AIS: Harnessing state of the art maritime solutions to uncover hidden fishing vessel movements in the Pacific
Uncover vessel and buoy behavior using state of the art AIS maritime solutions, AIS Position Validation and Starboard’s platform and expertise.
Fisheries are of critical importance to the Pacific region with many nations working collaboratively to combat IUU fishing and increase traceability. Learn how Starboard Maritime Intelligence and Spire Maritime have joined forces to help tackle this issue in the Pacific Ocean, using real-time AIS, AIS Position Validation, algorithms, and machine learning.
Kelly Rummins (00:10)
Welcome, everyone. I’m just going to give it a minute while everyone joins. I can see lots of attendees are coming in. If you’re just joining us now, we’re just going to wait just a moment, let a few more people join before we kick off. Great. Let’s get started. Welcome. I’m Kelly Rummins. I’m Starboard’s CMO. I’m really pleased to have Iain from Spire here and Joe from Starboard. I’ll ask them both to introduce themselves in a minute. But before I do, I just wanted to talk briefly about why we wanted to co-host this webinar with Spire Maritime and fill you in on who we have attending and do a little bit of housekeeping for the Q&A because we’re keen to get as many questions through as we can. Just quickly, we’ve had a really amazing response to the topic with over 400 participants registered. Many of you, I can see, as everyone joining, are online live tonight. I’m saying tonight because, Joe and myself, as you might be able to tell from my accent, we’re both based in New Zealand, and it’s a little bit later in the day for us, so we both got our coffees here.
But looking through the list of attendees, I can see you’ve joined us from all over the world. Thank you and good morning, good afternoon, and good evening. We’ve got quite a range of participants as well. Participants from regional fisheries management organizations, tech companies, fisheries agencies, national defense, a few financial institutions and financial services as well. Although today we’re definitely going to be focusing on the manoeuvres and behaviors of fishing vessels, both Spire maritime and Starboard will cover other vessel types. Hopefully a fair bit of what we talk about today can be applied to your areas as well. Just quickly, why did we want to do the webinar today? We wanted to share how our and our customers are making the most of AIS data to help combat illegal, unreported and unregulated fishing or IAU fishing. We also wanted to share a few new tools for uncovering the location of vessels that are not reporting their location on AIS, which can help support this. Often if you’ve been in a webinar with us before, we talk about remote sensing for this. But today we want to talk about Spire’s AIS position validation and Starboard’s AIS buoy analysis.
We’re going to share quite a few examples that come from the Pacific, where fisheries contribute well over a billion dollars to the economies of small developing island nations. The challenges in the area, they do range from underreporting of tuna in the Western Central Pacific to less regulated squid fisheries off the Coast of South America, and of course, illegal fishing as well. Finally, just before I ask Iain and Joe to introduce themselves, if you’ve got any questions or comments you would like to make, there’s a Q&A option down the bottom, so please do add those in at any time. Behind the scenes, we’ve got Dave and Fernando from Starboard and also Gabrielle from Spire maritime. Feel free to put a question in any time. We’re also recording the session today, and I’ll send out the recording probably early next week, as soon as it’s ready. So now I’ll go over to Iain. And it’s been great to be with you. And obviously we’ve been working with Spire’s AIS for some time, and it’s been great collaborating with you on this webinar and exploring your AIS position validation, can you tell us a bit about yourself and your work at Spire Maritime?
Iain Goodridge (04:06)
Yeah, sure. Thanks. Hello, everybody. Thank you for taking the time to join us live today. As Kelly mentioned, my name is Iain Goodridge, and I am the Senior Director of Radio Frequency Geolocation Products, which means that I spend a lot of time looking a way we can use our satellites to geolocate radio emissions. So signals such as cooperative signals such as AIS that we’re going to be talking about a lot today, and then signals that are non-cooperative when ships aren’t using AIS, as well as using the AIS to help validate what a ship is actually reporting. So it’s great to be here. I’m really looking forward to some Q&A. And my contact details are also available in the presentation if anyone wants to follow up direct. Thanks, Kelly.
Kelly Rummins (04:48)
Thanks, Iain. And, Joe, can you tell us a bit about yourself and your work at Starboard?
Joe Corbett (04:54)
Yeah. So thanks, Kelly. And hi, everyone. Thanks for being with us here today. So, I’m Joe Corbett. I’m the Chief Data Scientist at Starboard. I’ve been with the company about just over four years now, and my role here is helping to develop and implement some of the machine learning and algorithms that we have in the platform. Yes, I’m really excited to show everyone what Starboard is about and how we’re using AI as data.
Kelly Rummins (05:27)
Thanks, Joe. And just to get started, I mentioned obviously at the start that we’re going to be talking a lot about the Pacific fisheries today. Can you just tell us a little bit more about the region and why fisheries are so important in the region?
Joe Corbett (05:41)
Yeah. So fisheries is really vital to many of the Pacific Island nations. For many of the nations, it’s their key resource, the key natural resource, especially those ones with quite small landmass that tend to have very large ocean, EEZs, and a lot of fisheries resource contained in that. And it’s a big earner for a lot of these countries. And this income isn’t just through fishing and export of fishing, of fish. Fish and catch, but also through granting access to foreign fleets to fish in their EEZs. And so an example of this is the really innovative Vessel Day scheme, which comprises, I think, about nine countries in the Pacific, and it caps the total number of days that can be used for fishing tuna specifically in these countries, EEZs, and then they are allowed to then auction off these days to foreign fishing fleets, trade them amongst each other, which is a really good way of bringing money into the economies and also managing the tuna fisher in a really sustainable way. Okay. And so this monitoring and control and surveillance of fisheries in the Pacific is done by a number of organizations. There is the countries themselves, but also the wider inter-regional organizations such as the Regional Fisheries Management Organizations, which is WCPFC, the Western Central Pacific Fisheries Commission, along with the Forum Fisheries Agency, the FSA, who helps the countries manage these fisheries in their EEZs. This collaboration really helps to work to combat any illegal, unreported, or unregulated fishing that occurs in the Pacific tuna fisheries, which the estimate amounts to about 200,000 tons annually. Any steps to lower that amount is really positive for the Pacific Islands nations.
Kelly Rummins (08:09)
Thanks, Joe. Iain, I think obviously, so Joe talked a bit about the monitoring, control, and surveillance in the area, what are some of the ways that the vessel activity can be monitored using Spire’s data?
Iain Goodridge (08:23)
Yeah, thanks. I think everyone’s pretty familiar with, or should be pretty familiar with, the concept of AIS being used. Obviously, with the mandates requiring it to be used pretty much every vessel now, it’s still an absolute powerhouse data set. I think we see on average, about 270,000 ships a day throughout AIS collection. But it’s been interesting to continue to iterate on that. In the early days of Spire, we were hyper-focused on the satellite collection, which came with its own challenges. You have lower power class B signals, you have areas of the world that are incredibly busy. So it’s extremely hard for the satellites to actually extract one and those messages. So over time, we’ve really invested in alternative ways of getting that data out of those busy areas. And we have invested, obviously, in the terrestrial network, which is more the AIS that a lot of people are familiar with. That’s the land-based receivers that we’ll see at every single port or organization on the Coast. And we’ve also invested in what we refer to as enhanced satellite data, and that’s one of my favorite ones that has come to market. It’s very similar to the concept that we’re all probably familiar of when we’re driving and maybe someone’s got an Apple Maps app or Google Maps app, and it shows you that like, or in red, there’s traffic slowing down. That’s actually other people’s phones that are transmitting that information. And enhanced satellite, the AIS is exactly that. It’s other ships collecting AIS for us and then transmitting it by dedicated data channels so we can cut through that noise and those high traffic zones. And Gabrielle will back me up on this, but there was a moment when we turned that data stream on and we actually saw probably close to 18,000 chips that we’d never seen before in our data, which was one of those light bulb moments. So all that combined together has really helped with the tracking in these areas. Thanks, Kelly.
Kelly Rummins (10:25)
That’s great. Thanks, Iain. So, Joe, do you want to tell us a little bit more, I guess, about how Starwood is making use of this AI data? I can also see a few questions are coming in, so we might see a bit of action in the chat, but I’ll let Gabrielle and O’Near and Dave respond to those.
Joe Corbett (10:46)
Yeah, thanks, Kelly. If we’re looking at the Pacific Fisheries Region, it’s a really large area. Even a single country will often have an EEZ that is quite massive. And they’re often too big to adequately monitor with surface or air-based assets even. This is where satellite-based information like the AIS data or the vessel AIS that Iain was just talking about, where that really comes in and plays such an important role in the surveillance of these fisheries areas. Organizations that are tasked with monitoring the fisheries in these areas use these geolocation services to track where vessels are going through the area and make sure that they’re not… Or that they’re complying with the rules that they’re supposed to. We’re showing AIS here. The other one is the Vessel Management System, which is the private country-specific vessel tracking system. And yeah, and sothe organizations will use a tool to do this, to visualize this and to understand what’s going on. That’s really where Starboard comes in. So Starboard is a maritime domain awareness platform, and we ingest and display in real time millions of AIS points every day for thousands and thousands of vessels. Our philosophy is to take it a bit further than just displaying points, but to apply models and algorithms. And display it all in a really user-friendly, user-centric way so that analysts can go from just looking at points to really understanding what is happening in their area of interest.
Kelly Rummins (12:53)
That’s great. Thank you.
Joe Corbett (12:54)
What I think I’ll do is just give you a bit of a tour through Stablet. This is what we’re looking at now is the Stablet platform. What you’re seeing here is the last seven days of AIS tracks for all the fishing vessels in this area on Starboard. This is a really amazing view. It really shows you the extent of the fishing activity in the area. But if you were an analyst who was using Starboard or looking to gain some insight, you wouldn’t really look at this view. You’d be focused on a particular area or a region. What we can do is take a look at some specific areas. This could be anything really, an EEZ, a region you supply. But what we’re going to look at here is what’s known as the high-seiz pockets. Disease. These are pockets of the High Seas which are entirely surrounded by countries’ EEZs but are not themselves part of an EEZ. This means that they are free to fish in, so any vessel can go in there if it’s under the jurisdiction of the WCPFC, for instance. It can fish, and those countries that surround those and those that surround these pockets will not receive any compensation for this.
That just means that they’re not getting the compensation. It’s harder to enforce catch limits and just a bit harder to monitor what’s going on. And so what we’re going to do is just zoom in on this little pocket over here, which is surrounded by the EEZs of French Polynesia, Kiribati, and the Cook Islands. And what you can see is a lot of fishing activity has been happening in the last week or so. Each of these tracks is for a different vessel. You can see there’s this big, Taiwanese fleet in this pocket that is undertaking what looks a lot like long-line fishing. Yeah, definitely. What was really interesting or what stood out to me when I was looking at this was these couple of Chinese fleets up here and over here. This is the Cook Islands and this is Kira Bas. I was wondering what was going on there. If we just change our filter just to see what is on the screen, and give it a second to load, what we’ll see is that there is a big fleet of Taiwanese ships in the High Seas pocket, and then there’s two fleets of Chinese ships, one up here in the Cook Islands, EEZ, and one over here in the Kiribati, EEZ.
What will have happened is that the Cook Islands and Kiribati will have licensed… Will provide a license to these ships to fish in their EEZ in exchange for money, presumably.
Kelly Rummins (16:43)
Sorry, just on the screen there, I can see Stover’s different-colored tracks. Some are white, some pink tracks, there’s a couple of dotted lines. Can you explain to me a little bit about what those tracks mean and how that relates to some of the algorithms for the vessel movements?
Joe Corbett (17:01)
Yeah, sure thing. What we’ll do is we’ll just select one vessel and explore it in a bit more detail. This is the man-y-feng, number 16, and this is a Taiwanese long-line. What we’re looking at here is the track that is generated by the AIS points that we ingest. We can also just turn on the actual points. Here you can see each of the individual AIS points that we receive that then go on to form the track. You’ll notice these are a purplish color that indicates that they came from a satellite-based AIS receiver. This would have been picked up by the satellites and sent to us for ingestion to our platform. Okay. Excuse me. Just turned it off. When we ingest this data, we apply a number of different algorithms and models to the data in order to extract information and provide it to the analysts. In the case of fishing vessels, we apply a machine learning based fishing classifier. This is the same fishing classifier that Global Fishing Watch uses. The distinction is that we run it every hour, so we’re updating the fishing results in near real time. You can see on the track, this is indicated by this pink region here.
If I just scrub through, you can see what this vessel is doing. This will be a long-lying vessel. Down here, it’s putting the line into the water. Then it gets to the end here, lets it soak, and then it goes back again and picks it up, pulls it in, takes the fish off, and then heads off. Another great feature of Starboard that we’ve got is that we detect buoys that are transmitting on AIS. When we get a message that comes in, AIS message, we determine whether or not it comes from a vessel or comes from a buoy. What we can do is then associate buoys with vessels. If I just click on this link. Here we have that same vessel, but I’ve also selected some of the buoys that fish with it. A lot of long-line vessels, especially, will use buoys attached to the line, presumably, so they know where to go to pick it up again. If we just zoom in here for a bit and then scroll through. You can see what is happening. The vessel is dropping off the buoys, as it puts the line in the water. Then when it comes back, it comes back and picks up the buoys again and continues on its way.
This is a fairly consistent thing that we see with a lot of long-line vessels. You can see how the vessels travel with the buoys, travel with the vessels over time. What’s quite neat and quite an interesting feature is if you turn off the main vessel, you can still see the tracks of the buoys, and it still makes quite a clear fishing signal. You can actually track the vessel’s movements, even if it’s not transmitting on AIS.
Kelly Rummins (21:06)
Joe, with something like that, being able to see the vessel even when it’s not transmitting on AIS, I can see it’s close to an EEZ there on the edge. If the vessel wasn’t transmitting and the buoy was over in the EEZ, would that be a way to detect an incursion?
Joe Corbett (21:24)
Yes, it would be. This has actually been used as evidence in the court prosecution in the Pacific, and we can’t talk about that. But what we can do is show you another example. It’s actually over in the Indian Ocean. I’m just going to jump over to the Indian Ocean for a second. Just to get you orientate yourself, we’re just in the Indian Ocean here, just down from Sri Lanka. We’ve got selected here a vessel, Imula 0926. This vessel appears to have a number of buoys associated with it. There’s two ways we can tell that. One is that they travel with the vessel. But these ones also have quite a similar name to the vessel. In fact, they’re just the vessel’s name with a suffix attached to them. If we zoom out and we just play that through time, you can see how the buoys follow the vessel around. Where the vessel goes, the buoys go as well. What’s really interesting about this is that when we come down here, the vessel is down here doing some fishing activity. Then at this point here, it turns off its AIS. This dashed line indicates when we don’t have AIS from a vessel, but we can connect the two points that do have AAS.
We just represent it as a straight line. But what was really interesting is that this vessel with buoys then pop up down here within the Mauritius EEZ, and then when it goes up and then eventually goes away. What this is quite strongly indicating is that this vessel was down in the Mauritius, AUZ doing some fishing and attempting to hide its behavior by turning off its AIS transmitter.
Kelly Rummins (23:44)
It’s a great example. That’s likely to be a long-liner as well, using those long-line voice. Before, when you were sharing the example on the Pacific, there was a vessel subtype and it said that it was a long-liner, and it also showed some registrations. This doesn’t look like it has those. But can you tell us a bit more about how analysts might use that information for monitoring, control, and surveillance?
Joe Corbett (24:11)
Yeah. What we’ve done is we’ve gone through, I think it is 15 of the regional fisheries and management organization, the Missile Registries, throughout the world. We’ve taken that data and then matched it to our AIS data. Okay. This has a number of good uses. One is that it lets us get more detail about that vessel that we wouldn’t otherwise have from just self-reporting. For instance, fishing longline that comes from the RFMO registry, not from the AIS. But we can also then match that up with the registry records. For instance, you can then see the authorization period for a vessel. This vessel is authorized in the WCPFC through to February 2025, and it’s authorized to fish for tuna. And it’s authorized to transship on the high seas. So if you’re an analyst, if you’re investigating a ship or you’re looking to see what a vessel is up to and whether or not it’s allowed to be fishing in those areas, then this is a really helpful way to have all that data in the same place, in the same location, and allow you to really efficiently do your job. And just linking this back to the RUU, the illegal, unreported, unregulated fishing we mentioned earlier.
If a vessel is fishing in a region, it’s not supposed to or is not authorized to, then that would be considered the illegal part of the IUU phishing. The other quite neat thing over in this panel here is this journey summary. If we click on this, we can see the activity of the vessel since its last port visit. It’s been out fishing for the last 62 days since it visited Pago Pago Harbor in American Samoan. It’s had about 34 days, 35 days of fishing. It’s had one encounter, and this is quite interesting, it’s had one likely trans-shipment with one vessel. The Encounters algorithm is another algorithm that we apply to data as we ingest into our system. That’s really looking to see when two vessels are close to each other for a defined period of time. We’ve started taking that one step further and trying to classify the encounters into different types. The likely transshipments are our first attempt at that, and that is when a fishing vessel meets with a fish carrier specifically, and is encountering them for a duration of, I think, at least three hours, which is generally enough time to at least transship a bit of fish.
If I click on this transshipment event, it will take us to the location. Just zooming in a bit, you can see here where the fishing vessel, Mani Feng, has met up with the fish carrier, the Lien-Jie-Seang. If I just scrub through that a bit, you can see how they’re moving in pretty much concert until they get to a point where they split up and head off. That is quite likely to be a transient event. If we just link this back to the WCPFC record here, you can see that this is authorized to train ship in the High Sea, so it is allowed to be doing that. I can also just zoom out a bit more and you can see how this works in a wider context. The fishing vessel initially leaves America and Samoa and goes over and starts to fish in this high sea pocket, then heads up north and meets up with the fish carrier. And this fish carrier has come down from Taiwan. Over here, it’s come down. And these little icons indicates where it’s had encounters or transshipments with other vessels. And so it’s come down and met up with a number of vessels.
And then it goes up and heads back up to Taiwan, where it presumably offload as the catch. This is just quite a nearly neat way of seeing how these distant water fleets operate with the transshipment and the resupply. And if you were interested in, say, tracing the supply chain of these fish, then this gives you an idea of where a particular fish might have been caught in the Pacific or before it ends up in the supply chain. It’s also quite useful if you will say a fisheries officer at a port that was under the port state measures, and then you could use this data to verify the self-reported trans-shipment events that fish carriers are supposed to report when they come into those ports.
Kelly Rummins (30:20)
That’s great. Thanks, Joe. I think we’re going to come back a bit later to chat about the squid fisheries as well in the Pacific. But Iain, we’ve been talking quite a lot about AI data in general. What are the different geolocation target services and how does this align with Spire’s Maritimes main AI services?
Iain Goodridge (30:42)
Yeah, that’s a great question. This actually all started about two and a half, maybe three years ago. We were challenged by a government agency to have an opinion or a way to come up with a accuracy or not an accuracy, like a cell anywhere in the world and actually have a likelihood that we would be able to pick up a message there. We could destroy a 200 by 200 kilometer block on somewhere in the world’s oceans. And we needed to come up with a way to say we’re 98 % or 99 % confident that if a ship’s in that area and it’s transmitting, we would be able to capture it through one of our capture methods. And that led into what we refer to as position validation, which was the to be able to take a message from a ship and ignore the actual contents of the message, so not decode it in any way, not worry about what’s actually in it, and then use Doppler geolocation to validate where that transmission came from. You can only do that when you have a large amount of satellites. Obviously, you need multiple of these messages to be able to do that Doppler geolocation.
We also wanted it to be near real time. It’s really important. There’s a lot of services out there that are after the fact. We can do all kinds of analysis and go back through and say, Oh, yes, that fishing vessel did cross into that EEZ, but we know that a day later or two days later. This position validation service runs globally about every 25 minutes, and that’s the data that we can now feed out to give you those alerts. I do believe you got a slide that shows the different types of alert it’s looking for. And It know we’ve got a good example as well. I’m ready to go for Joe, which is worth sharing. But let’s take a look at the slides here. Yeah, great. Thanks. This slide just talks a little bit about the Doppler effect. Best way to think about Doppler is next time you’re out on the street and you see an emergency vehicle, you hear that siren. And just take a second to think about how different that siren sounds as it comes towards you, and then as it passes past you and it goes away from you, you’re actually, now I’ve said that, next time you listen, you’ll be like, it is different.
It is. There’s a frequency difference there. And that’s the best way to think about how Doppler works. So these AIS messages are being transmitted. We pick those up from our satellites, and we know how long it should take to get there. Obviously, the speed of light is constant, so we know literally how long that radio signal should have taken to get there. And then when we have enough of those, we’re able to run it through the algorithms and then geolocate that. It’s actually a lot of… When we work through it, we noticed that there was three main issues that we were detecting. And I think, Kelly, the next slide has those on there. We can talk about those for a second. We have What we refer to is good vessel behavior. That’s the top left image there where we have a reported position. So if you decode the message, the lat long in that message is a pretty good track to estimate a position, which is the cyan color there, the blue. You can see the dots hovering around it. The other three things here are the ones that we’re going to touch on in the demo, probably the 181.91, I’ll get to that in a second.
The out-of-footprint, though, is one that I think quite a few people in the call may have come across before. It’s actually one of the favorite spoofing techniques that’s out there. And for us, it’s actually one of the ones that comes to the top quite easy. Out of footprint literally means you have a ship transmitting a lat long that is impossible for a satellite to be picking that up because the satellite footprint is nowhere near that lat long. And that’s one of the first triggers that we have that there’s an issue there. But what’s interesting, and you can see that on the image there, is the ship is reporting this position up the Coast of heading north up there, the yellow dots heading up the Coast of Africa. And it actually is heading northwest. And it’s really fun when you dig into that data there because you’ll see that the spoofed track is a very consistent speed, maybe seven knots, maybe eight knots, little heading changes, very little adjustments. And then the actual estimated Doppler location, so we’re actually creating that location through the independent source, independent service, the ship is flat out. I mean, it’s going as fast as it possibly can. It’s ever gone before. I mean, it’s touching its maximum knots. It literally is racing to go and do its nefarious behavior. And then what’s really interesting is when you see these play out, the ship will actually catch back up with its reported position. So you see this beautiful concert where position, real position, and then they catch back up with the spoof, and then they go on their merry way. That’s the data that’s coming out of this API in near real time. The other two to briefly cover those, a fixed GPS is where the ship is saying it’s not moving, and it is. We do see that with fishing vessels. And then 181, 91, all that literally means is the GNSS receiver that’s on the ship, however it’s connected to the AIS equipment, has been disconnected or been interfered with. The AIS is still active. It’s transmitting on those VHF frequencies. This is popular because the fishing vessel then benefits from showing up to all the ships that are around it. And most importantly, the fishing vessel that’s doing this can also see what’s around it, and it can go back and get its buoys because it is still active on its AIS.
You could also say that there’s a requirement there that the logs show that the AIS was never turned off, right? We always get questions about how do you know if AIS is turned off. A lot of these fishing ships, we’ll try and on because they know that’s being checked. But what the 181.91 means is we don’t have a valid GPS position and the ship would normally show up out of range, we’re still able to locate that one. And that’s the one we see quite a amount with fishing because it’s a really good way of quickly pulling a plug out, so to speak, and then wandering off into an easy and then wandering back without any red flags. I think actually that ties in well to a demo you’ve got, Joe, where you’re going to show that over in Argentina. Is that correct?
Joe Corbett (37:01)
Yes. Bring that up.
Yes, we’ve got a couple of examples, but yeah, one is exactly that. In Starboard, you can see that the AIS is missing, but the vessel position validation is showing a bit more information.
Yeah, I’ll just give a bit of context here if you want, Anne. This is just off the Coast of Argentina, and the Lureong 1U, 186 is one of the square jiggers that frequently populate this area. From a starboard point of view, this dashed line here indicates where we don’t have an AIS position, but we’re just creating a straight, interpolated line between the last or between two points that we do have. What is interesting is that in this time period that the data is missing, we were able to find a number of these position validation points provided by Spire.
Kelly Rummins (38:33)
Iain, do you want to describe a little bit more? I guess this will be the one anyway.
Iain Goodridge (38:40)
Yeah, exactly. This is what I was referring to. We have seen this before. I’m a very close to an easy. Again, the position validation service works on a global collect and then analysis, which gives it that 25-minute delivery there. We’re not going to ever see perfect track lines from that. We’re going to see a lot of estimated positions as we absorb those Doppler transmissions and then collect that out there. But you can see here, and the timestamps line up here on this example, where that ship actually has technically lost its position. What that means is it’s gone out of range. The 181.91 is the lat long. Often a lot of systems will exclude that as not a good message and things like that. It literally disappears. But at the exact same time that was happening, we were able to pick up these Doppler transmissions out there to the EEZ. And this would have happened in near old time as well. So it’s a great example where if there’s other assets that could be in the area to validate that, obviously on the Spire side, we’d obviously take this data to go a little further and potentially look at maritime VHF or radar use there to see if we could come up with another location in that EEZ. This has happened, I’ve probably seen this 10 or 20 times a week just from getting into the data. It’s quite a popular technique. Thanks. It’s a great demo.
Joe Corbett (40:12)
Awesome. I think we’ve got another one which shares a little bit about the similar missing, but also an example of the outer footprint as well, and definitely an example of a vessel that’s been spoofing as well, which you talked about before.
Iain Goodridge (40:26)
Yeah, well, Joe gets that up, I mean, the outer footprint is one that we see regularly, usually 10 to 15, sometimes 20 ships a day are showing up on our alerts as out of footprint, often following the same route, which is the run over the Venezuelan area, right on the Atlantic there. Although recently, within the last two weeks, we actually have seen a ship fake its position all the way into the Gulf of Mexico, which interesting coming close to US waters and things like that. It was interesting one to track. We have a lot of partners at Spire, as you can imagine, in the space community. We’re able to get fast access to tasking and we’re able to collect images over these reported positions and then validate there’s actually no ship there. It’s interesting to just pull back a SAR image or a photographic image and say, Yep, there’s a position. There’s absolutely no ship in this area. Go ahead, Joe. Sorry I talked a little bit over you there.
Joe Corbett (41:29)
No, that’s fine. This is the Arman 114. Now this is a bit of a jump from the fishing. This is actually an Iranian oil tanker. I think in this example, it was transmitting its position to be up here or near the Suez Canal. But the Spire PV picked it up transit in across the Indian Ocean area here. So this, I presume, would have been one of the outer footprint detections.
Iain Goodridge (42:06)
Correct. Yeah, the satellites have about a thousand kilometer footprint, just in a circle for the antenna. So think about that range when you’re looking at those outer footprints.
Joe Corbett (42:16)
Yeah. And so this one’s interesting because we’re not shown in this view, but in July, it reappears on AIS and was, I believe, apprehended by the Indonesian authorities for illegal oil transfers. It was quite clearly up to no good, and the position of validation is a really cool tool to spot vessels that are trying to hide.
Kelly Rummins (42:55)
We’re getting pretty close to time. I think we’ve had a lot of questions coming in through the chat, which is awesome. I’m just going to quickly ask you to give us a bit of a teaser about what’s coming up for Spire, and then I might grab a couple of questions that have come through to get you to answer them live. We will probably run maybe five minutes over. Apologies to anyone if you do have to head off. But again, I will share the recording probably early next week. But yes, Iain, can you let us know a bit about what’s coming up?
Iain Goodridge (43:27)
Yeah, thanks. Obviously on the position validation side, there’s engineering work happening to reduce the latency of that data run and also improve the accuracy that’s going to be ongoing. We’re actually also looking at onboard processing of that data. Right now it’s a transmit, receive, download process. We’re able to put that onto our satellites to actually run on board the satellites and then download that finished result. That obviously will really speed up the time there. That’s moving away from the cooperative signal. So AIS is a cooperative signal. It’s something that we’re open to collect. Moving to non-cooperative signals such as maritime, VHF radio, and SACOM devices, Spire is obviously heavily invested in delivering services there that could match that latency as well. So getting out of the AS world for a second and getting into the other types of radio frequency that are out there. In the US, Spire is under contract with the NRO to deliver that type of data as well. And as you can imagine, it would fall into all those usual maritime friendly radio frequencies, VHF, KNAKU bands, satellites, radar and things like that. Good. We can take some questions now, Kelly, if you want.
Kelly Rummins (44:44)
Yeah, that sounds good. Joe, is there any questions there in the chat that’s come through that you want to answer live as well? I Can see quite a lot around, I guess, AI’s data and looking for when there’s gaps. I don’t know if there’s anything there.
Joe Corbett (45:07)
If you can find me a good question, I can’t really…
Kelly Rummins (45:10)
There’s quite a lot to go through.
Joe Corbett (45:12)
Kelly Rummins (45:18)
I think there’s a question here about what countries and organizations that are currently live monitoring their AIS:ON. Is anyone using this data in real time? Looking to see what’s happening and intervening either at sea or at port?
Joe Corbett (45:39)
Yeah, there is a number of countries, especially across the Pacific. I think Marshall Islands is a good example of a country that is using AIS and using Starboard in particular, but they have really taken to using it to see what activity is happening in their territorial and EEZ waters.
Kelly Rummins (46:09)
I think they also just recently won an award that IMCS network for their work in illegal fishing, stopping illegal fishing, which is excellent. And I think we had a couple of questions come through over email, actually, so they might not have appeared in the chat yet. One was about what are the limitations with the RFGL estimates for this broadcasting an AIS message at the time a satellite is overhead, and what is required for an RFGL estimate to be created? I think you might have answered some of this a little bit in your presentation already, but that was from there?
Iain Goodridge (46:55)
Yeah, no, great question. Currently we need about five transmissions from a ship in order to trigger the Doppler geolocation. And we need that to go to one satellite. That’s the current limitation to the service. As the satellite is passing and receiving those signals, if we get those five, we can calculate that position. Obviously looking to improve that and obviously looking to be able to cut through more of the heavy trafficked areas. Right now, this is a satellite-based service. So talk about the early, when we started the presentation, we talked about the issues with really busy parts of the world. Although, oddly enough, a lot of the nefarious activity actually happens in the open ocean. Those two things are working with each other there. We do actually attribute MMSIs. Obviously, we can still decode those messages if we choose to, but we’re able to do that on the process level. So if we see a shift that’s acting normally and then it starts to go out of footprint, we’ll grab those MMSIs there and tie those two together. I think that answers the question that came in.
Kelly Rummins (48:10)
Yeah, that was an interesting one about being able to the MMSI numbers. Yeah, that’s fair.
Iain Goodridge (48:15)
I think there’s a longer conversation there about fingerprinting that transmission, and obviously there’s ongoing work to bring those services out too.
Kelly Rummins (48:27)
Awesome. Thanks, Iain. Joe, I don’t know if you want to answer maybe one more question before we head too far, but I can see everyone… Most attendees have stayed, which is awesome. Thank you all for continuing on past our allotted time.
Yeah, there’s a question. I’m interested in tracking the vessel using the buoys, and they turn off AIS, can you elaborate a bit more on this? How do you confirm that a buoy belongs to a particular vessel?
Joe Corbett (49:00)
At the moment, we don’t attach buoys and vessels together. You have to look at the tracks manually and work out which buoys might belong to a vessel. And it’s usually pretty obvious because they don’t switch vessels very often, and the buoys will stay with the vessel for a very long time. So you’ll see that behavior where the vessel comes along, drops buoys off, picks it up, and then moves on to the next fishing ground, and those same buoys go with that vessel. You have to look over a bit of a time period to make sure that those buoys belong to that vessel. But it becomes pretty obvious when you look at it in some detail.
Kelly Rummins (49:48)
That’s awesome. Thanks, Joe. And I don’t know if there’s anything else you wanted to add to the live, but otherwise, I think we might wrap up.
Joe Corbett (49:58)
I’m good. Thank you, guys.
Kelly Rummins (50:00)
Awesome. Thank you both so much. I think we covered a lot. There’s a lot of information there. As I said, we will send out the recording. It’s been an absolute pleasure working with you, Iain, and of course, Joe, as I often do, and also Gabrielle at Spire behind the scenes as well. Thank you very much. Also thank you to Dave and Fer for answering all those questions today. Thank you for the attendees. It’s been a great topic and it’s been really enjoyable to host this. We’ll sign off. Thank you.