nice framework to what to do with this data. We have millions of data, so what? Because there’s a nicer framework to work with the data in a manageable way, we have different things that we can do with the data. 

On top of that, we also have sensors. So, sensors is like a palette that we can work with because there's a finite amount of sensors out there. So, we can measure all those variables that's out there. So, we have worked with that palette and that palette keeps increasing. For instance, there are recent sensors that have come to the market, which are way more high resolution radar, at a short range of course, which is very relevant to us. And also, way more high resolution LiDARs with the capability to get the Doppler velocity. For instance, for an ice floe, velocity is very important. Now there are LiDARs which can capture the velocity by a single measurement using the Doppler velocity principle used in radars. So, this palette keeps increasing and the amount of data keeps increasing and we have a way to kind of put them in a nicer framework to work with. So, that made all the difference. 

One last point would be connectivity. For instance, Ed’s recent trial down in the Antarctic. We had the luxury of having video calls with Ed back and forth and debugging the system. So, the nice thing is Ed basically gave us remote access to his computer and we could debug the system then and there and update the code then and there. So, that was pretty cool, which was made available by satellite internet connectivity that's available in high bandwidth. So, all these things kind of come together to make this thing possible. So, it's exciting.

Rob (06:00)
There's a really little story, well not a story, related to that but the change that Ed experienced. So, we did trials with the Sir David Attenborough three years ago and so Ed joined the second trials. And on the first set of trials, we basically didn't have internet for five weeks. It was wonderful. Like everybody just, everybody said, okay, well they're off in the Antarctic. There's no connection, you know, peace and quiet. But this time around, we had full connectivity. So, not only did Ed manage to kind of get in contact with Oscar all the time, but we were in contact with everybody. And that really is a huge step change in how operations in the Arctic will develop in the future. Just being able to download multiple satellite images as you would do at home. It's really a game changer.

Brad Cox (06:41)
Yeah, so I would like to talk a little bit about the ICESIGHTS system in itself. So, Ed, you know, obviously we mentioned that you were there at the trials in Antarctica. What could ICESIGHTS actually change in a day-to-day bridge operation?

Ed Moakler (06:54)
So, right now, ICESIGHTS is still in its infancy. Like Oscar said and Rob said, we were sitting up in the observation deck for I think six hours and Oscar and I were debugging live over Microsoft Teams and fiddling around and making it work. So, we're still in a data collection phase, but the ultimate goal is to provide some sort of guidance to the crew. You can never replace the judgment of the master and the captain and the crew, XO, whatever the structure is, you can't replace 30 years of experience with an AI. We want it to be an operational guidance tool. What that guidance will look like will differ on the ship type and the captain and what they like. But yeah, some sort of guidance. We're going to use the cameras and LiDARs and radars, whatever tech, like Oscar said, whatever's on the palette for that specific ship and profile and system, we’ll tailor that to their needs to provide some sort of guidance. 

Maybe it's a red-yellow-green of, this ice is bad, this ice is good. It could be lead detection. It could be changing conditions. Maybe optimizing fuel, reducing loads. We can, kind of the world's our oyster there. It's going to be a tool for the crew to use without replacing it. I'm not sure we're going to end up there in, not at least in my career, maybe after that. For the next 20 or 30 years, yeah, we want it to be just a tool to just make operations that much safer and a little bit more efficient for the operators.

Brad Cox (08:12)
And I'm sure there's a tangent we could go down with how this technology could inform autonomous operations. So, Oscar, you mentioned AI. So, where does the AI part really matter here? What does it actually let crews do that traditional ice navigation methods can't?

Oscar De Silva (08:27)
Regarding that, Brad, so first off, disclaimer, physics models, we are not getting rid of it. Physics models is what allowed us to go to the moon and basically, you know, go to space. We didn't have like much of data to go on those voyages. 

So, what AI models can do is work in unison with physics models to fill those gaps where you don't have the accurate information. So, for instance, I can give you an example from the work that I did with ABS back in the day, back in 2015, when we were trying to assess pack ice conditions using vision. The methods back then were pretty rudimentary. So, you use the sensors, geometry models to figure out that's ice, that's not ice floe and the size looks like this. So, this kind of information were captured using what we call models of the sensor, inverse models of the sensor rather than relying on newer methods like data-driven AI models or neural networks which are trained for the purpose. But as time went by, this research developed into incorporating AI into it and that solved a lot of problems. 

So, for instance, sunlight reflections that happens on ice floes and on the surface of the sea is hard to deal with if you're dealing with a rudimentary vision-based system. But if you're using AI, that's pretty much easy to deal with because you train the AI so much that it knows, okay, that's sunlight, that's not an ice floe. That kind of training can be done using AI. 

So, that's where AI fits nicely, perceiving scenes and understanding the scene so we can capture, okay, what’s ice, what's not ice, what type of ice is there and is there a threat there? So, that kind of thing, it's very good at. For example, in the ICESIGHTS system, in the current generation, we have a way to create a map of the surrounding environment when the ship is traversing in ice conditions. But these maps are currently built in unison with AI models and also physics-based models. So, for instance, geometry is used to basically figure out what the LiDAR is seeing, how to translate it into a map is figured out using geometry. Down the road, it might turn into an AI model, but we’re not there yet. That'll happen down the road as the amount of data increases. 

We see parallels of this with the Tesla autonomous fleet. When Tesla first released its fleet, it was basically hardware capable to sense all these things. It was not software capable to do all these autonomous driving. So, what happened was as the data accumulated, they were able to build better and better models. And then eventually have those models running on the cars themselves through an overnight firmware upgrade using an internet connection. So hopefully, we go into something like that where we make ships hardware capable at least using different generations of ICESIGHTS systems on ships. Eventually using our data centers we'll have better and better models which you can upload on those systems to view better and better tools for our mariners to navigate with.

Brad Cox (11:07)
I think in most cases more data is always better. You see Formula One right now with their new engines and things and they’ve just got to get more data so they can figure out what they're going to do with it. So, Rob, from the designer perspective, how does a tool like ICESIGHTS change the way an owner can think about operating an ice-class ship?

Rob (11:24)
I kind of want to touch on the data bit, but I know we're also going to talk about the data bit in a bit. But of course, more data is better as long as you actually use that data. The idea with ICESIGHTS from a designer perspective or the benefit to us is that essentially, ICESIGHTS is a data collection tool that would be mounted on every ship that is operating in ice. And that is a really phenomenal set of data that we can use to inform how the ships are designed. And we'll talk about design maybe a bit later on, but we have to remember that the design tools we have as well as the classification rules are based on an understanding of ice loads and how ships and ice interact that's based on a relatively small number of ship trials and dedicated data sets. 

What's really important when we come to looking at how to understand how capable a ship is operating on ice and what ice's risk is to be able to match what the loads are on the ship and what the ice properties are. And that you can't do without characterizing the ice, without understanding what the strength is, what the thickness is. And that combination of measuring the ship response and the actual ice characteristics is really only, to date, it's only really been done during dedicated full-scale trials.

And dedicated full-scale trials, I think, I haven't looked this up and maybe I should have done, but there's probably less than 200 ever done of any ship in ice ever. And out of those, there's a certain number in a specific sea area like the Baltic, which is not directly comparable to the Arctic, for example. So, the data sets that we're using are quite small and ICESIGHTS essentially allows us to potentially in the future gather all that data all the time from all the ships that are everywhere to feed into what we know about ship ice interaction. 

From a designer's perspective, that's really exciting. But I think that's a way off because of course we have to calibrate how ICESIGHTS works first and that's what we were doing in the Antarctic. We were measuring loads directly on the hull and then looking at what the ICESIGHTS equipment is doing in terms of characterizing what the ice is. 

But I think the other point related to how this helps us is actually more about the conversation to have with operators and about what the operational envelope is. I am not a boat driver. I am not a captain, but I do understand that it's very subjective. Ice is a very complex regime, it's not uniform. And the ship's response to ice is very much dependent on the ship. So, as Ed alluded to earlier, there are 30 years of experience, a captain can feel what ice is a danger to the ship, where he or she is pushing the ship in terms of its limits. 

But how to translate that into something where we can sit down and draw an engineering envelope around it and say, okay, well for these types of conditions, you know, these are the limits. These are—and if you want to call them operate operational envelopes or capabilities or whatever—we need to have a stronger link between the operational side and what the ship is experiencing in terms of loads. And then to be able to have that conversation with operators and then to set the ship specification in terms of the structural strength, in terms of the ship performance in ice around those conversations I think is really valuable because at the moment we are very much dependent on, well, more than anecdotal experience, but certainly it's more semi-empirical than it is based on, you know, a direct relationship in the real world of operation. So yeah, it's a really, really exciting direction that we're going.

Brad Cox (14:48)
Pivoting to some of the broader implications, Ed, how do you see this kind of data changing conversations around class, safety margins and life cycle management in Arctic shipping?

Ed Moakler (14:57)
I will answer your question, Brad. I just wanted to pile on to something Rob said about the trials and everything. That was one of the great things on this expedition on the Sir David Attenborough is just the amount of data that was collected. Obviously, we had our ICESIGHTS system, but there's an ice load monitoring system. There was stuff all along the shafting. It was everywhere. It was all along the hull. The amount of data—it may be one of the biggest data sets ever collected on an ice trials.

If we can actually get all of that data together and time stamped and everything talking to one another, I think there's so much potential. And if this is something we can maybe eventually try to make a bit of a framework around and maybe set the standard, then we can try to get more of this ice load data and matching it and helping build the ICESIGHTS system. It was just something I wanted to reiterate that Rob had mentioned because it's absolutely key to this whole thing of that categorization and then those ice loads. 

But if we assume that we can get all these ice loads, Oscar talked about physics models, Rob talked about the measurements on the hull, and we've got ICESIGHTS. If we can get all that together, we can estimate the impact, the impact loads and their location along the hull. From there, maybe we can look at some of the digital tools that we're seeing now in class around 3D plan review, digital twins, these sorts of technologies. We can start looking at things like fatigue life. If there's particularly high loads, if we saw a really nasty piece of ice when we were clipping along at 10 knots, 12 knots, maybe there's a potential for damage there. So, that can inform surveys. Or coating degradation. All of these sorts of things so we can try to take all that and then inform the surveys. 

And it'll help reduce costs and increase efficiency throughout the life cycle of the vessel. If we collect enough data and we're able to kind of correlate all these things, maybe then it informs the next generation of rules as the ice conditions change. And if we're seeing changes in loads for ships that are set up to measure loads. It's a bit cyclical and it's going to take decades, but I think there's a lot of potential here to have more pointed surveys, or maybe point is not the right word, but to have more focused surveys on those areas that we're seeing higher impacts and then the rule development can follow on from the outcomes of that.

Brad Cox (17:05)
Okay, and Rob, we kind of talked about this a little bit already, but a lot of ice-class design builds on more than a century of experience. But if you had that kind of reliable high-resolution data we're talking about, what could that unlock in future designs?

Rob (17:18)
I think we did touch on it, but I think it is worth expanding on. The ice-class rules are kind of like the foundation for what a ship from a structural perspective or a strength or a safety perspective are built on and the data points used around those rules are relatively few. And I can certainly kind of go into what Ed was talking about. There is a real potential to develop the rules in a way that align more closely with the ice conditions. 

Or let me put it another way. If you look at the ICE class rules, the descriptions about what type of conditions they are expected to operate in are very loose. They're almost nominal and they don't give a very good understanding to either regulators themselves or administrations or operators about what the limitations are in terms of the ice conditions you can operate in. And we've tried to develop some tools. Polaris, which is the IMO based system for assessing ice conditions and operational limitations, which both ABS and me in a former role have been closely involved in the development of, goes some way to that. But it really doesn't paint the whole picture.

And of course, if you don't have that kind of close understanding of what the ship can do, then you obviously have to design for the worst case. You have to put in conservatism in terms of structural strength, which of course, more strength means more steel, which means reduced cargo carrying capacity, et cetera. So, the more we can bring together the data we get from ICESIGHTS, which brings this connection between what the ice conditions are and what the loads are, we can really start to set the level of design to align with the ice conditions and the operations that the ship is expected to do. And that is somewhat an optimization target. But I think it does have a kind of a safety perspective of it or related to it, which is of course that it's not always optimization for minimum weight, it's optimization around the knowledge in terms of what ice is dangerous and designing for those conditions. So, I think it's really going to feed into our designs in the future.

Brad Cox (19:19)
And, you know, safety is certainly core to everything we do and always kind of a sub theme throughout every podcast episode. So, Oscar, how do you see research institutions like Memorial fitting into this loop, you know, between data collection models and next generation designs?

Oscar De Silva (19:34)
Yeah, so we have different ways that we connect to this overall ecosystem of research that's going on in this area. One of the ways is, for lack of a better word, labeling crew. That's what we have as students. So, they have interesting projects going on in this area, which labels a lot of this data to make use of it. That's one of the areas that we can incorporate students a lot because they get exposure to the area as well, like different subject domains when you're engaging in this labeling activity, because it's application specific. And they also get to play around with the AI models at the infancy of their engineering program. So, that's pretty cool and a lot of students have high praise for it because rather than filling out some Excel sheets or some kind of other thing, they get to do an AI model and they get to get some technical knowledge in this area. So they have very good experience coming out from these types of activities. 

So, HQP training. So, in other words, highly qualified personnel training is one of the main outcomes that we see is very impactful from this type of initiatives with the industry engaged because there are many examples that I can delve into where a student gets into a project which is in a very specific area and they get hired by the company. There's a lot of examples where these things happen and that's a good pipeline even for industry because you have a student which has worked on a project for a long period of time with a lot of training in this area. 

Another last thing is industry and academia, they work together in a lot of things. So, for instance, Ed and myself and Rob and there's a lot of other partners as well, we are doing it because we are very interested in the problem, because there's a lot of things to explore in the problem. From my end, I'm a robotics geek, so I'm trying to fit robotic knowledge everywhere I can in the overall framework. Ed is coming from a different angle, so he's looking at, okay, well, you can do that, but that won't fit, because that's contradictory to some sort of a safety aspect that's there. Likewise, when all these parties work together, there's a optimization happening there as well to figure out the correct solution for the problem at hand. Overall, it's a very good experience for the students and also myself. It keeps things interesting.

Brad Cox (21:45)
So, I think that overlaps with the next question quite a bit actually. So, you know, I wanted to talk a little bit about the future of all of this and I know we've touched on it at different points, but Oscar, this project brings together class, designers and academia. So, from your viewpoint, why is that mix important for Arctic innovation to actually reach ships?

Oscar De Silva (22:03)
One of the things is we are engaged in engineering research. Unlike other faculties, engineering research is always looking at, ok, what's the new science and how to apply it to a problem. So, the mix helps us with identifying a good problem to apply the science. 

Another thing is, as academia, we have the luxury of working on something for a very long period of time without somebody firing us. So that's a good luxury to have and that's a good luxury that the industry can make use of because they have some dream that should come true in 10 years’ time. This is like a nice boiling pot to kind of try things out and develop the technology to like levels which you can market and commercialize. 

One last point is coming from a design side of things, we always try to build it, break it, then fix and repeat. So you go through that cycle, right? So, since you want to do that cycle, unless you have the industry, especially for this type of project, you cannot break it and fix it because we won't be able to test it out in the real-life situation. We can do it in a lab setting, which I think is not a good way to do this type of research because you really need to know, what are the failure cases? You really need to know, okay, what are the edge cases to deal with? 

One last point would be like Ed talked about standards. So, for standards, there should be a lot of data and lot of scientific rigor going into validating a particular idea. So, for instance, if it's ship ice load monitoring is the goal, there should be some scientific quality results generated and reviewed by the scientific community and also validated through different parties to get the amount of confidence needed to go to the next step of standardizing it and also going to quality level of it. So likewise, when you're working with industry and academia, you have the ability to take that interesting problem and apply that scientific rigor into it, like a larger time scale, so you can build it to the level that's needed for commercialization. 

Brad Cox (23:54)
And so, Ed, obviously I know we're still in a very early developmental stage, but what does validation look like for a system like ICESIGHTS? How do you build trust with operators who will rely on this in difficult conditions?

Ed Moakler (24:06)
Yeah, it's a tough one. So, like most things, validation comes down to your inputs and the outputs and correlating that data and making sure it's actually correct. We had the advantage on the SDA of having Railotech and crew come on and put some strain gauging on, build up all the ISO monitoring math models. And we've got great load data that Rob and I and Oscar have been fiddling around with. That's kind of the first step is getting that there.

There's other things. We can use different simulation tools that we have in-house, other things we're developing. There's kind of multi-prong there. We can look at the physical data, measured right on the hull. Not always possible. These things are not easy to install. Of course, there's costs. Commercial ships are going to struggle with that. So, maybe we look on more of the simulation side and validate based on the other measurements. And all of these things are also ship-specific because the response is based on that ship structure and layout. Validation is definitely a bit difficult. But again, it's data driven. If we do it the right way, we'll get there on the validation side. 

One of the things, in talking with lots of captains, commercial, Coast Guard, Naval, they do not want more bells and whistles or alarms on their bridge. They've got enough of that already. So, our system has to fit their need. The needs are going to change depending on how the ship is operated and their drivers. A search and rescue operation using ICESIGHTS is going to look a lot different than a cargo ship going through the Northwest Passage or through the Baltic Sea. It's going to look very, very different. So we're going to have to tailor what the system looks like and provides to the individual ship and crew. 

So, we've got to keep that in mind at these early stages. The data we collect, we've got to be able to pivot in different directions to provide the crew with what they need. And then when they provide us feedback, we've got to be able to action that. It's the only way to build trust with these folks. Being on the bridge, it's not very often I get to go to sea, but being up on the bridge, there's enough stress, especially when you're in ice.  And if our system isn't providing what they need, they're not going to have that confidence in it. So, we've got to be able to truly take their feedback and then action that into something they're going to use and prove to them that it's another tool in their tool belt to help them operate safely. So, that's probably the biggest thing. So, we'll get that data validation done and then being flexible and able to create the guidance that they need in that moment.

Brad Cox (26:21)
You mentioned alarms and I know alarm management is a hot topic right now. 

Ed Moakler (26:25)
Oh yeah. Yeah.

Brad Cox (26:27)
So, Rob, Ed mentioned the bells and whistles, but when you look at the technology trends, even outside of ICESIGHTS, what are you seeing that's going to be really the most disruptive for Arctic shipping?

Rob (26:37)
I think actually it's a big picture view. What comes after ICESIGHTS is what does ICESIGHTS fuse with and this kind of idea of sensor fusion. Essentially what ICESIGHTS, at least my vision of it, is it's a near-field sensing tool. And I love this word tool. I think it's really important. So, Ed, thanks for, I'd already written down that I was going to say tool. Because it is a tool to help the operator, just as the ice charts are, just as radar is, just as the eyeball is, it's a tool. 

But I think what we have the advantage of going forward, which is, is I think close to game changing, is then the potential to fuse all this together. So, to be able to overlay what we're getting from ICESIGHTS, which is a characterization of the near-field ice environment, with what we're getting from satellite imagery, which we can now download at vastly higher speeds and higher resolutions in the middle of the Arctic Ocean, to overlapping that with radar and then maybe crowdsourcing. Like what happens if there are five or six ships in the next 100 nautical miles that all have an ICESIGHTS box on that are all getting this data, fusing all that together to really paint a picture of what the environment is and knowing what that environment really means to the ship.

And I'll go back to the point that I made at the beginning, that ice trials and ice measurements are relatively few and far between. That is principally because of the cost, the cost of taking the ship off hire and measuring the ice and doing all of this as dedicated trials. The fusion of all this technology essentially means we're gathering data wherever the ship goes. What are the risks? What are the conditions? And from that, we can really start to paint informed risk maps of what it means for different ships for different ice classes or maybe there won't be ice classes in the future. Maybe it will be some sort of rating on the ICESIGHTS scale. Let's see where we go with this. Sorry, that just came off the top of my head.

But the main point is that with all of this data which ICESIGHTS will help us validate that we can characterize the ice in this way, it really does open up a lot of possibilities not just for the near-field tactical navigation, but more related to routing, ice avoidance, long-term voyage planning. Once we start overlaying these layers of data and these layers of inputted information then it really becomes more than just one tool, it becomes a multi-tool. And that I think is really exciting in terms of being able to operate ships in more remote environments with a known risk level. That's what it's about. It's not saying we can go everywhere with it, but it's saying that in these more remote environments, we get to understand what the risk level is in near real time and then also, when we're voyage planning, to understand, should we be committing to routes or voyages through these areas and be able to quantify that in a way that we do our best at the moment with the tools that we have, but this really is a big step forward.

Brad Cox (29:17)
Before we wrap up, I want to leave listeners with some closing thoughts from each of you. Rob, in one or two sentences, what do you most want shipowners, operators and charters to take away from this discussion about Arctic shipping?

Rob (29:28)
I would say let's look to the future a bit in terms of, climate change means that ice conditions will be more variable and there will be a need to actually use the data we have to characterize and understand what the ice is. We can't necessarily rely on what we've been doing before. We need to be able to get the data to understand what ice we are facing and what we will face in the future. And then I think out of that, the takeaway is that we can and this type of project is getting us closer to a stage where we can really understand the risks more completely and we should all work together on that. The point that I made earlier about crowdsourcing or sensing technologies on different ships, this only works in collaboration or at least we get a lot further with collaborating a lot quicker. I think this project is a good example of that but I think the Arctic maritime community could take a step in that direction too.

Brad Cox (30:18)
And Oscar, any closing thoughts from you?

Oscar De Silva (30:20)
So, if you look at the industries like self-driving cars and even generative AI and humanoids, these are taking those industries by storm. They didn't know how it's going to change and all that. And if you look at maritime shipping, Arctic maritime shipping, that also has the potential to have that kind of a step change because of this cloud connectivity and because of all these sensors and a lot of this data that's been accumulated and the AI that can help with that data. So, thanks to that, my envision is it would look like, if you get on a Tesla, you have that nice bird's eye view of what's going on around the environment. And it keeps getting better and better of understanding of what's the near-field is looking like and how to plan the path. Do you want something like that on the hull of the ship? That's the question. If ICESIGHTS can do and can develop the AI models to bring it to that kind of a display using sensors and certifiable AI, that's a very good outcome of the project. I’d say cloud AI-driven risk awareness will allow us to boldly go where no ice-class ship has ever gone before.

Brad Cox (31:18)
And Ed, closing thoughts from you.

Ed Moakler (31:21)
Yeah, I just wanted to echo something that Rob said, and I think it's that collaboration. This expedition that we went on over Christmas was a truly international collaboration. It allowed us to take this next step forward and develop some new technology. So, that's something I wanted to echo again.

As the Arctic becomes more accessible and those ice conditions change and the risks change, it's going to be more and more important to do things the right way. And by right, I mean safely. So, this means proper planning and getting the correct training and the correct tools and technology to help ensure that the ship and the crew and the environment, we're reducing that impact and we're keeping everybody safe and everybody gets back home. 

There's tons and tons of opportunities. We all know schedules and economic drivers are more than they ever have been. But these areas are extremely remote. They come with unique hazards. And systems like ICESIGHTS are going to help us there. But we've still got to try to do these things the right way and do it safely.

Brad Cox (32:18)
Okay, great. So that's great place to leave it. Arctic shipping is clearly moving into a new era where better ice data, tools like ICESIGHTS and close collaboration are changing how the industry designs and operates ice-class vessels. So, thanks again to Rob, Oscar, and Ed for helping us explore what that could mean on board and across the fleet. Thanks, guys.

Rob (32:36)
Thank you.

Ed Moakler (32:37)
Thanks, Brad.

Oscar De Silva (32:37)
Thanks, Brad.

Brad Cox (32:38)
And for everybody listening, thank you for joining us for another episode of Setting Course. Be sure to subscribe, leave a review, and share this episode. To learn more about how ABS is supporting ice-class vessels, visit us at www.eagle.org. Thank you for listening.

End (32:52)