27. Sécurité en mer
Durée 00:34:20
Spotify
Apple Podcasts
YouTube
Transcription
CORDIScovery podcast- Episode #27 - Safety at sea
This is an AI transcription.
00:00:10:05 - 00:00:40:09
Abigail Acton
This is CORDIScovery. Welcome to this episode of CORDIScovery. With me, Abigail Acton Energy World Grain Supplies, Consumer Goods. So much more. Ships transport a remarkable 90% of the world's commodities, as the United Nations Secretary-General Antonio Guterres said, without ships and the women and men who work on them. Economies would stall and people would starve. Vital to the flow of the commodities we all depend on.
00:00:40:14 - 00:01:08:11
Abigail Acton
Seafarers often work in dangerous environments, and the pandemic revealed how vulnerable they are to social restrictions. The United Nations estimated 400,000 seafarers were prevented from going home in September 2020, often stranded on board as the pandemic forced the imposition of travel bans. In recognition of all we owe them. The International Maritime Organization founded the International Day of The Seafarer back in 2011, and it falls on the 25th of June.
00:01:08:12 - 00:01:37:15
Abigail Acton
So this episode, we're looking at how EU funding is working to make life at sea safer for passengers and for crew. Today's guests, whose work has been funded under the Horizon 2020 program, are Franz Evegren, who is the director of the Fire Safe Transport Unit at the Research Institute of Sweden, Rise. Franz focuses on using innovative technologies, operations and applications to reduce the incidence and impacts of fires.
00:01:37:15 - 00:01:40:21
Abigail Acton
On board. Roll on. Roll off car ferries. Welcome, Franz.
00:01:41:01 - 00:01:43:03
Franz Evegren
Thank you very much, Abigail. Nice to be here.
00:01:43:05 - 00:01:55:16
Abigail Acton
Also at the Research Institutes of Sweden, is Luis Sanchez-Heres, senior researcher at the Maritime Department. Luis is interested in harnessing the power of AI and machine learning to improve safety at sea. Welcome, Luis.
00:01:55:18 - 00:01:57:20
Luis Sanchez-Heres
Thank you. Abigail. Thanks for inviting me.
00:01:57:22 - 00:02:13:15
Abigail Acton
Lazaros Karagiannidis is concentrating on how smart applications and wearables can make ship evacuations faster and safer. Lazaros is based at the National Technical University of Athens, where he is senior researcher and project manager at the AI Sense Group. Hello, Lazaros.
00:02:13:17 - 00:02:15:09
Lazaros Karagiannidis
Hello. Thank you for having me.
00:02:15:11 - 00:02:39:12
Abigail Acton
Very welcome. Wonderful that you're here with us. Franz, I'm going to turn to you first. The Lash Fire Project is developing operational and design solutions to improve fire prevention and where fires do break out. Ensure better management on roll on, roll off ferries. Franz, the public doesn't necessarily really hear much about fires on Ro-Ro Ferries. But the International Maritime Organization has decided that more needs to be done to prevent them from breaking out.
00:02:39:14 - 00:02:42:04
Abigail Acton
Can you tell us a little bit more about the context, please?
00:02:42:06 - 00:03:05:03
Franz Evegren
Yes, in fact, there have been quite a few fires on this so-called roll on roll off ships or Roro ships, as we refer to them, carrying vehicles. So you roll on and roll off the vehicles. And if you look at the past incidents, there have been about two and a half major fires on these Roro ships every year.
00:03:05:05 - 00:03:32:23
Franz Evegren
You might have heard of a couple of them in the Mediterranean. There have been a few significant fires on the Norman Atlantic and the Sorrento, but there have been quite a few fires. And this was picked up by the International Maritime Organization, which is a United Nations specialized agency, more like a global parliament for ship safety, including all the nations and a number of the recognized organizations internationally.
00:03:33:00 - 00:03:40:11
Franz Evegren
So they decided that something needs to be done, and they also open up for regulations to change. And that's where our research really came in.
00:03:40:16 - 00:03:57:00
Abigail Acton
Right. Super. Thanks very much for that explanation. Okay. That's interesting. So clearly there was there is a need to to work out what can be improved. So what are some of the potential solutions that the project identified relating to the issues that you've just talked about here, about the incidents of fire?
00:03:57:03 - 00:04:21:15
Franz Evegren
So yeah, in our our project is based on a number of previous projects where we identified a number of challenges that have caused these fires. We looked into past accidents, of course, and we also looked into the future on the coming changes in like the reformation of the vehicle fleet into a larger use of batteries, for example, or hydrogen as fuel.
00:04:21:17 - 00:04:56:21
Franz Evegren
And the issues came down to a list of 20 challenges, which we grouped into partly the four stages of a fire, if you may say so prevention, detection, extinguishment and containment. And we also had two work packages focusing one, focusing on the operations so the people and what they do on board firefighting, for example, and also on the design, the design interface between the humans and the technologies, which has often been a cause of confusion and costs larger consequences of fires, you might say.
00:04:57:00 - 00:05:03:19
Abigail Acton
Can you give me some examples of what you just mentioned there about the human interface with the technology of where that has gone wrong in the past?
00:05:03:24 - 00:05:28:09
Franz Evegren
Absolutely. And so, for example, it might be the case that when you get an alarm on the bridge, it says that the fire has occurred by frame this and deck that in the starboard part of this and then you might have a CCTV camera with a numbering and then you have if there is an established fire or a confirmed fire, you need to activate the extinguishing system in a section.
00:05:28:12 - 00:05:45:22
Franz Evegren
And all these systems are individual and causes or gives a lot of ground for mistakes. You can say so. And even if when you come into the drenched room where you activate the direction sprinkler system, there might be a different numbering on those valves compared to the sections on the deck, so to say.
00:05:46:01 - 00:05:51:11
Abigail Acton
And all of this under a time of great stress and urgency, way, way, really, you want things to be smooth flowing.
00:05:51:11 - 00:05:58:03
Franz Evegren
Right? Right. And in a panic or a very stressful situation. Yeah. Things need to be easy to make checked.
00:05:58:05 - 00:06:14:18
Abigail Acton
Yes, exactly. Not trying to struggle to find the right answer. Okay. That's a great example. And you talked about the likely move towards electric batteries for vehicles and stuff like that. The hydrogen you mentioned as well. So what were some of the solutions that you were thinking with regards, for example, to anticipation of fire?
00:06:14:20 - 00:06:41:21
Franz Evegren
Well, we have looked into detection of the thermal runaway of a battery in an electric vehicle as well. That, I would say, is the area which has not come so far. We weren't addressing that in the project, but we were looking into it a little bit now. And in a lot of aspects of the project, we have actually transformed a little bit because of the large need coming from the operators and from the crews asking how to handle this new situation.
00:06:41:23 - 00:07:03:16
Franz Evegren
What we have done is we have looked into how to safely charge a vehicle on board. We have looked into whether or not the current extinguishing systems would be sufficient to control an electric vehicle fire, for example, and things like that. So we have adapted parts of the project to really consider the future reformation of the vehicle fleet.
00:07:03:18 - 00:07:10:14
Abigail Acton
And as you say, the request from the crew, because I know that you're very interested in having the feedback from crew for sure, from the actual seafarers.
00:07:10:14 - 00:07:30:07
Franz Evegren
Right. And like I said, many of these challenges that we are addressing, they really come from the crew and the concerns and the hazards that they see. And from the like I said, the different fires, the incidents in the past. And we really want to make sure that our research is not becoming a shelf warmer, but really comes to.
00:07:30:09 - 00:07:44:08
Abigail Acton
Yeah, not just academic work but has practical, practical solutions of the project. Yeah. Frans what sort of cutting edge technology are you using, for example, for fire detection? I know that you mentioned perhaps looking under the vehicles, right?
00:07:44:10 - 00:08:06:17
Franz Evegren
So we thought it would be very exciting to look into the potential of using drones. So we've done that in different ways. We've done it, for example, you might be aware of that. Drones are now used for maneuver board exercises. So you can send out a drone to if a man falls overboard to locate the person. And we also use it for getting an overview of the fire.
00:08:06:23 - 00:08:27:09
Franz Evegren
But we also use not flying drones, but rolling drones on the deck that can roll underneath the vehicles. And with the heat cameras and the sniffers or smoke detectors and etc.. And and they can be quite well-equipped because you don't need so many and they can kind of control the area when no one is on board any way.
00:08:27:11 - 00:08:29:05
Abigail Acton
And maybe areas that are inaccessible.
00:08:29:07 - 00:08:37:18
Franz Evegren
Right. And in particular, since the batteries are usually located very low in the vehicle, it makes sense to kind of be underneath the vehicle when you scan them.
00:08:37:19 - 00:08:44:22
Abigail Acton
Yeah. And I guess you could have them sort of like caterpillars going along most of the time throughout the voyage, really, and just kind of doing tours.
00:08:45:00 - 00:08:56:24
Franz Evegren
Right, Right. Yeah. So that's what we're looking into. And we've developed a pilot or a kind of an example of such a drone can be developed and, and used onboard and it's quite exciting to see it work.
00:08:57:01 - 00:09:12:01
Abigail Acton
Yeah, I know it must be and very satisfying and also very satisfying for the seafarers to feel that they're actually being heard by someone in a position to, to come up with some solutions. Okay. Right. Any other kind of solutions or problems that were identified that you are beginning to think of ways of handling them now?
00:09:12:03 - 00:09:33:22
Franz Evegren
Yeah, we are trying to use a quite high so called technology readiness level. So so we want to really make make solutions that can have an impact right away, not not core research, you might say. One of the things that we have looked at is screening all the cargo before it goes on board to look for hotspots on board.
00:09:33:24 - 00:09:50:02
Franz Evegren
These are systems that exist for other valuable infrastructure, such as tunnels, and we thought it would be very useful to also do it for the ships. So developed the system there, and that is now being demonstrated for some of the operators that we work with.
00:09:50:04 - 00:10:05:00
Abigail Acton
So if I can just stop you for a minute, France, I'm really interested in what you've just said about the notion of something that can scan the contents of vehicles to see. I suppose if something is potentially a flashpoint, could be a risk for fire hazard that exists already for going into tunnels you mentioned, right?
00:10:05:00 - 00:10:28:16
Franz Evegren
Yeah, there are companies focusing on this. And so several or a couple of kilometers before the tunnel, usually they scan the vehicle. So that they well in advance can set these vehicles aside, which have ignition sources or seem to be overheated. And combined with AI, these systems are quite smart. They can really see if it's a part of a vehicle that should be hot or if it shouldn't.
00:10:28:16 - 00:10:37:03
Abigail Acton
So fascinating. Yeah. And of course, it's really clear what you say there. There's clearly another process that could be useful there for for for ships and and boats coming aboard.
00:10:37:05 - 00:11:00:10
Franz Evegren
Right. And then we have also looked into extinguishing systems on the weather deck. For example, the weather deck is is the deck which usually see from outside the ship where you can see the actual cargo, the vehicles on top. So they're totally exposed to to wind and to to the weather. And previously they haven't really been and there haven't been any extinguishing systems or detection systems for these decks.
00:11:00:10 - 00:11:06:11
Franz Evegren
So now we're really trying to identify find systems that would work in such a harsh environment. Yeah.
00:11:06:11 - 00:11:13:22
Abigail Acton
I mean, as soon as you mentioned it, of course, it must be very difficult because you usually think of sprinklers and things like that as trenches coming from above. And if there's no roof.
00:11:13:24 - 00:11:15:12
Franz Evegren
Right. You don't have a ceiling. Yeah.
00:11:15:12 - 00:11:21:21
Abigail Acton
Yeah. And a fascinating. Thank you so much. This is really, really interesting. Franz, why did you get involved in this research? What was it that pulled you into this area?
00:11:21:24 - 00:11:54:14
Franz Evegren
Well, I've always been kind of drawn to the sea, I guess. And I knew I wanted to become a firesafer engineer and a risk analyst. So? So I thought that would be a great combination. And I also sailed for a few years and and I really know how it feels to be alone at sea. And when you have felt the big waves of the Biscay or actually was in one experience I had at the North Sea when when our ship was breaking down a little bit.
00:11:54:14 - 00:12:18:05
Franz Evegren
So, so we had to search emergency harbor. And I really felt small and far away from help. And and even if you're on a larger ship, I think that the resources are very limited. And in case there is a fire on board, you really need to be able to manage it on your own on the ship. And so that kind of drew me in to this area and I hope to make an impact and to make it make it safe from them.
00:12:18:10 - 00:12:36:05
Abigail Acton
I'm sure you will. It sounds already as if some of the ideas you were having are, and the ability to give seafarers a voice as well are definitely going to be making a difference, if not already. Thank you very much, Franz. I'm going to turn now to Lazaros. Lazaros SafePASS wanted to redefine evacuation processes, systems and equipment.
00:12:36:07 - 00:12:58:15
Abigail Acton
Your project looked at a better ways of monitoring hazardous situations on board and informing passengers and crew about the safest, quickest evacuation routes. So we've been talking here with Franz about shipboard fires and how to prevent them. But Lazaros, if a fire does break out, that can't be controlled, can you tell us what's involved in an evacuation in current circumstances, please?
00:12:58:17 - 00:13:26:13
Lazaros Karagiannidis
Yes. So once the situation is assessed and the passengers are not safe, anymore on board and evacuation will be, of course, initiated. And this involves the announcement of an evacuation signal. The announcement through the public announcement system to the passenger to move to a safe place, which is called Master Station. And in this process, the crew actually facilitates and guides the passengers to the master stations.
00:13:26:15 - 00:14:01:12
Lazaros Karagiannidis
And then to their obligation stations. But this whole process currently is a quite static process process meaning that passengers need to move to their pre allocated and designated master stations and they have to follow the static, let's say, exit signs. When all exit signs. And they have to wear their lifejackets and so on. And the common problem with this situation is that there might be situations where passengers are located far away from their designated master stations.
00:14:01:14 - 00:14:30:00
Lazaros Karagiannidis
Secondly, there might be evolving hazards like fires or flooding that may block existing evacuation routes or even the whole master station. So we need solutions that actually can, first of all, adapt to the evolving nature of hazards and adapt a bit to the evacuation strategy and also introduce novel means of evacuating passengers not only relying, you know, by the instructions by the crew.
00:14:30:01 - 00:14:50:05
Abigail Acton
Okay, that's clear. And indeed, I mean, any of us who've been on ferries, we've all seen those master signs. And clearly, you know, if you're at the other end of the ship getting a coffee, it's not necessarily the most practical place to go back to. So who did you work with in order to come up with a definition of what needed to be changed?
00:14:50:07 - 00:14:57:00
Abigail Acton
And once you realized what needed to be changed, what kind of ideas did you have to improve the situation?
00:14:57:03 - 00:15:24:02
Lazaros Karagiannidis
Yes. First of all, in our consortium, we have the chance to collaborate with different disciplines and backgrounds. For example, with academia, with shipyards, with cruise line operators, with ICT software providers. But even with lifeboat manufacturers, as well as organizations that are dealing with regulations and also independent ethics and personal data experts.
00:15:24:04 - 00:15:34:14
Abigail Acton
let me stop you for a second. I can see the logic of all the various groups that you're talking about there. But the personal data experts, this is fascinating. Why would they be brought on board? What's the relevance of of data protection there?
00:15:34:17 - 00:16:00:05
Lazaros Karagiannidis
Yes, because the solutions we are implementing also deal with personal data such as the ID of a passenger, such as the location of a passenger. So in order to be compliant with the current GDPR regulations, I mean, the purpose of data protection, we had to ensure that we are really compliant with those regulations and we need to take into account these our solutions, all these regulations.
00:16:00:07 - 00:16:12:00
Abigail Acton
So you've got this really kind of diverse team of of experts, but practical and legislative. Excellent. And what were some of the ideas that you came up with with the help of these these diverse people?
00:16:12:04 - 00:16:47:23
Lazaros Karagiannidis
Just some of the ideas, and especially from the shipyard and the cruise line operators was to, first of all, identify the mission and operational requirements during an evacuation. So we found out that the static application process is not that efficient. And this could be or let's say, an area of modification. So we came up with the idea of a dynamic application and let's say process, which means that we actually know the location of the passengers at any time using localization systems, their mobile devices and the smart life jackets that we have implemented.
00:16:47:24 - 00:17:16:05
Lazaros Karagiannidis
We can calculate the safest and fastest evacuation route based on a current situation on board, especially when we are having fires or flooding that can block some routes. We can calculate this safest evacuation route in real time and then inform the passengers through their different mobile devices or the dynamic exit signs. In contrast to the, let's say, static exit signs.
00:17:16:07 - 00:17:23:01
Abigail Acton
So back in a second. So by dynamic exit signs, we mean basically exit signs that would illuminate depending on a context.
00:17:23:03 - 00:17:36:02
Lazaros Karagiannidis
Exactly right. So these are digital displays that in case there is a blocked route, it will indicate that you should not follow this route. You should follow another route, another corridor, or to go to your muster station. Exactly.
00:17:36:04 - 00:17:46:13
Abigail Acton
And I believe you also considered what would happen in the case of a horrible idea this. But I mean, you know, I guess it does happen. Unfortunately, in the case of limited visibility where people can't see things very clearly.
00:17:46:17 - 00:18:16:15
Lazaros Karagiannidis
Yes. This was also one of the major aspects we had to look in. So we came up and came up with an ideal for smart life jackets. So in case you have a limited visibility, you can activate haptic navigation process, which means you can feel vibrations on your lifejacket that indicate the direction you have to follow or even activate and audio navigation through a speaker which tells you which direction you have to follow.
00:18:16:15 - 00:18:30:07
Lazaros Karagiannidis
And all this information is processed at the, let's say, control central processing platform that shares this information to the lifejacket, but also the mobile devices and so on.
00:18:30:11 - 00:18:48:16
Abigail Acton
And there's absolutely wonderful you're already making me feel safer. Both you and Franz are making me feel safer about my future ferry trips. And I use ferries a lot, so I'm a nervous flier. So I like boats. So, Lazaros, So can you tell me a little bit about training? Because this is all when we're in the middle of a really tough situation.
00:18:48:22 - 00:18:53:03
Abigail Acton
But I believe you've also done some anticipatory work in the form of training.
00:18:53:05 - 00:19:22:20
Lazaros Karagiannidis
Yes, exactly. And the main, let's say, need was to provide some novel training means, especially for the handling and operation of the lifeboats which are deployed on the on the cruise ships. So we implemented an augmented reality application that provides intuitive and easy to use, let's say, air glasses and playback, different kinds of training scenarios as well as maintenance.
00:19:22:20 - 00:19:35:04
Lazaros Karagiannidis
And our resource, the crew is, let's say, better trained and has a more a better experience, a better training experience through these devices.
00:19:35:04 - 00:19:47:01
Abigail Acton
So they can have an understanding of the impacts of the choices they make in a particular context and what that actually means in terms of how easy it is to deploy the lifeboat or and if it's deployed at the right angle.
00:19:47:03 - 00:20:07:24
Lazaros Karagiannidis
Exactly. And also, the good thing is that this type of training can actually be conducted at any time and at any place. So you don't need actually actual, let's say, a lifeboat or you don't need to be on board to make these trainings. And this will it was well appreciated by the trainers.
00:20:08:01 - 00:20:18:24
Abigail Acton
As well appreciated by me as well. That's excellent. Thank you. Very good. This is wonderful. Can I open this up and ask, Does anyone have any observations to make to Lazaros or any questions?
00:20:19:01 - 00:20:20:20
Franz Evegren
Yeah, yeah, I have a question. Yes.
00:20:20:20 - 00:20:22:00
Abigail Acton
Franz, please.
00:20:22:02 - 00:20:55:17
Franz Evegren
I was wondering also, by the way, as I understood it, you look at how you could alter the route to the master station or assembly station in a smarter way, and based on that, you know, the actual location or local positioning of the passenger. Have you also looked at maybe optimizing instead the master station designation for the different passengers based on the location of the passengers so that you go more to the closest master station instead?
00:20:55:20 - 00:21:30:04
Lazaros Karagiannidis
Yes, this was actually exactly the principle or the process. We we followed, of course, taking into account always the aim or regulations and to the specifications of the master stations. But knowing the exact position of the passengers, we could actually guide them to the closest and safest master station, of course. And in some cases even we did not even go through the master mustering process and directed them directly to the embarkation stations, which reduces very much the evacuation time.
00:21:30:06 - 00:21:59:19
Franz Evegren
It's a very interesting subject. I think we actually used a similar principle in the last ferry project where we there's something called a runner or someone gets an alarm to to go to the place of the alarm to confirm whether or not there is a fire. And most of the times it's not. It's a false alarm. And instead of having one designated person to do that, we send an alarm to the closest person, maybe the three closest crew members who are trained to investigate whether or not it's a fire.
00:21:59:21 - 00:22:03:23
Franz Evegren
So with the same principle that you know where all the crew members are.
00:22:04:00 - 00:22:05:20
Abigail Acton
It's about maximizing efficiency.
00:22:05:22 - 00:22:09:06
Lazaros Karagiannidis
Right? The complementarities between the projects.
00:22:09:08 - 00:22:29:21
Abigail Acton
absolutely. And I think you two have a of a lot of overlaps as well. I mean, I think it's it's fascinating. Thank you so much. I'm going to turn to Luis now. Luis, onboard safety also depends on highly accurate location positioning. So we've talked about what would happen on board with the fire and how we would deal with that if we, you know, unfortunately, it did actually occur.
00:22:29:23 - 00:22:49:18
Abigail Acton
But PREPARE Ships your project is exploring the way artificial intelligence, advanced sensing and Earth observation, data all interconnect not only to determine where a ship is now with greater accuracy and you're doing it with very great accuracy. I look forward to talking about that. But also where the ship will be, which is an important element for autonomous vessels.
00:22:49:24 - 00:22:53:07
Abigail Acton
So can you tell us how current systems work and what the drawbacks are?
00:22:53:10 - 00:23:15:10
Luis Sanchez-Heres
Well, currently in a ship that you have a, you're basically using GNSS system, which is the global navigation satellite system, the most common one that we talk about it every day is the G.P.S., the Global Positioning System, which is from the United States. But there's other GNSS such as Galileo or Beidou from China. Galileo is from Europe.
00:23:15:12 - 00:23:51:12
Luis Sanchez-Heres
And all these systems, they tell you where you are in the world and then you can pinpoint your location in a map. But this location that you get has a certain accuracy. So, for example, GPS, this has certain accuracy of a couple of meters, but then you can use different or extra systems to get higher accuracy. For example, you can use something called satellite based augmentation service that gives you maybe one meter accuracy, and then you can have other type of corrections like something called Arctic, that it will give you a centimeter corrections or a PPP correction that will give you dosimeter accuracy location.
00:23:51:18 - 00:24:01:05
Abigail Acton
Okay. But at the moment, I believe you think of the positioning systems as being either dynamic or, as you say. Kinematic Can you explain what these two types are?
00:24:01:05 - 00:24:28:05
Luis Sanchez-Heres
Well, one thing is determining where you are. So where you are in a ship, it is determined by usually GPS or visual submissions or radar. That's that. Those are where you are. Where are you going to be in the future? That's a completely different problem. So usually you use your velocity to determine where you're going to be in the future and that you could call it a dead reckoning predictor or a kinematic predictor, like a moving in that direction of moving at ten meters per second.
00:24:28:11 - 00:24:55:11
Luis Sanchez-Heres
Therefore, in a second they're going to be ten meters ahead. That's it. But the alternative to that and it's important for ships is that you, instead of predicting your future position with your current elicited, you can predict your future position with the current forces being exerted on the vessel. And this is very meaningful for vessels because there are so big they have such a large mass and moment of inertia that these forces are acting and they have consequences that are evident forward in the future.
00:24:55:12 - 00:24:59:05
Abigail Acton
So we're talking about things like currents and wind. And when you say Forces.
00:24:59:07 - 00:25:19:15
Luis Sanchez-Heres
and the propeller and the rudder and what it is. Right. I never say of that real vessel when in a simulator you turn the rudder and nothing happens and then you get desperate and you turn it even more. And then suddenly the basin is turning and you lose control because you don't have the right feeling that what you do right now, you're going to see the effects in the future.
00:25:19:17 - 00:25:39:02
Luis Sanchez-Heres
So from the point of view of control, it's actually quite challenging for somebody who not trained. And because of this problem or even for people who are experienced, if you have wind and currents and other vessels around that then, okay, how where am I going to be in one meter time in the current situation? So you have an idea that is challenging.
00:25:39:02 - 00:26:00:05
Luis Sanchez-Heres
Then seeing another vessel that it's turning and looking at its velocity vector and say, okay, where is that vessel going to be in? I made a time that's even more challenging. So in this preparation project, the was one part of it was the predictor system which actually uses information from on board a vessel and machine, learning to predict where that ship is going to be at a certain amount of time.
00:26:00:08 - 00:26:21:19
Abigail Acton
Okay. So let me ask you then, can you break that down a little bit more in terms of how that all comes together? So because I know that you're very interested in the use of artificial intelligence to crunch big data. So where are the sources of the data coming from Earth Observation and as you say, the forces and so on, And how are they coming together for you to to interpret?
00:26:21:21 - 00:26:48:18
Luis Sanchez-Heres
All right. So we have a Galileo, which is the European GNSS and then we also use corrections from from the Galileo system, which school has security system service. And that gives you this corrections. And we get a position in a map says, okay, latitude, longitude, we are here with Nascimento accuracy. Then we use that information and would mix it with information from the radar, from the anemometer, which is a thing that measures that wind speed and so on to calculate forces.
00:26:48:18 - 00:27:11:06
Luis Sanchez-Heres
And we put these forces into a model, a dynamic model. Now, this dynamic model, it's a bunch of equations with some parameters and coefficients. And these coefficients, we have learned them through experience. We record where we have been on where we are, and we use that information to adjust with machine learning these parameters, and then we can make a prediction out of that.
00:27:11:08 - 00:27:27:10
Abigail Acton
That's fascinating. I know that you were doing a pilot in Gothenburg Harbor. Could you tell us a little bit more about that? And your notion obviously, is that you want the artificial intelligence to do the heavy lifting when it comes to the the assessment, all the data and the interpretation of the data, how did that work out in Gothenburg?
00:27:27:10 - 00:27:28:00
Abigail Acton
Tell us more.
00:27:28:03 - 00:27:49:20
Luis Sanchez-Heres
Well, we actually got the three different types of vessels. We got a pilot boat, a yacht, Princeton's yacht, and we used to look for a system in a ropax, which is a railroad that also carries passengers on and there we installed the system. Some we do some situations where we got somewhat close to each other and we got to see the predictions off of the other ships.
00:27:49:20 - 00:28:11:12
Luis Sanchez-Heres
So that's an important component of the system that we developed. My predictions can be shared to all their vessels so they can see where I'm going to be and I can see where they're going to be and then try to evaluate what the people sailing these vessels thoughts about these predictions, how they're useful to resolve situations. When it comes to the machine learning part, a machine learning part that goes inside of this this system that we made.
00:28:11:12 - 00:28:30:20
Luis Sanchez-Heres
And it's the the calculation of these coefficients, but also the monitoring of the predictions. So the moment that the system says, like your predictions are actually not that good at the moment because of whatever reason, it shuts the system of a safety so that the person operating the vessel doesn't rely on predictions that are all reliable.
00:28:31:00 - 00:28:40:08
Abigail Acton
Yeah, no, that's great. So it removes the idea of false confidence and that's excellent. Okay. And how precise and accurate did you manage to go the predictions?
00:28:40:08 - 00:29:00:10
Luis Sanchez-Heres
Yeah well it's it's an here it depends quite a lot on the weather conditions on what kind of shape the larger the bathtub, the more accurate it is. A small vessel like a pilot boat they're planning. And then the models that we use to model the motions of the ship, they're much more complicated when it's a planning vessel.
00:29:00:10 - 00:29:05:13
Luis Sanchez-Heres
But for big ships, I would say that we got quite sufficient accuracy to make the system useful.
00:29:05:16 - 00:29:07:14
Abigail Acton
So what more variables in a smaller ship?
00:29:07:15 - 00:29:27:16
Luis Sanchez-Heres
Well, the smaller ships, as they start planning, Right. So there's two types of you can say that there's two types of vessels like the ones that, that displays the water like big ships and then the ones that really go so far that they lift off and they're kind of like planning on the surface. Okay. But the physics of that is much more complicated than the other one.
00:29:27:16 - 00:29:35:22
Luis Sanchez-Heres
And so coming up with a new a good set of equations to describe that and how the forces are going to move the ship, it's quite difficult.
00:29:35:24 - 00:29:37:10
Abigail Acton
So that's the next phase.
00:29:37:12 - 00:29:40:07
Luis Sanchez-Heres
That's yeah, yes.
00:29:40:09 - 00:30:00:08
Abigail Acton
This is excellent. Thank you so much. Work behind this. It's very complicated. And there's a you know, there's computational work involved and obviously mathematics and so on. It's very intricate. Louis Why did you get drawn into all of this? It sounds simple. With no mathematical ability whatsoever. It sounds quite kind of intense. What attracted you to to this?
00:30:00:10 - 00:30:16:20
Luis Sanchez-Heres
Well, I'm a naval architecture and been naval architect by training, and I like maneuvering. I think it's it's nice to do simulations about where a ship could be. And then suddenly we have all these tools and machine learning neural networks. And then the question is, can we use them?
00:30:17:01 - 00:30:25:01
Abigail Acton
Yeah. And also for the future because of autonomous vessels. Yeah. Can you tell me a little bit more about how that would feed in? I mean, you know, in your own words, why is it important?
00:30:25:05 - 00:30:48:07
Luis Sanchez-Heres
Well, for example, the model that we have, the predictive system that predicts the future motions of a ship, it has a dynamic model of a vessel and an autonomous vessel is going to have is going to require a dynamic model of the vessel. So this could be a component in an abdominal vessel to actually allow for an autonomous navigation system to steer that vessel.
00:30:48:09 - 00:31:06:10
Abigail Acton
Absolutely. I mean, we think of smart motorways with autonomous vehicles. But it's true, of course, the sea is also like a highway. And yeah, so then that that kind of infrastructure for the vehicles to in this case the vessels to be able to sense where they are and let other ships also predict where they're going to be obviously is crucial.
00:31:06:10 - 00:31:07:01
Abigail Acton
Yeah.
00:31:07:03 - 00:31:15:15
Luis Sanchez-Heres
Well it's more about like it's it's about control. Like if you have a computer and you want to control something, usually the computer needs to have understanding of what is it controlling.
00:31:15:17 - 00:31:16:02
Abigail Acton
Yes.
00:31:16:03 - 00:31:35:22
Luis Sanchez-Heres
So then in some sense where he got created in the context of other vessels, it could be like that model that helps the computer to control. So that's one thing. Then in this project, we also have this communications system. So then you could think of two autonomous vessels. They communicate to each other their route, their current motions and where they're likely to be.
00:31:35:22 - 00:31:41:17
Luis Sanchez-Heres
And they can use this information to plan their movements in a better way and avoid incidents. Super.
00:31:41:19 - 00:31:48:19
Abigail Acton
Absolutely. Thank you. Thank you very much. Very interesting. Does anyone have any questions for Louis? Yes, Franz. What would you like to say? Maybe it's an observation.
00:31:48:21 - 00:32:16:21
Franz Evegren
Well, yeah, it's it's more of an observation. In our project, we have looked into the positioning, the most optimal positioning of the ship. So it's not really predicting maybe the movement of the ship, but in case of a rather large Roro ship fire, how to best position the ship to make sure that you, for example, limit fire spread or that you give as good conditions as possible for evacuation.
00:32:16:23 - 00:32:24:21
Franz Evegren
So so I don't know if if any of the research you have done, Louis, do you think any of that could impact like in a fire situation?
00:32:24:23 - 00:32:52:14
Luis Sanchez-Heres
I mean, I guess it is dynamic model if you think of, okay, there's a fire and then you want to position the ship in the best possible way because help is on their way. But you don't want people in the breached start to move the controls because you want them to evacuate. Yeah. So then you could think of an autonomous navigation system that steers that ship just to make it easy for everybody to get out and for fire control or the fire brigade to take care of the situation.
00:32:52:16 - 00:33:09:16
Abigail Acton
But meanwhile, on board Lazaros’ system is letting everybody know whether to turn left or right by buzzing at them. If they're dark and it's confusing and it's all stressful. Look at you guys. You've got us all safely taped up. Thank you very, very much for all your work. It does genuinely make me feel much more positive about the future.
00:33:09:16 - 00:33:13:20
Abigail Acton
Pinkham votes. Thank you. Thanks for your time, guys, and for being with us today.
00:33:13:20 - 00:33:15:08
Franz Evegren
Thank you for having me. Thanks a lot.
00:33:15:11 - 00:33:16:15
Lazaros Karagiannidis
Thank you very much.
00:33:16:16 - 00:33:17:12
Luis Sanchez-Heres
Thank you very much.
00:33:17:12 - 00:33:39:13
Abigail Acton
Yeah, very welcome. That was super. Thank you. if you enjoyed this podcast and are interested in the latest scientific research coming out the EU, have a listen to previous episodes. Follow us on Spotify and Apple Podcasts or whichever podcast provider you favor. Are you curious about what other EU funded projects are doing in the area of maritime safety?
00:33:39:15 - 00:34:00:19
Abigail Acton
The Cordis website will give you an insight into the results of projects funded by Horizon 2020 and Horizon Europe that are working in this area. The website has articles and interviews that explore the results of research being conducted in a very broad range of domains and subjects from nematodes to Neanderthals. There's something there for you. Maybe you're involved in a project or would like to apply for funding.
00:34:00:21 - 00:34:14:22
Abigail Acton
Take a look at what others are doing in your domain. So come and check out the research that's revealing what makes our world tick. We're always happy to hear from you. Drop us a line. Editorial at cordis dot Europa. Dot EU. Until next time.
00:34:14:24 - 00:34:19:08
Unknown
Earlier this week.
This is an AI transcription.
00:00:10:05 - 00:00:40:09
Abigail Acton
This is CORDIScovery. Welcome to this episode of CORDIScovery. With me, Abigail Acton Energy World Grain Supplies, Consumer Goods. So much more. Ships transport a remarkable 90% of the world's commodities, as the United Nations Secretary-General Antonio Guterres said, without ships and the women and men who work on them. Economies would stall and people would starve. Vital to the flow of the commodities we all depend on.
00:00:40:14 - 00:01:08:11
Abigail Acton
Seafarers often work in dangerous environments, and the pandemic revealed how vulnerable they are to social restrictions. The United Nations estimated 400,000 seafarers were prevented from going home in September 2020, often stranded on board as the pandemic forced the imposition of travel bans. In recognition of all we owe them. The International Maritime Organization founded the International Day of The Seafarer back in 2011, and it falls on the 25th of June.
00:01:08:12 - 00:01:37:15
Abigail Acton
So this episode, we're looking at how EU funding is working to make life at sea safer for passengers and for crew. Today's guests, whose work has been funded under the Horizon 2020 program, are Franz Evegren, who is the director of the Fire Safe Transport Unit at the Research Institute of Sweden, Rise. Franz focuses on using innovative technologies, operations and applications to reduce the incidence and impacts of fires.
00:01:37:15 - 00:01:40:21
Abigail Acton
On board. Roll on. Roll off car ferries. Welcome, Franz.
00:01:41:01 - 00:01:43:03
Franz Evegren
Thank you very much, Abigail. Nice to be here.
00:01:43:05 - 00:01:55:16
Abigail Acton
Also at the Research Institutes of Sweden, is Luis Sanchez-Heres, senior researcher at the Maritime Department. Luis is interested in harnessing the power of AI and machine learning to improve safety at sea. Welcome, Luis.
00:01:55:18 - 00:01:57:20
Luis Sanchez-Heres
Thank you. Abigail. Thanks for inviting me.
00:01:57:22 - 00:02:13:15
Abigail Acton
Lazaros Karagiannidis is concentrating on how smart applications and wearables can make ship evacuations faster and safer. Lazaros is based at the National Technical University of Athens, where he is senior researcher and project manager at the AI Sense Group. Hello, Lazaros.
00:02:13:17 - 00:02:15:09
Lazaros Karagiannidis
Hello. Thank you for having me.
00:02:15:11 - 00:02:39:12
Abigail Acton
Very welcome. Wonderful that you're here with us. Franz, I'm going to turn to you first. The Lash Fire Project is developing operational and design solutions to improve fire prevention and where fires do break out. Ensure better management on roll on, roll off ferries. Franz, the public doesn't necessarily really hear much about fires on Ro-Ro Ferries. But the International Maritime Organization has decided that more needs to be done to prevent them from breaking out.
00:02:39:14 - 00:02:42:04
Abigail Acton
Can you tell us a little bit more about the context, please?
00:02:42:06 - 00:03:05:03
Franz Evegren
Yes, in fact, there have been quite a few fires on this so-called roll on roll off ships or Roro ships, as we refer to them, carrying vehicles. So you roll on and roll off the vehicles. And if you look at the past incidents, there have been about two and a half major fires on these Roro ships every year.
00:03:05:05 - 00:03:32:23
Franz Evegren
You might have heard of a couple of them in the Mediterranean. There have been a few significant fires on the Norman Atlantic and the Sorrento, but there have been quite a few fires. And this was picked up by the International Maritime Organization, which is a United Nations specialized agency, more like a global parliament for ship safety, including all the nations and a number of the recognized organizations internationally.
00:03:33:00 - 00:03:40:11
Franz Evegren
So they decided that something needs to be done, and they also open up for regulations to change. And that's where our research really came in.
00:03:40:16 - 00:03:57:00
Abigail Acton
Right. Super. Thanks very much for that explanation. Okay. That's interesting. So clearly there was there is a need to to work out what can be improved. So what are some of the potential solutions that the project identified relating to the issues that you've just talked about here, about the incidents of fire?
00:03:57:03 - 00:04:21:15
Franz Evegren
So yeah, in our our project is based on a number of previous projects where we identified a number of challenges that have caused these fires. We looked into past accidents, of course, and we also looked into the future on the coming changes in like the reformation of the vehicle fleet into a larger use of batteries, for example, or hydrogen as fuel.
00:04:21:17 - 00:04:56:21
Franz Evegren
And the issues came down to a list of 20 challenges, which we grouped into partly the four stages of a fire, if you may say so prevention, detection, extinguishment and containment. And we also had two work packages focusing one, focusing on the operations so the people and what they do on board firefighting, for example, and also on the design, the design interface between the humans and the technologies, which has often been a cause of confusion and costs larger consequences of fires, you might say.
00:04:57:00 - 00:05:03:19
Abigail Acton
Can you give me some examples of what you just mentioned there about the human interface with the technology of where that has gone wrong in the past?
00:05:03:24 - 00:05:28:09
Franz Evegren
Absolutely. And so, for example, it might be the case that when you get an alarm on the bridge, it says that the fire has occurred by frame this and deck that in the starboard part of this and then you might have a CCTV camera with a numbering and then you have if there is an established fire or a confirmed fire, you need to activate the extinguishing system in a section.
00:05:28:12 - 00:05:45:22
Franz Evegren
And all these systems are individual and causes or gives a lot of ground for mistakes. You can say so. And even if when you come into the drenched room where you activate the direction sprinkler system, there might be a different numbering on those valves compared to the sections on the deck, so to say.
00:05:46:01 - 00:05:51:11
Abigail Acton
And all of this under a time of great stress and urgency, way, way, really, you want things to be smooth flowing.
00:05:51:11 - 00:05:58:03
Franz Evegren
Right? Right. And in a panic or a very stressful situation. Yeah. Things need to be easy to make checked.
00:05:58:05 - 00:06:14:18
Abigail Acton
Yes, exactly. Not trying to struggle to find the right answer. Okay. That's a great example. And you talked about the likely move towards electric batteries for vehicles and stuff like that. The hydrogen you mentioned as well. So what were some of the solutions that you were thinking with regards, for example, to anticipation of fire?
00:06:14:20 - 00:06:41:21
Franz Evegren
Well, we have looked into detection of the thermal runaway of a battery in an electric vehicle as well. That, I would say, is the area which has not come so far. We weren't addressing that in the project, but we were looking into it a little bit now. And in a lot of aspects of the project, we have actually transformed a little bit because of the large need coming from the operators and from the crews asking how to handle this new situation.
00:06:41:23 - 00:07:03:16
Franz Evegren
What we have done is we have looked into how to safely charge a vehicle on board. We have looked into whether or not the current extinguishing systems would be sufficient to control an electric vehicle fire, for example, and things like that. So we have adapted parts of the project to really consider the future reformation of the vehicle fleet.
00:07:03:18 - 00:07:10:14
Abigail Acton
And as you say, the request from the crew, because I know that you're very interested in having the feedback from crew for sure, from the actual seafarers.
00:07:10:14 - 00:07:30:07
Franz Evegren
Right. And like I said, many of these challenges that we are addressing, they really come from the crew and the concerns and the hazards that they see. And from the like I said, the different fires, the incidents in the past. And we really want to make sure that our research is not becoming a shelf warmer, but really comes to.
00:07:30:09 - 00:07:44:08
Abigail Acton
Yeah, not just academic work but has practical, practical solutions of the project. Yeah. Frans what sort of cutting edge technology are you using, for example, for fire detection? I know that you mentioned perhaps looking under the vehicles, right?
00:07:44:10 - 00:08:06:17
Franz Evegren
So we thought it would be very exciting to look into the potential of using drones. So we've done that in different ways. We've done it, for example, you might be aware of that. Drones are now used for maneuver board exercises. So you can send out a drone to if a man falls overboard to locate the person. And we also use it for getting an overview of the fire.
00:08:06:23 - 00:08:27:09
Franz Evegren
But we also use not flying drones, but rolling drones on the deck that can roll underneath the vehicles. And with the heat cameras and the sniffers or smoke detectors and etc.. And and they can be quite well-equipped because you don't need so many and they can kind of control the area when no one is on board any way.
00:08:27:11 - 00:08:29:05
Abigail Acton
And maybe areas that are inaccessible.
00:08:29:07 - 00:08:37:18
Franz Evegren
Right. And in particular, since the batteries are usually located very low in the vehicle, it makes sense to kind of be underneath the vehicle when you scan them.
00:08:37:19 - 00:08:44:22
Abigail Acton
Yeah. And I guess you could have them sort of like caterpillars going along most of the time throughout the voyage, really, and just kind of doing tours.
00:08:45:00 - 00:08:56:24
Franz Evegren
Right, Right. Yeah. So that's what we're looking into. And we've developed a pilot or a kind of an example of such a drone can be developed and, and used onboard and it's quite exciting to see it work.
00:08:57:01 - 00:09:12:01
Abigail Acton
Yeah, I know it must be and very satisfying and also very satisfying for the seafarers to feel that they're actually being heard by someone in a position to, to come up with some solutions. Okay. Right. Any other kind of solutions or problems that were identified that you are beginning to think of ways of handling them now?
00:09:12:03 - 00:09:33:22
Franz Evegren
Yeah, we are trying to use a quite high so called technology readiness level. So so we want to really make make solutions that can have an impact right away, not not core research, you might say. One of the things that we have looked at is screening all the cargo before it goes on board to look for hotspots on board.
00:09:33:24 - 00:09:50:02
Franz Evegren
These are systems that exist for other valuable infrastructure, such as tunnels, and we thought it would be very useful to also do it for the ships. So developed the system there, and that is now being demonstrated for some of the operators that we work with.
00:09:50:04 - 00:10:05:00
Abigail Acton
So if I can just stop you for a minute, France, I'm really interested in what you've just said about the notion of something that can scan the contents of vehicles to see. I suppose if something is potentially a flashpoint, could be a risk for fire hazard that exists already for going into tunnels you mentioned, right?
00:10:05:00 - 00:10:28:16
Franz Evegren
Yeah, there are companies focusing on this. And so several or a couple of kilometers before the tunnel, usually they scan the vehicle. So that they well in advance can set these vehicles aside, which have ignition sources or seem to be overheated. And combined with AI, these systems are quite smart. They can really see if it's a part of a vehicle that should be hot or if it shouldn't.
00:10:28:16 - 00:10:37:03
Abigail Acton
So fascinating. Yeah. And of course, it's really clear what you say there. There's clearly another process that could be useful there for for for ships and and boats coming aboard.
00:10:37:05 - 00:11:00:10
Franz Evegren
Right. And then we have also looked into extinguishing systems on the weather deck. For example, the weather deck is is the deck which usually see from outside the ship where you can see the actual cargo, the vehicles on top. So they're totally exposed to to wind and to to the weather. And previously they haven't really been and there haven't been any extinguishing systems or detection systems for these decks.
00:11:00:10 - 00:11:06:11
Franz Evegren
So now we're really trying to identify find systems that would work in such a harsh environment. Yeah.
00:11:06:11 - 00:11:13:22
Abigail Acton
I mean, as soon as you mentioned it, of course, it must be very difficult because you usually think of sprinklers and things like that as trenches coming from above. And if there's no roof.
00:11:13:24 - 00:11:15:12
Franz Evegren
Right. You don't have a ceiling. Yeah.
00:11:15:12 - 00:11:21:21
Abigail Acton
Yeah. And a fascinating. Thank you so much. This is really, really interesting. Franz, why did you get involved in this research? What was it that pulled you into this area?
00:11:21:24 - 00:11:54:14
Franz Evegren
Well, I've always been kind of drawn to the sea, I guess. And I knew I wanted to become a firesafer engineer and a risk analyst. So? So I thought that would be a great combination. And I also sailed for a few years and and I really know how it feels to be alone at sea. And when you have felt the big waves of the Biscay or actually was in one experience I had at the North Sea when when our ship was breaking down a little bit.
00:11:54:14 - 00:12:18:05
Franz Evegren
So, so we had to search emergency harbor. And I really felt small and far away from help. And and even if you're on a larger ship, I think that the resources are very limited. And in case there is a fire on board, you really need to be able to manage it on your own on the ship. And so that kind of drew me in to this area and I hope to make an impact and to make it make it safe from them.
00:12:18:10 - 00:12:36:05
Abigail Acton
I'm sure you will. It sounds already as if some of the ideas you were having are, and the ability to give seafarers a voice as well are definitely going to be making a difference, if not already. Thank you very much, Franz. I'm going to turn now to Lazaros. Lazaros SafePASS wanted to redefine evacuation processes, systems and equipment.
00:12:36:07 - 00:12:58:15
Abigail Acton
Your project looked at a better ways of monitoring hazardous situations on board and informing passengers and crew about the safest, quickest evacuation routes. So we've been talking here with Franz about shipboard fires and how to prevent them. But Lazaros, if a fire does break out, that can't be controlled, can you tell us what's involved in an evacuation in current circumstances, please?
00:12:58:17 - 00:13:26:13
Lazaros Karagiannidis
Yes. So once the situation is assessed and the passengers are not safe, anymore on board and evacuation will be, of course, initiated. And this involves the announcement of an evacuation signal. The announcement through the public announcement system to the passenger to move to a safe place, which is called Master Station. And in this process, the crew actually facilitates and guides the passengers to the master stations.
00:13:26:15 - 00:14:01:12
Lazaros Karagiannidis
And then to their obligation stations. But this whole process currently is a quite static process process meaning that passengers need to move to their pre allocated and designated master stations and they have to follow the static, let's say, exit signs. When all exit signs. And they have to wear their lifejackets and so on. And the common problem with this situation is that there might be situations where passengers are located far away from their designated master stations.
00:14:01:14 - 00:14:30:00
Lazaros Karagiannidis
Secondly, there might be evolving hazards like fires or flooding that may block existing evacuation routes or even the whole master station. So we need solutions that actually can, first of all, adapt to the evolving nature of hazards and adapt a bit to the evacuation strategy and also introduce novel means of evacuating passengers not only relying, you know, by the instructions by the crew.
00:14:30:01 - 00:14:50:05
Abigail Acton
Okay, that's clear. And indeed, I mean, any of us who've been on ferries, we've all seen those master signs. And clearly, you know, if you're at the other end of the ship getting a coffee, it's not necessarily the most practical place to go back to. So who did you work with in order to come up with a definition of what needed to be changed?
00:14:50:07 - 00:14:57:00
Abigail Acton
And once you realized what needed to be changed, what kind of ideas did you have to improve the situation?
00:14:57:03 - 00:15:24:02
Lazaros Karagiannidis
Yes. First of all, in our consortium, we have the chance to collaborate with different disciplines and backgrounds. For example, with academia, with shipyards, with cruise line operators, with ICT software providers. But even with lifeboat manufacturers, as well as organizations that are dealing with regulations and also independent ethics and personal data experts.
00:15:24:04 - 00:15:34:14
Abigail Acton
let me stop you for a second. I can see the logic of all the various groups that you're talking about there. But the personal data experts, this is fascinating. Why would they be brought on board? What's the relevance of of data protection there?
00:15:34:17 - 00:16:00:05
Lazaros Karagiannidis
Yes, because the solutions we are implementing also deal with personal data such as the ID of a passenger, such as the location of a passenger. So in order to be compliant with the current GDPR regulations, I mean, the purpose of data protection, we had to ensure that we are really compliant with those regulations and we need to take into account these our solutions, all these regulations.
00:16:00:07 - 00:16:12:00
Abigail Acton
So you've got this really kind of diverse team of of experts, but practical and legislative. Excellent. And what were some of the ideas that you came up with with the help of these these diverse people?
00:16:12:04 - 00:16:47:23
Lazaros Karagiannidis
Just some of the ideas, and especially from the shipyard and the cruise line operators was to, first of all, identify the mission and operational requirements during an evacuation. So we found out that the static application process is not that efficient. And this could be or let's say, an area of modification. So we came up with the idea of a dynamic application and let's say process, which means that we actually know the location of the passengers at any time using localization systems, their mobile devices and the smart life jackets that we have implemented.
00:16:47:24 - 00:17:16:05
Lazaros Karagiannidis
We can calculate the safest and fastest evacuation route based on a current situation on board, especially when we are having fires or flooding that can block some routes. We can calculate this safest evacuation route in real time and then inform the passengers through their different mobile devices or the dynamic exit signs. In contrast to the, let's say, static exit signs.
00:17:16:07 - 00:17:23:01
Abigail Acton
So back in a second. So by dynamic exit signs, we mean basically exit signs that would illuminate depending on a context.
00:17:23:03 - 00:17:36:02
Lazaros Karagiannidis
Exactly right. So these are digital displays that in case there is a blocked route, it will indicate that you should not follow this route. You should follow another route, another corridor, or to go to your muster station. Exactly.
00:17:36:04 - 00:17:46:13
Abigail Acton
And I believe you also considered what would happen in the case of a horrible idea this. But I mean, you know, I guess it does happen. Unfortunately, in the case of limited visibility where people can't see things very clearly.
00:17:46:17 - 00:18:16:15
Lazaros Karagiannidis
Yes. This was also one of the major aspects we had to look in. So we came up and came up with an ideal for smart life jackets. So in case you have a limited visibility, you can activate haptic navigation process, which means you can feel vibrations on your lifejacket that indicate the direction you have to follow or even activate and audio navigation through a speaker which tells you which direction you have to follow.
00:18:16:15 - 00:18:30:07
Lazaros Karagiannidis
And all this information is processed at the, let's say, control central processing platform that shares this information to the lifejacket, but also the mobile devices and so on.
00:18:30:11 - 00:18:48:16
Abigail Acton
And there's absolutely wonderful you're already making me feel safer. Both you and Franz are making me feel safer about my future ferry trips. And I use ferries a lot, so I'm a nervous flier. So I like boats. So, Lazaros, So can you tell me a little bit about training? Because this is all when we're in the middle of a really tough situation.
00:18:48:22 - 00:18:53:03
Abigail Acton
But I believe you've also done some anticipatory work in the form of training.
00:18:53:05 - 00:19:22:20
Lazaros Karagiannidis
Yes, exactly. And the main, let's say, need was to provide some novel training means, especially for the handling and operation of the lifeboats which are deployed on the on the cruise ships. So we implemented an augmented reality application that provides intuitive and easy to use, let's say, air glasses and playback, different kinds of training scenarios as well as maintenance.
00:19:22:20 - 00:19:35:04
Lazaros Karagiannidis
And our resource, the crew is, let's say, better trained and has a more a better experience, a better training experience through these devices.
00:19:35:04 - 00:19:47:01
Abigail Acton
So they can have an understanding of the impacts of the choices they make in a particular context and what that actually means in terms of how easy it is to deploy the lifeboat or and if it's deployed at the right angle.
00:19:47:03 - 00:20:07:24
Lazaros Karagiannidis
Exactly. And also, the good thing is that this type of training can actually be conducted at any time and at any place. So you don't need actually actual, let's say, a lifeboat or you don't need to be on board to make these trainings. And this will it was well appreciated by the trainers.
00:20:08:01 - 00:20:18:24
Abigail Acton
As well appreciated by me as well. That's excellent. Thank you. Very good. This is wonderful. Can I open this up and ask, Does anyone have any observations to make to Lazaros or any questions?
00:20:19:01 - 00:20:20:20
Franz Evegren
Yeah, yeah, I have a question. Yes.
00:20:20:20 - 00:20:22:00
Abigail Acton
Franz, please.
00:20:22:02 - 00:20:55:17
Franz Evegren
I was wondering also, by the way, as I understood it, you look at how you could alter the route to the master station or assembly station in a smarter way, and based on that, you know, the actual location or local positioning of the passenger. Have you also looked at maybe optimizing instead the master station designation for the different passengers based on the location of the passengers so that you go more to the closest master station instead?
00:20:55:20 - 00:21:30:04
Lazaros Karagiannidis
Yes, this was actually exactly the principle or the process. We we followed, of course, taking into account always the aim or regulations and to the specifications of the master stations. But knowing the exact position of the passengers, we could actually guide them to the closest and safest master station, of course. And in some cases even we did not even go through the master mustering process and directed them directly to the embarkation stations, which reduces very much the evacuation time.
00:21:30:06 - 00:21:59:19
Franz Evegren
It's a very interesting subject. I think we actually used a similar principle in the last ferry project where we there's something called a runner or someone gets an alarm to to go to the place of the alarm to confirm whether or not there is a fire. And most of the times it's not. It's a false alarm. And instead of having one designated person to do that, we send an alarm to the closest person, maybe the three closest crew members who are trained to investigate whether or not it's a fire.
00:21:59:21 - 00:22:03:23
Franz Evegren
So with the same principle that you know where all the crew members are.
00:22:04:00 - 00:22:05:20
Abigail Acton
It's about maximizing efficiency.
00:22:05:22 - 00:22:09:06
Lazaros Karagiannidis
Right? The complementarities between the projects.
00:22:09:08 - 00:22:29:21
Abigail Acton
absolutely. And I think you two have a of a lot of overlaps as well. I mean, I think it's it's fascinating. Thank you so much. I'm going to turn to Luis now. Luis, onboard safety also depends on highly accurate location positioning. So we've talked about what would happen on board with the fire and how we would deal with that if we, you know, unfortunately, it did actually occur.
00:22:29:23 - 00:22:49:18
Abigail Acton
But PREPARE Ships your project is exploring the way artificial intelligence, advanced sensing and Earth observation, data all interconnect not only to determine where a ship is now with greater accuracy and you're doing it with very great accuracy. I look forward to talking about that. But also where the ship will be, which is an important element for autonomous vessels.
00:22:49:24 - 00:22:53:07
Abigail Acton
So can you tell us how current systems work and what the drawbacks are?
00:22:53:10 - 00:23:15:10
Luis Sanchez-Heres
Well, currently in a ship that you have a, you're basically using GNSS system, which is the global navigation satellite system, the most common one that we talk about it every day is the G.P.S., the Global Positioning System, which is from the United States. But there's other GNSS such as Galileo or Beidou from China. Galileo is from Europe.
00:23:15:12 - 00:23:51:12
Luis Sanchez-Heres
And all these systems, they tell you where you are in the world and then you can pinpoint your location in a map. But this location that you get has a certain accuracy. So, for example, GPS, this has certain accuracy of a couple of meters, but then you can use different or extra systems to get higher accuracy. For example, you can use something called satellite based augmentation service that gives you maybe one meter accuracy, and then you can have other type of corrections like something called Arctic, that it will give you a centimeter corrections or a PPP correction that will give you dosimeter accuracy location.
00:23:51:18 - 00:24:01:05
Abigail Acton
Okay. But at the moment, I believe you think of the positioning systems as being either dynamic or, as you say. Kinematic Can you explain what these two types are?
00:24:01:05 - 00:24:28:05
Luis Sanchez-Heres
Well, one thing is determining where you are. So where you are in a ship, it is determined by usually GPS or visual submissions or radar. That's that. Those are where you are. Where are you going to be in the future? That's a completely different problem. So usually you use your velocity to determine where you're going to be in the future and that you could call it a dead reckoning predictor or a kinematic predictor, like a moving in that direction of moving at ten meters per second.
00:24:28:11 - 00:24:55:11
Luis Sanchez-Heres
Therefore, in a second they're going to be ten meters ahead. That's it. But the alternative to that and it's important for ships is that you, instead of predicting your future position with your current elicited, you can predict your future position with the current forces being exerted on the vessel. And this is very meaningful for vessels because there are so big they have such a large mass and moment of inertia that these forces are acting and they have consequences that are evident forward in the future.
00:24:55:12 - 00:24:59:05
Abigail Acton
So we're talking about things like currents and wind. And when you say Forces.
00:24:59:07 - 00:25:19:15
Luis Sanchez-Heres
and the propeller and the rudder and what it is. Right. I never say of that real vessel when in a simulator you turn the rudder and nothing happens and then you get desperate and you turn it even more. And then suddenly the basin is turning and you lose control because you don't have the right feeling that what you do right now, you're going to see the effects in the future.
00:25:19:17 - 00:25:39:02
Luis Sanchez-Heres
So from the point of view of control, it's actually quite challenging for somebody who not trained. And because of this problem or even for people who are experienced, if you have wind and currents and other vessels around that then, okay, how where am I going to be in one meter time in the current situation? So you have an idea that is challenging.
00:25:39:02 - 00:26:00:05
Luis Sanchez-Heres
Then seeing another vessel that it's turning and looking at its velocity vector and say, okay, where is that vessel going to be in? I made a time that's even more challenging. So in this preparation project, the was one part of it was the predictor system which actually uses information from on board a vessel and machine, learning to predict where that ship is going to be at a certain amount of time.
00:26:00:08 - 00:26:21:19
Abigail Acton
Okay. So let me ask you then, can you break that down a little bit more in terms of how that all comes together? So because I know that you're very interested in the use of artificial intelligence to crunch big data. So where are the sources of the data coming from Earth Observation and as you say, the forces and so on, And how are they coming together for you to to interpret?
00:26:21:21 - 00:26:48:18
Luis Sanchez-Heres
All right. So we have a Galileo, which is the European GNSS and then we also use corrections from from the Galileo system, which school has security system service. And that gives you this corrections. And we get a position in a map says, okay, latitude, longitude, we are here with Nascimento accuracy. Then we use that information and would mix it with information from the radar, from the anemometer, which is a thing that measures that wind speed and so on to calculate forces.
00:26:48:18 - 00:27:11:06
Luis Sanchez-Heres
And we put these forces into a model, a dynamic model. Now, this dynamic model, it's a bunch of equations with some parameters and coefficients. And these coefficients, we have learned them through experience. We record where we have been on where we are, and we use that information to adjust with machine learning these parameters, and then we can make a prediction out of that.
00:27:11:08 - 00:27:27:10
Abigail Acton
That's fascinating. I know that you were doing a pilot in Gothenburg Harbor. Could you tell us a little bit more about that? And your notion obviously, is that you want the artificial intelligence to do the heavy lifting when it comes to the the assessment, all the data and the interpretation of the data, how did that work out in Gothenburg?
00:27:27:10 - 00:27:28:00
Abigail Acton
Tell us more.
00:27:28:03 - 00:27:49:20
Luis Sanchez-Heres
Well, we actually got the three different types of vessels. We got a pilot boat, a yacht, Princeton's yacht, and we used to look for a system in a ropax, which is a railroad that also carries passengers on and there we installed the system. Some we do some situations where we got somewhat close to each other and we got to see the predictions off of the other ships.
00:27:49:20 - 00:28:11:12
Luis Sanchez-Heres
So that's an important component of the system that we developed. My predictions can be shared to all their vessels so they can see where I'm going to be and I can see where they're going to be and then try to evaluate what the people sailing these vessels thoughts about these predictions, how they're useful to resolve situations. When it comes to the machine learning part, a machine learning part that goes inside of this this system that we made.
00:28:11:12 - 00:28:30:20
Luis Sanchez-Heres
And it's the the calculation of these coefficients, but also the monitoring of the predictions. So the moment that the system says, like your predictions are actually not that good at the moment because of whatever reason, it shuts the system of a safety so that the person operating the vessel doesn't rely on predictions that are all reliable.
00:28:31:00 - 00:28:40:08
Abigail Acton
Yeah, no, that's great. So it removes the idea of false confidence and that's excellent. Okay. And how precise and accurate did you manage to go the predictions?
00:28:40:08 - 00:29:00:10
Luis Sanchez-Heres
Yeah well it's it's an here it depends quite a lot on the weather conditions on what kind of shape the larger the bathtub, the more accurate it is. A small vessel like a pilot boat they're planning. And then the models that we use to model the motions of the ship, they're much more complicated when it's a planning vessel.
00:29:00:10 - 00:29:05:13
Luis Sanchez-Heres
But for big ships, I would say that we got quite sufficient accuracy to make the system useful.
00:29:05:16 - 00:29:07:14
Abigail Acton
So what more variables in a smaller ship?
00:29:07:15 - 00:29:27:16
Luis Sanchez-Heres
Well, the smaller ships, as they start planning, Right. So there's two types of you can say that there's two types of vessels like the ones that, that displays the water like big ships and then the ones that really go so far that they lift off and they're kind of like planning on the surface. Okay. But the physics of that is much more complicated than the other one.
00:29:27:16 - 00:29:35:22
Luis Sanchez-Heres
And so coming up with a new a good set of equations to describe that and how the forces are going to move the ship, it's quite difficult.
00:29:35:24 - 00:29:37:10
Abigail Acton
So that's the next phase.
00:29:37:12 - 00:29:40:07
Luis Sanchez-Heres
That's yeah, yes.
00:29:40:09 - 00:30:00:08
Abigail Acton
This is excellent. Thank you so much. Work behind this. It's very complicated. And there's a you know, there's computational work involved and obviously mathematics and so on. It's very intricate. Louis Why did you get drawn into all of this? It sounds simple. With no mathematical ability whatsoever. It sounds quite kind of intense. What attracted you to to this?
00:30:00:10 - 00:30:16:20
Luis Sanchez-Heres
Well, I'm a naval architecture and been naval architect by training, and I like maneuvering. I think it's it's nice to do simulations about where a ship could be. And then suddenly we have all these tools and machine learning neural networks. And then the question is, can we use them?
00:30:17:01 - 00:30:25:01
Abigail Acton
Yeah. And also for the future because of autonomous vessels. Yeah. Can you tell me a little bit more about how that would feed in? I mean, you know, in your own words, why is it important?
00:30:25:05 - 00:30:48:07
Luis Sanchez-Heres
Well, for example, the model that we have, the predictive system that predicts the future motions of a ship, it has a dynamic model of a vessel and an autonomous vessel is going to have is going to require a dynamic model of the vessel. So this could be a component in an abdominal vessel to actually allow for an autonomous navigation system to steer that vessel.
00:30:48:09 - 00:31:06:10
Abigail Acton
Absolutely. I mean, we think of smart motorways with autonomous vehicles. But it's true, of course, the sea is also like a highway. And yeah, so then that that kind of infrastructure for the vehicles to in this case the vessels to be able to sense where they are and let other ships also predict where they're going to be obviously is crucial.
00:31:06:10 - 00:31:07:01
Abigail Acton
Yeah.
00:31:07:03 - 00:31:15:15
Luis Sanchez-Heres
Well it's more about like it's it's about control. Like if you have a computer and you want to control something, usually the computer needs to have understanding of what is it controlling.
00:31:15:17 - 00:31:16:02
Abigail Acton
Yes.
00:31:16:03 - 00:31:35:22
Luis Sanchez-Heres
So then in some sense where he got created in the context of other vessels, it could be like that model that helps the computer to control. So that's one thing. Then in this project, we also have this communications system. So then you could think of two autonomous vessels. They communicate to each other their route, their current motions and where they're likely to be.
00:31:35:22 - 00:31:41:17
Luis Sanchez-Heres
And they can use this information to plan their movements in a better way and avoid incidents. Super.
00:31:41:19 - 00:31:48:19
Abigail Acton
Absolutely. Thank you. Thank you very much. Very interesting. Does anyone have any questions for Louis? Yes, Franz. What would you like to say? Maybe it's an observation.
00:31:48:21 - 00:32:16:21
Franz Evegren
Well, yeah, it's it's more of an observation. In our project, we have looked into the positioning, the most optimal positioning of the ship. So it's not really predicting maybe the movement of the ship, but in case of a rather large Roro ship fire, how to best position the ship to make sure that you, for example, limit fire spread or that you give as good conditions as possible for evacuation.
00:32:16:23 - 00:32:24:21
Franz Evegren
So so I don't know if if any of the research you have done, Louis, do you think any of that could impact like in a fire situation?
00:32:24:23 - 00:32:52:14
Luis Sanchez-Heres
I mean, I guess it is dynamic model if you think of, okay, there's a fire and then you want to position the ship in the best possible way because help is on their way. But you don't want people in the breached start to move the controls because you want them to evacuate. Yeah. So then you could think of an autonomous navigation system that steers that ship just to make it easy for everybody to get out and for fire control or the fire brigade to take care of the situation.
00:32:52:16 - 00:33:09:16
Abigail Acton
But meanwhile, on board Lazaros’ system is letting everybody know whether to turn left or right by buzzing at them. If they're dark and it's confusing and it's all stressful. Look at you guys. You've got us all safely taped up. Thank you very, very much for all your work. It does genuinely make me feel much more positive about the future.
00:33:09:16 - 00:33:13:20
Abigail Acton
Pinkham votes. Thank you. Thanks for your time, guys, and for being with us today.
00:33:13:20 - 00:33:15:08
Franz Evegren
Thank you for having me. Thanks a lot.
00:33:15:11 - 00:33:16:15
Lazaros Karagiannidis
Thank you very much.
00:33:16:16 - 00:33:17:12
Luis Sanchez-Heres
Thank you very much.
00:33:17:12 - 00:33:39:13
Abigail Acton
Yeah, very welcome. That was super. Thank you. if you enjoyed this podcast and are interested in the latest scientific research coming out the EU, have a listen to previous episodes. Follow us on Spotify and Apple Podcasts or whichever podcast provider you favor. Are you curious about what other EU funded projects are doing in the area of maritime safety?
00:33:39:15 - 00:34:00:19
Abigail Acton
The Cordis website will give you an insight into the results of projects funded by Horizon 2020 and Horizon Europe that are working in this area. The website has articles and interviews that explore the results of research being conducted in a very broad range of domains and subjects from nematodes to Neanderthals. There's something there for you. Maybe you're involved in a project or would like to apply for funding.
00:34:00:21 - 00:34:14:22
Abigail Acton
Take a look at what others are doing in your domain. So come and check out the research that's revealing what makes our world tick. We're always happy to hear from you. Drop us a line. Editorial at cordis dot Europa. Dot EU. Until next time.
00:34:14:24 - 00:34:19:08
Unknown
Earlier this week.
L’énergie, les réserves mondiales de céréales, les biens de consommation et bien plus encore: les navires transportent 90 % des produits de base dans le monde, ce qui est remarquable. Comme l’a déclaré(s’ouvre dans une nouvelle fenêtre) le secrétaire général des Nations unies, António Guterres: «Sans les navires, et sans les femmes et les hommes qui y travaillent, les économies s’enliseraient et les gens mourraient de faim.» Les Nations unies ont estimé que 400 000 marins(s’ouvre dans une nouvelle fenêtre) n’ont pas pu rentrer chez eux en septembre 2020, souvent bloqués à bord, la pandémie ayant imposé des interdictions de voyager. Essentiels à la circulation des marchandises dont nous dépendons tous, les marins travaillent souvent dans des environnements dangereux. Et la pandémie a révélé à quel point ils sont vulnérables aux restrictions sociales. À l’occasion de la Journée des gens de mer(s’ouvre dans une nouvelle fenêtre), cet épisode se penche sur la manière dont le financement de la recherche contribue à améliorer la sécurité en mer, tant pour les passagers que pour les travailleurs. Franz Evegren(s’ouvre dans une nouvelle fenêtre) est le directeur du département de sécurité incendie des transports au Research Institutes of Sweden(s’ouvre dans une nouvelle fenêtre). Franz, qui a coordonné le projet LASH FIRE, étudie les moyens de réduire le nombre d’incendies à bord des navires rouliers. Luis Sanchez-Heres(s’ouvre dans une nouvelle fenêtre) travaille également au Research Institutes of Sweden. Il a étudié le potentiel de l’IA et de l’apprentissage automatique pour prédire l’emplacement futur des navires dans son projet PREPARE Ships, ce qui pourrait révolutionner le monde des véhicules autonomes. Lazaros Karagiannidis(s’ouvre dans une nouvelle fenêtre) se concentre sur la façon dont les applications intelligentes et les dispositifs portables peuvent accélérer, et accroître la sécurité, des évacuations de navires. Lazaros est basé à l’Université technique nationale d’Athènes(s’ouvre dans une nouvelle fenêtre) et a été le coordinateur du projet SafePASS.
Vos retours sont les bienvenus!
Si vous avez des commentaires, nous serons toujours heureux que vous nous en fassiez part! Envoyez-nous vos commentaires, questions ou suggestions à l’adresse électronique habituelle, editorial@cordis.europa.eu.