You're listening to a stage talk titled, Using the Open Source Munitions Portal. This week we spoke to Joe Dyke and Joe Emmett from Airwalls and Nick R. Jensen -Jones from Armant Research Services. Together they work on the Open Source Munitions Portal, a database of different munition types for open source research analysis. We spoke on ways to use the tool, their latest updates and why resources like theirs are important.
This talk was hosted by me, Charlotte Ma, on Thursday the 13th of March 2025 in the Bell and Cactus Discord server. Today we're talking about weaponry, specifically munitions identification. I'm joined on stage by the team behind the open source munitions portal. The portal is a free to use tool that aims to provide a key resource to democratize information about conflict and open up access to understanding munitions use. Bellingcat and our wider volunteer team use it
frequently. The OSMP offers nearly a thousand reference images of a variety of munitions from active conflict zones, each reviewed and assessed by at least two munition experts. Key information is provided about each image, including reported location, tentative model of munition, and where relevant analyst notes written by experts. You can also find guides to munition types, explainer videos, and a glossary of key terms on site.
Basically, they have everything you would need to know about munition types on their website. Recently, the team launched the first of a series of articles using 3D interactive models to explain how to identify some of the most widely used munitions, which we will hopefully dive into today. The site has agreements with a number of other research organizations, including ourselves at Bellingcat, Human Rights Watch and Myanmar Witness, allowing access and re -publication
of their archive of munitions images. The team next to me is Joe Deit, director of programs at Air Wars and a co -founder of the OSMP. Nikar Jensen -Jones is also a co -founder and an arms and munitions specialist and a director of armament and research services. And Joe Emmert, a researcher for Air Wars with the ears of open source intelligence experience, focusing on conflict zones and munitions use in the Middle East and Ukraine. If you'd like to ask a question, please do so in the chat
as we talk. And please note, as I just mentioned within your question, if you do not want me to read your username out, because this is being audio recorded and will be posted on podcast platforms following this session. Okay. I'll be back with the questions in a second, particularly those that you ask in the chat. But first, let me pass the mic to the OSMP team so we can learn a little bit more about this fantastic resource. Guys, take it away. Okay. Well, yeah, thank you
first of all for coming. It's really interesting. We're going to try and keep our talks like you asked to 20, 25 minutes. So we do, we will obviously be talking and explaining the site, but we know it's a podcast as well. So we'll try and keep it broad enough that If you're not looking at what we're looking at, you can still kind of understand. I'm going to talk briefly about where the idea for the OSMP came from, and particularly
the problem it was trying to solve. Nick's then going to talk briefly about how he and the Ares team designed the kind of verification and... methodology to a certain extent. And Joe then we'll talk about what's actually on the site and some of the new developments, including, as you mentioned, 3D models, et cetera. So as a kind of brief background of where this kind of came from, I worked as a journalist in the
Middle East for a decade. My last job being chief correspondent for the West Bank in Gaza with AFP, which is a big global news agency. I was more of a kind of traditional, no, non -OUSINT journalist. working on conflicts, writing about conflicts, but I never really had the time and accessible resources to learn properly about
munitions. And I know that was a thing from my peer group, people that, you know, other journalists working in their world, that we all kind of felt like it was something we really wanted to learn about, but didn't really have the time or the expertise. And then in 2021, I joined Air Wars. and set up the investigations unit. And we began using open source techniques to investigate civilians
being killed in conflicts across the globe. For those of you aren't aware of what we do, Air Wars is the world's leading open source documenter of civilian harm, assessing civilian harm claims in countries across the conflicts across the world, including Syria, Yemen, Ukraine and Gaza.
We were set up actually the same year as Bellingcat in 2014, and actually a lot of the techniques that Bellingcat has pioneered, geolocation, et cetera, we have built into our systems and the two organizations have overlapped many years. But then when I started doing investigations at Airwalls, essentially I kind of realized after doing a few that I was doing the same process every single time as an OSINT process, where I would kind of gather all images of munition
remnants, craters, et cetera. And then I would send them to one of the five or six munition experts that I knew to review. And kind of over time, I kind of realized this was both inefficient and pretty undemocratic. It was inefficient in that every time I was asking a favor from the same few experts, sometimes they would come back and say, yep, that's a Russian artillery shell model number, blah, blah, blah. And sometimes they'd come back and say, yeah, you've just sent
me some metal. That doesn't really tell me anything. Nick maybe can speak a little bit more to that, but you had a few experts being inundated requests, often multiple journalists sending them the same images about the same incident asking for verification. And it was also kind of like inefficient in that I wasn't actually learning or getting better. I was just doing the same thing and kind of outsourcing
that every time. And it was kind of undemocratic because I work for Air Wars and so we have contacts with munitions experts and those kind of things. The vast majority of journalists do not have that. And that includes journalists in conflict zones, journalists in smaller news organizations, people who are kind of experts in OSINT investigations. So they don't have that network of people that they can go and check information. And so there really was no resource available to fill that
gap. And so a colleague of mine and I basically said, wouldn't it be great to set up an accessible library of munition images that had been reviewed by actual experts? And so that was kind of the idea. And then I called up Nick, who will speak in a second, who I'd known for a little while by that stage. And he is one of the world's leading experts in munitions and was one of the people that I would call up and ask if they could do
me a favor. And it turned out he'd kind of been having the same idea for a number of years at the other end of the telescope. Aries have trained hundreds of journalists and researchers and others in munitions, but at the end of the courses, et cetera, they don't really have easily accessible material to direct them to. Maybe six months down the line, a journalist would be like, oh, I did that really interesting course, but now I've forgotten it all and now I kind of want
to come back to that information. And there was nothing in the public. domain that's like easy and accessible. Images also, you know, referent images tended to be munitions in their entirety, not how people, investigators, journalists, et cetera, find images of munitions in the real world, you know, once they have exploded, you
know, a particular fragment, et cetera. And even as like practitioners, experts like Nick, who can maybe talk a little bit more, but the of reference material meant that like going and saying, okay, I want to check my assumptions. I'm pretty sure that this is this, but I want to just see five other pictures of that just to verify was sometimes unnecessarily time consuming.
So for that, Nick and I basically kind of came up with the idea for the tool, which we described as a tool for journalists, human rights researchers and lawyers who often receive little training in munitions and have few accessible resources. And the aim was not and is still not to turn anybody into a specialist. Munitions, as I've learned over the last few years, are incredibly
complicated. Every time I think I've understood a munition, Nick or somebody will say, well, actually, there's an additional new variant. And you always have to be aware of false assumptions and matches. And we do not say that the tool
can make you a munitions expert. But what we aim to do is to raise everybody's level up a little bit of understanding, a little bit, and to build an archive that would allow you to learn from investigations that have been done over the last 20, 30 years in an easy and accessible fashion. And so we applied and got a grant from the Google News Initiative and now from the European Media and Information Fund. And we built a prototype, which we launched in 2023. a month or so after
the war in Gaza had begun. That was a very kind of janky prototype. So we then spent five months getting in -depth feedback from journalists, lawyers, human rights researchers, and others, specifically about how the tool could be more useful to them, what functionalities that we didn't have, et cetera. built a whole new version. It's not actually the same websites that we started from scratch. And that was eventually launched
in June last year. And we see this tool and its evolution is kind of pushed towards more accurate and responsible reporting on conflict. And we at Air Wars are kind of geared around this mission to focus attention on the full impact of conflict on, particularly on civilians. And yeah, that is... We, the, I'm going to hand over to Nick here, but kind of my last point would be that we would, we have, we still see this as like
a collective tool. So like, we do genuinely want feedback because the things that aren't there that could be really, really useful, uh, to people, uh, where we also want to know those. So yeah, that's, um, that's me and Nick, maybe I'll hand over to you. Yeah, thanks Joe. Um, right. Uh, I think probably, uh, some of you are familiar with. with ARIES, but for those of you who aren't ARIES, Armament Research Services, we call ourselves a specialized technical intelligence consultancy.
And essentially what we do is provide detailed answers on questions about arms and munitions. So everything from proliferation questions, how did that get there? Where might it end up next? To very technical, what we'd call pins and springs questions. How does it function? You know, one
of the key components we can look at. So we came on board in part when Joe and I spoke the first time because we had, as Joe said, received so many requests for the same sort of answers from various customers, if you like, or mostly unpaying customers that really needed sort of fairly basic information from our perspective, but needed it to be verifiable and needed to be confident
in the answers we were giving them. So you can imagine that a group of technical specialists with really niche skill sets and Joe is very kind in what he said about me. I would maybe push back a little and say my greatest strength is that I know a lot of really brilliant experts in very niche fields. We have a guy who, the example I love to give is we had a guy once upon a time who specialized in East German ammunition packaging from the 1970s. That was his focus.
And he would look at the fonts and the kerning and look at when they changed the spacing between the R and the E and this sort of thing. You think, oh, okay. So having access to those people, they're often coming in at a 10, if you like, at full bore. nails, pins, springs, bolts, whatever it is, they're really down the details. And a journalist or an NGO user, or even an independent researcher
might need a five or a six, right? And what's out there on the internet might be a one or a two, or it might come across as a five or six, but be completely unverifiable. So the OSMP in many ways is about confidence. And what Aries does, I think what we bring to the table for this is giving the user the confidence in what they're reading. So you've got to have the really high level as you say Charlie, niche expert sources, but you also need to have some of those generalist
introductory sources. And ideally you need to have something in between. And that was the bit that Joe and I really felt was missing. We thought you could get some introductory training from people like us and others. You get some really competent technical specialty work from, again, us and others, but you really had a hard time if you wanted to get on -ramp into some of the sort of medium sized questions about arms and
munitions. So Aries in contributing to the OSMP, a large part of what we do is the verification process. So we contribute in some ways to identifying important information that might go into the tool, but really what we're doing is verification. So we're checking those images, making sure that they're of sufficient quality, that the object, the munition can be classified, and then formally assessing them through a process that sees it pass through at least the hands of two different
ARIES reviewers. So sometimes it's three. If something's really specialized, it might be four. But generally speaking, if you see something on the OSMP, Two ARIES technical specialists have looked at it and they've made a tentative identification wherever possible. We make a distinction between identifying and classifying. And those are two kinds of operations that are really useful in arms and munitions investigations. Sometimes people need to look at weapons at a much more
granular level. They need to know for attribution purposes, for example, the make and model and the manufacturer, what country it came from. For forensic purposes, you need to go even more granular and it might need to match a projectile to a... gun barrel, you might need to match a powder residue to a rocket launcher blast, for
example. But other times you need to look at things at a much more macro level and you need to understand, you know, what types of weapons they are in broad categories or how many air delivered munitions were used in this conflict, how many guided munitions, whatever it might be. So OSMP tries to sort of give people both sets of data where possible. But as Joe said, we're not pretending that we can put a tool on the internet that makes you into a specialist
with decades of experience. Instead, we're helping people walk themselves down that path and then when they can make an informed decision about what kind of information they might need next. So if you know you only need that sort of macro level classification type information, then maybe you don't need to go and pay a lot of money to consult someone like me to tell you something
you don't need to know. On the other hand, if you go down the path of understanding the weapons in a particular investigation that you're interested in, and those arms and munitions, turn out to be really complicated or you need to interpret them in a way that requires specialist expertise, then you at least have a starting point to go to someone and have a conversation with them and explain what you need and what you've already done. So that was the gap we thought needed filling.
The verification process, like I say, is I think where we put our best foot forward so that we give people confidence. When you look at the OSMP, you're confident that, you know, there's a, there's a small likelihood that there's an error in there. We're fairly confident. You know, obviously it comes with all the usual caveats. We're doing this from open source material. We're
doing this from a distance. In most cases, some of these we've, you know, there might've been a physical interaction, but in almost all cases, these are done from a distance from test -based research. So there are all the typical caveats for identification, but at least the level of confidence is hopefully higher for the average user than it would be if they'd stumbled across
the photos themselves. I think that's probably the gist of it in the time that we have, but Joe can talk a little bit about the process that each entry goes through. Cool. Okay. Yeah. Hi everyone. I'm the other Joe. I joined Airwars in March of last year and essentially my job is more at the start of the documentation process.
So I will be the one sifting through social media, so telegram pages, Twitter accounts, Facebook groups to essentially find the actual munition remnants and then add them to the portal with the initial tags before they go to the technical specialists from Ares. I kind of want to start by taking you through the evolution of the OSMP. So where we've come from and where we are now.
Essentially, as Joe said, we initially launched in beta in November of 2023, and then we relaunched last year in June 2024 with a brand new design. And throughout this process, we've sort of been aiming to add somewhere around 50 new munition images each month. And that's allowed us to expand from around 230 to now about 860 munition images. And within these munitions, they cover around 200. different models and types from around two dozen different locations, regions, countries
across the world. And throughout this, we've also been trying to expand the categories of munitions. So not just air delivered bombs or rockets or missiles. We've now been adding tank gun projectiles, our recoilless rifle projectiles, and most recently drones or UAVs such as the Iranian shayad 131 and 136 or the Israeli harap that was used by Azerbaijan quite a lot in Nagorno
-Karabakh. We've also recently begun to sign agreements with certain organizations, one of them being Bellingcat as well as Human Rights Watch and Myanmar Witness to essentially allow us to tap into their world -leading research and investigative practices so we can use their images from the field. We can add them to the portal, add tags and ID them. And a good example would be one from Bellingcat. Look at the research, different research organizations on the portal.
And essentially when you click on each of the images, it will allow you to look at the original source, as well as an external reference link where you can go to the organization's page, talking about the munition. Another The feature we've added to sort of aid understanding of munitions is what we refer to as the Analyst Note feature. So that's when we finish reviewing a munition and Ares technical staff will add information to provide basically greater detail surrounding
a unique munition capability. So for one example, might be a GBU -39, one of these ones we've got from Gaza. And Aries added that this image shows various munition remnants, including a fuse well and two nose cone fragments from GBU -39 small diameter bombs. The presence of two different nose cones indicates that these remnants are
from at least two distinct munitions. So not only can we show you what these pieces of metal essentially are, but we can also say something that you may miss is that we've actually got two nose cones. So it suggests in this instance that there might be two GBU -39. air -delivered
bombs that hit this location. Talking about GB39s, they were widely used in Gaza, which brings me on to my next point, which is how we actually not just document munitions, but try to link them to incidents of civilian harm, that we don't overlook the actual impact of the munitions in conflict zones. Through this process, we essentially usually work with members from the Air Wars team. to match munition debris found in a particular location to a specific civilian casualty incident
that's being investigated. It can also match images obviously we've found from Human Rights Watch and other organisations to their work documenting the munitions as well. And we'll always add an external resource link to these munitions so you can look at the original work essentially
as well. The reason behind adding this is that we think it provides important context, not only of what the munition is, but most likely for people who are investigating say a potential war crime, you can look and see who actually may have deployed the munition, so who has that in their inventory, what platform was likely used. So is it an artillery piece, a tank, a
fighter aircraft, something like that. And in some instances where we have markings or data panels on the fragments, we can actually look at who potentially made that, who the weapons manufacturer of that specific part of the munition was as well. And we, essentially in doing this, we're trying to help the Samoan community gain an understanding of what the telltale signs of the aftermath of a strike may look like for GB 39 or Mark 84 and what to look for in the future.
But I think as Joe mentioned, you've always still got to be very cautious and look around for any false positives in this. In sort of recent months, we've also moved beyond just documenting munition images to actually helping you visualize the munitions. So we've been essentially creating 3D interactive models of some of the most widely used munitions in conflict zones. which a lot of you may have seen with the GBU -39. So essentially it will give you a 3D model with extra information
around it. And you can click on the model and look at the model from various angles, as well as learn about the different parts and match them directly to images in the OSMP to corresponding parts. And in each of these parts, images linked to the certain sections, you can click on them and you can look at the original image, where the source came from, just to get some extra
context around it. And we thought the GBU -39 was a pretty important starting point as it was one of the ones that the New York Times called Israel's weapon of choice in Gaza. We're also looking at more widespread munitions that are going to be used in the future as well. I'm working on some other 3D models and different iterations of how we can basically mess around and change with the way in which the model is visualized and maybe help you compare different munitions
side by side as well. Lastly, I kind of want to talk about how the OSMP has become more reactive due to the nature of the conflicts we document munitions use in. Important incidents occur on short notice in a short period of time and they often generate large amounts of public interest.
And so where there is this particular interest or potentially misinformation, we're going to try and respond as accurately and quickly as possible, essentially to provide as accurate as possible information about the munitions used
in certain instances. So for example, after the strike on Hassan Nasrallah, the Secretary General of Hezbollah in September of last year by the Israeli Air Force, the Israelis actually posted on the Israeli Air Force Twitter account an image of one of the F -15s involved in the strike. And there was a lot of reporting about bunker
busters being used. And we were actually one of the first to be able to identify that the munitions hanging off The aircraft were what we, what were found to be BLU 109 JDAMs, which are essentially bunker buster and what is commonly referred to in the media as bunker busters, confirming that. Another example might be the recent strike on the Chernobyl new safe confinement structure. And there was a reported drone strike on that. And then there was a lot of back and forth claims
about who had done it. Russia said that it wasn't, that they weren't essentially behind the strike and often even accused Ukraine potentially of being behind it. The issue with that is that there were images of debris and remnants from
the scene. And we matched one of these images to a video and some other images from one of the gantries, the walkways at the upper levels of the New South confinement, which showed what we found to be an MD550 motor, which is the engine essentially of Iranian and Russian made Russian variant Shahed 136 UAVs or kamikaze drones as they're often referred to, which would suggest that Russia was most like the one behind the
strike and not Ukraine. And yeah, we've essentially tried to expand that and remain as reactive as possible and responding to it. So we've done other ones when there was a storm shadow strike in Kursk. We were able to identify some elimination remnants and confirm that a British made air -launched cruise missile was used inside Russia for the first time by Ukrainian forces. So yeah, that's kind of how the OSMP came to be and how
it's moved forward. And maybe in a little bit, we can talk about the actual future of where we think it might be going. But yeah, apart from that, we'll be happy to take some questions. Amazing. Thank you so much for doing a little tag team there and really embedding us in the growth of the OSMP. OSMP. It's really interesting to hear. Keep your questions coming in. I've
seen a few pop in already. I wanted to start just by asking though, because this is something that often comes up when you're looking at building a tool or building a resource, particularly for open source research where you find a gap, is the difficulty in getting the resources and the time to keep it maintained and to keep it up to date with the flood of information that comes in. When setting up the OSMP, did you find it at first difficult to get time investment for
documentation? How did you learn to balance resources? For example, Bellingcat has just introduced toolkit maintainers to keep our toolkit up to date and they are doing an incredible job. Before it was very, very difficult for us to keep up with the pace of different tools being launched and to make sure that our toolkit was representative of the latest open source tools out there. So how has the OSMP team been able to manage that process? Yeah, it's an interesting question.
I think when we first set out, we very much said that this was about quantity, not quality. And I think, uh, Nick, uh, and so I think Joe mentioned that we had about 50 images. Now that's not actually, you know, by OSIN standards, that doesn't feel like very much, but they each go through quite a very rigorous process and it takes 50 images per month, you know, is actually a lot of work. And so it's about selecting the images that we
think are the most, uh, useful. Yeah. I mean, the Bellingcat toolkit is such a useful, uh, thing, I guess what, what We're doing, we understand that like the archive has cumulative effects. If you know what I mean, that the bigger it gets, um, over time and you had the same methodology over time, it becomes more useful and more, more
useful. And, but we, we kind of also as, and then, uh, you know, very much you'll notice there's lots of things that aren't on there right now that we, we were like, okay, let's start with. We initially started with full munition categories. We've now added tank projectiles and recoilless weapons, and then some drones. But yeah, there's a lot of stuff. Mines is not on there. There's also things that are not. So we're trying to
do it well. And I think Joe E always says we need a minimum of 10 images of a particular munition. for it to be useful as a reference, right? Cause you kind of want to see it from the same thing from multiple different angles, et cetera. And so, yeah, I think we're not trying to kind of keep up in that sense. And we're not in the same way that Bellingcat is a, like a hub of all of
these methodology. Like what we're trying to do is like build kind of slightly methodically a really reliable and in -depth archive, I think. Yeah, exactly. That's amazing. We also spoke previous to when setting up this call about the fact that you also allow people to submit images to the site for verification. Can you talk a little bit about that submission process, what you're looking for and if there's any caveats to content that you can upload for verification?
Yeah, sure. I'll take this one. In terms of It's open to anyone to submit an image. We do say generally they should be a decent quality image. So it should be fairly up close to the munition
or the munition remnant should be visible. And ultimately you can go to the website and you can, anyone can upload them so long as they have a corresponding source image as well that we can add and we can reference because part of our sort of verification process, which I was going to get into is that We try and look for the most reliable sources that we can take munitions from and get as close to the original source as possible. And then we put it through TinEye
and Google Lens. We reverse image search it to essentially make sure that we're getting as close to the original as possible. And we're checking the veracity essentially of the image and making sure that it's as accurate as possible before we even start the process of going through the methodology behind classifying it. But it is open to anyone. Anyone can submit. It sends it directly to us. We get an email each time and then we can review it and potentially upload
it. That being said, you do have quite a large PSA on your site about not approaching munitions under any circumstances. You mentioned that you get quite a lot of imagery from organizations like Bellingcat and things like that. Where are most of your images coming from? How do you make sure that people aren't taking unnecessary risks to help document munitions on the ground, for example? Well, I can answer the first bit and then maybe Nick is best equipped to answer the
second, the last part of that. Essentially, most of our imagery is coming... maybe predominantly from social media. So different accounts, we search in Arabic language or Facebook groups. We often save them and archive them so we can go back because for example, the Palestinian police force in Gaza have an explosive ordnance disposal team and we're often looking and scouring through that. But it's a whole range of sources. I wouldn't say we try and balance it out, but
it's predominantly social media. And then we're also in or signing more and more potential agreements with other organizations to use their imagery that they've actually got people in the field who are going there and taking photographs of the munition. Very often in the past, we have some images from, I think, Human Rights Watch of US Atacans, which is a ballistic missile with cluster munitions from Iraq in, I think, 2000.
three or four. So a lot of our images date quite far back as well, especially when we take them from other organisations. But we're trying to essentially get a wide range of sources to do that. In terms of approaching munitions, I think Nick maybe could talk a bit more about the dangers of that. Yeah, sure. I mean, I don't think there's too much that needs to be said. You know, I'll give the obligatory safety warning, which we have on the page, which is don't do it unless
you're unless you're trained to do so. As we partner with different organisations, Obviously we want to have confidence that they're not putting their employees or their sources in positions of risk, but ultimately you can only go so far down the chain. So I think the remedy to this potential threat to sources or third order effects that we might not be able to visualize is just promoting good safety practices around arms and
munitions. So we use the arms acronym, which is avoid the area, record all the information, mark the area from a safe distance to warn other people. And then seek assistance from the relevant authorities, whoever that might be in a place. So at the simplest form, that just means if you see a munition and you happen to be in the field, you stay away from it. You talk to someone competent and it's really that simple. Very sound advice.
And Mike, the spelling expert asked earlier, would you ever look at creating a similar database for full weapons systems, effectively having someone watch a military parade in Russia and Uzbekistan and identifying the vehicle scene, for example? Is that something you've ever thought about? Is that kind of resource out there that you're aware of? Joe, do you want me to tackle this or you want to chime in? We would have thought
about that. Yes. We love the idea. I think, and Joe probably has some words to say, but in general terms, there's a lot of things we could do with this platform. We have a big document that we all are very excited about. Of course, it's all about time and money at the end of the day. how we prioritize those different avenues for future development, I think is an ongoing discussion.
But as Joe the Younger mentioned, we've already added recoilless projectiles, for example, and we're adding new types of munitions that we're looking at and we're adding more detail on those that are already in there. Going forward, yeah, I mean, in a perfect world, I'd love to see us tackle all arms and munitions and then maybe
even look at platforms. Yeah, just really to reiterate, yeah, like it was kind of trying to bite off what we can chew relatively slowly in history, in the sense of we'd rather feel like we built a really strong database of this. of specifically the ammunition types that we're looking at. And then, but yeah, like Nick and I have talked about small arms, there's lots and lots of worlds in which this would be incredibly
useful. And the main thing that we did with the second version of the portal was we made it kind of fully scalable, like the structure is scalable. And so that was actually uh, some of the data stuff that, but like it involves some of like, um, stripping back some of the things that we had coded for previously and making it a bit less so that like, yeah, if at some point we wanted to add in mines or we want to add in small arms, yeah, all of that is doable within the
portal. And like, um, I think that would be a really useful potential, uh, thing that, you know, over time would be of use to people. There you go, Mike. There's appetite for it. Uh, That being said, while we're talking about resources, we mentioned that people can submit. Is there an opportunity for people to apply or offer to help with the verification process? Is that kind of lopped away to ARIES or is Air Wars volunteers
also helping with the verification process? Is there any way that open source researchers can apply to be part of the verification process at OSMP? Might be another one for both of us, but I think I'd say, you know, not to be too proprietary because the point of this, of course, is, as Joe said, to democratize information and get some of this out there, not behind a paywall where it often is. I think the real value in what we're doing lies in the confidence we give
users. I want people to be able to look at ammunition and know that they have confidence in how that's gone through the process. So certainly there are avenues for people to join Ares and Air Wars. And we're always looking for more talented people with a specific background and skill set. in identifying arms and munitions in our case, and in other kinds of open source research as well
in Airwar's case. But equally, I think, you know, it's important that if people, you know, attach to this project or end up working on the OSMP, they've come from a really strong background that gives confidence to the user base. I think that's what's more important than the volume. A lot of the tools, I had a conversation with this, with a... open source intelligence specialist about this not too longer. And he was talking to me about the sort of different waves of OSINT
tools that have come out. And in wave one, a lot of it was all about capturing data, crawling, getting as much material as possible. And the sort of second and third waves that we're looking at now, it's more about refining and sorting through the immense amount of data that's out there. So I think we're quite good at what we
do within the framework that we do it. As Joe said, that's a little bit slower than maybe the adrenaline rush you get from jumping on Twitter and firing off an ID if you think you know something. But what it gives people at the other end of that pipeline is that confident result that they can use in a court case or a human rights investigation or a really important article for the New York Times or CNN or whatever. So that to me is the
focus. I think where we might expand a little bit in the future, and this is probably in line with what you're asking, is in getting a bigger funnel at the front. So getting more data in for us to process. So time and money permitting,
that would be really cool. Yeah, exactly. I think the like, you know, I've been doing this a while now and I can most of the time I can say, okay Well, I think that's a hellfire right and but like I but I don't think that's actually of use to anyone in the like a there's a million, you know a number of variants that I might not have understood and like what you really want is like somebody like Nick or some of the people are from the Aries team who have like Real expertise
and to say okay, that's definitively a hell file as you know, you know, there's still tentative but you know that like the value of the tool is that like the people who are verifying it are not You know, it's not a it's not quite like a community value like what we are trying to look into and I think we talked about this a lot is like trying to make the tool more community focused. And we've actually been really open
to ideas about that and how we do that. Like how, you know, whether it be more like questions around, okay, well, the community, you know, the thing that people keep saying we're missing is this and some kind of process. by having that,
I think might be more, more useful. Um, but yeah, like, I think the, the, like Nick said, the value is in having certainty that, uh, yeah, that somebody, you know, the methodology that like two people who've been trained in expert in munitions, uh, identification have reviewed that image rather than, yeah. Um, so it's a slightly different process, but yeah. You know, I might just take one more thing on the end there to say. that we really want this to be a useful tool. I think
Joe said this right at the very beginning. We want this to be something people use. We're not doing this as a vanity project. No one's getting rich off this. So if this is something that is informing your work as a professional or even as an enthusiast, if there are ways in which the tool can do a better job of informing your work and supporting your work, we'd like to hear about it. That for us is probably the biggest community drive at the moment is how can we make
this tool more useful? Part of that's going to be educating people about it, doing talks like this one. Part of it's going to be sitting down with some groups of specialized professions. So sitting down with some journalists, sitting down with some human rights researchers and saying, how do you use this? How could you integrate it into your workflow? Does it actually help? Like does integrating this into your workflow make your day slower and not help you? If so,
we need to change something. So yeah, we're definitely very open to hearing from people about how we can improve what we're doing. The very next simple question comes is how do people give you that feedback? How do people get in touch with you? If they do have comments or success stories or things that they feel like could be improved. Yeah. Well, I mean, there's a, um, an email, which I should know off by heart, but is in the about page. I think it's just info at OSMP. rsmp
at airwars .org Which yeah, and to be honest
like We're quite open on social media. Please follow us on the like we've actually had some really good feedback on Twitter and Like blue sky and People and she's saying oh like even even like yeah, really really interesting feedback and saying like saying a couple of people say, oh, I found this image here, that look, you know, like when something happens, being able to use the OSMP as an immediate reference and then like linking it on social media, that kind of thing
is really, really helpful. And so yeah, like I think we're pretty open to that. There's an email that Joe, the Joe is in charge of, but yeah, we'd be very open to kind of social media, blue sky, Twitter, et cetera. Amazing. I've just linked the blue sky in the chat. I did the X account earlier. And Izzy Bakehoven has just also covented the OSMP at airwas .org as the main place to email. So thank you so much for that. There's another question from Mike that
I want to throw in. Is there a way of seeing known users? So for example, if someone finds a shell that they can... they can see the countries that are known to use that type of munition. You mentioned, for example, you can see the different regions in the tool. Is there a way of seeing different munitions that are used across regions? I suspect, uh, Joe Emmett might want to say something here as well, but I'll, I'll jump in by saying that's much, much, much harder than it seems
at first blush. I think, yeah, that would be lovely. What I'd love to have, again, in a sort of perfect world is a really deep dive data page for each of the munition models of at least those of particular interest, which included, amongst other things, information on known users. But
that is quite challenging to verify. So because we're, as Joe the Elder said, trying to keep to verified content and giving people confidence in those results, we are... probably being a little more stringent than maybe some other sites or resources in really fact -checking everything that goes onto those, onto those model descriptions and pages. So it's possible. I think it would be incredibly helpful, but I think it would be a surprisingly large and time -consuming job.
Yeah. I'll just hop in on that. Yeah. It's, and Joe, sorry I passed you in a second, but yeah, it's something that we've talked about a lot. And it's not like, You can do things like, you know, you can search by location and specific munition and that filters automatically. It gets very complicated very quickly. I mean, yeah, if you take something like a Grad rocket, like who hasn't used a Grad rocket at this point,
right? And so like it's easier in the context of Gaza, for example, where it's fairly clear. who is conducting airstrikes there's the app, but in, in other contexts, it becomes very complicated, very quickly. So it's something we're trying to work through. It's not currently available and it, uh, it's just something that like we would like to do longer term, I think. Um, yeah,
but it's about doing it right. And, and maybe Joe, you've got some, uh, yeah, I just, uh, think re -upping what you guys have said that, um, with certain munitions due to the broad sway of actors that might be both state actors and non -state actors that might be using a munition, it won't always be necessarily as helpful as well in a conflict zone because the same munition might have been used by both sides or in the case of when there's multiple parties involved
in a conflict, all of them could be using the munition. So it won't always be that helpful. If, for example, there's an Israeli airstrike in Syria and you have an Israeli made ammunition, it might be helpful in that sense. We have discussed it a lot, I think, as the other said, but it might be something for the future. But again, it's just, it's really hard to pin down essentially sometimes. Okay. We've got about 10 minutes left. So if you have a question, get your question
coming in now. Quickly going back to one of the updates that you mentioned you were doing was kind of this faster response, talking about ongoing conflicts. And you gave a few examples, Jo, of, for example, the Chernobyl attack. Have you seen an increase of disinformation impacting your work a little bit with the weapons identification? For example, Bellingcat last year published a
really fun article called Oh Shit. Seven deadly sins of open source research because we saw, you know, a trend of people claiming to do open source research and not following example rules that we would want people to follow. Have you seen a similar trend in UNITION's identification on social media, particularly in ongoing conflicts or fast moving conflicts? And how are you dealing with that within your own reporting on social
media? trying to kind of get the word out there that you've kind of identified it successfully. That fills that one for Nick, I think. Yeah, I've got some thoughts. You know, I'm going to sort of, right, we're in an environment, we're in an information environment that's becoming in many ways increasingly more challenging. I think people know that. People are used to, you know, some of the disinformation campaigns that
we've seen in recent years in particular. However, I think I will say that On the whole, since I've been poking around Twitter and online communities looking at arms and munitions identification, i .e. some time, we've got better as a whole. I think the level of knowledge has actually risen. I think people are less often wrong in a very public way. I know when I founded Aries, that
was 2013. A couple of our very earliest clients were some NGOs and in one case a government that had made embarrassing public mistakes in misidentifying munitions. I mean embarrassing the way that actually affected the conclusions of their report or their claims about attribution in particular. So bad stuff. When some organizations, NGOs and other organizations that work in the open source space got started, there were some errors pretty publicly
in some of their work as well. And a lot of that has been based around the fact that these groups often start small and with a tremendous willingness to assist and jump to conclusions. So I think as a whole, the OSINT community has matured. It's got better at not jumping to conclusions. It's got better at sort of self reflection and self verification and the tools that we have to try and combat disinformation have also improved. So I actually will strike an optimistic tone,
I think, in answering that question. Amazing. It's actually nice to hear some optimism when it comes to the range of disinformation out there. So that's great. I wanted to also ask you a quick question about the 3D models, because obviously that's a new thing that you've introduced and you've spoken about it a little bit. Why was that decision made that the 3D models was the
way to go forward? And what process goes into creating them to ensure accuracy when you're describing the different parts of each model? Yeah, sure. So we were thinking of the best way to essentially help visualize ammunition. And we thought an interactive 3D model was probably the best way because you can view every single angle. We can break it into parts and we can match those parts to real world examples. And then we could add extra detail along the side
with it. And essentially the process is we have a really good a 3D modeler called Azul with Air Wars. And I will work with her and sort of initial stage. We'll discuss it with Nick and Joe about what is the next munitions. We have a few in the pipeline mainly to do with drones. We might be focusing on in the coming months. And what we aim to do is to essentially draw from technical documents. actually the arms manufacturers pages and their images of the munition, as well as
dimensions. So on some of the future ones, we might potentially add dimensions to the munitions so you can get the correct measurements. And we thought, yeah, a 3D model essentially is the best way to go to help people understand the munition. Cause often we'll have images of different parts of the munition and maybe one image of Say a Gb39 hanging from or a market E4 bomb hanging from underneath fighter jet in one angle. But you wouldn't always be able to see the full thing.
So matching these images with a model will help you sort of understand what the bomb looks like, particularly how the munition looks like, particularly before it's actually delivered. So before it impacts the target essentially. Yeah. And I would just add on that, that it's also like, I think it's really, for me, I found it, I found it really Yeah, useful to understand and visualize but also it's kind of like Taking it beyond like I'll just trust us to munition experts. I've
looked at this right. It's It's like, okay You can you can also understand if like by linking to the specific images and so if you're on the 3d model You can see okay Well here is here are there and then it takes you back to the original images that we've already classified They're in the archive as that particular munition. So it's taking it beyond like Like it's an extra layer of kind of openness and transparency around almost like, you know, here is, here is this
model. Like here are the images from our archive that maps these specific parts and it helps you both understand it. And also like, I think myself that like, yeah, now I, now I see things that I wouldn't have seen before I was working on this project. Um, yeah. I might, if I can quickly also just say that, you know, Joe said at the
very start. that the genesis of this project was in many ways responding to some of those challenges that he had as a journalist and later as a researcher and indeed that I've had as a specialist with journalists coming to me and others. In fact, I can see a few people in the crowd here who field queries from journalists
occasionally. So you'll probably know if you have an area of specialty and people ask you, I don't want to make this sound like we've invented a whole thing to make my life easier, but in many ways it has. We've got a tool now we can point people to that sort of shows them some of the working. You know, oftentimes, as Joe was saying, You know, as a journalist without a specialty expertise in the subject, you're
going to struggle to get repeatability. You're going to struggle to get confidence in the answers. And you can't really check the working if you like, you just get a, you go to an expert and you say, Hey, tell me this. And they say, yep, it's, it's one, two, three. And you go great. And you take it as read. And this at least is just our step towards showing a bit more of the
working out. So you'll see at the bottom of that GVU 39 page, for example, there's resources, there's links to other external sites with even more information and brochures and manufacturers information and that sort of thing. So I'd love us to keep going that way. Obviously they're quite time consuming to make, but if we can keep doing it, then yeah, we'd love to just sort of show some of the workings so people have even more confidence in the findings on OSMP. Amazing.
Yeah, at Bellingcat we've also found 3D models really help with kind of showcasing how we've determined what munition is what. We found it really useful when comparing airstrike models. really quickly squeeze this question. Any plans for expanding the geographical coverage of the datasets? As someone reporting on conflict in Latin America and the Caribbean, this sort of resource would have enormous implications here.
So in the future plans, is there any plans for expanding the geographical coverage of these datasets? Yeah, I think, sorry, I think it goes back to what we said earlier. Like, At the moment, it's restricted to explosive munitions and kind of heavy munitions. And Latin America tends to be a place where there's a lot of small arms, right? That may or may not change if stuff happens
along the Mexican border, but yeah. But at the moment, we would cover Latin America if there were... like the types of munitions being used at scale. At the moment, it tends to be small arms and there is some stuff, but yeah, we definitely wouldn't rule out Latin America. There's also some other gaps. It's not that we've deliberately ignored certain places. It's just that we've focused on the places where there's the... heaviest
use of kind of these kinds of weapons. And yeah, um, I was noticing the other day that we only have one image from Sudan, for example. Now that's a place where there's been quite a lot of explosive munitions. And so at some point I would like to expand that. And so, yeah, there's no, there's no, um, restrictions in that sense. Yeah. Um, and just, just on this, I know a lot of our munition images are currently from Ukraine and the Middle
East and to an extent the Caucasus. But we are starting to add more munitions from especially parts of Africa. So in the pipeline, we have a lot of images of the MAM -L, which is a Turkish made air delivered bomb by the Bayraktar TB2 drone. And that's seen a lot of use in Ethiopia, Libya, as well as parts of Mali and Burkina Faso. So we are starting to expand the geographical sort of range of the munitions we are documenting,
but it does take time. And as Joe said, there are certain parts of the world where it's more small arms rather than the munitions we're currently documenting. But again, we'll have to look at that maybe in the future. Yeah. And just to hop back in briefly before I let Nick speak, sorry, is to say that, yeah, like one of the reasons for that has been the airwars' geographical focus has historically been the Middle East and Ukraine.
And so like there are contexts that we don't know that much about, which is why we're hoping to expand these partnerships. Somewhere like Myanmar, for example, like we didn't, and like Nick can maybe talk about like the munitions used in Myanmar. They're very unusual and different. Um, but like we worked with Myanmar, uh, witness, uh, to build that bit of the archive because like, like it wouldn't be something that we would necessarily be focusing on as an organization.
So like that's kind of where you would like to build a kind of more collaborative tool over time. And yeah. Um, yeah. Sorry, Nick. No, that's all right. You've said essentially what I was going to say, which is, you know, we all have our own backgrounds and geographical biases and the areas that we've worked on. And in many cases, we're trying to get sort of bang for buck in terms of the time and the funds that we have
that are both limited for this project. So we're trying to, you know, look at in some ways the low hanging fruit and gather as much solid materials we can to put through the pipeline and get into the OSMP and into people's hands. If for no other reason than as Joe said before, it's important to scale this up and show the kind of aggregate
value of the archive. But as we push out into the future, and if you're sitting there listening to this and you've got a couple of hundred thousand bucks you want to send our way, that would help. But as we push into the future, obviously we'd love to bring on people with different expertise and partner with organizations with different expertise so we can expand both the geographic coverage, but also the types and munitions and maybe the types of details that we're giving
people as well. Love that little plug there. That was fantastic. Yeah, there's quite a lot of love in the chat. for their plan to expand, particularly in Africa. Subtle Knife says, Sudan, yes. Uri says, I'd love to see more from Mali, Burkina Faso, for example. And somebody else has said that they can help with Sudan. And they'll definitely email open source Neosha's portal now. So thanks so much for that. It's towards the end of the hour. Unfortunately, we do need
to wrap. I can see people popping questions in still. So if the team Doesn't mind kind of staying a little bit just to answer questions in the chat. I'm sure our team would really appreciate that. But thank you so much for joining us today, Joe, Joe and Nick. It's been a pleasure to hear about this incredible project that you've launched and have been working on for so many years now. It's been absolutely vital to so many incredible investigations that have gone on over those past
years. Yeah, thank you so much for investing your time there and for spending some time here. I just wanted to mention that quickly that we have in two weeks time, we're continuing this conversation on munitions and identification. And we're actually going to have a talk from a previous spelling cat fellow called Andrew Matthewson. He actually created a fantastic key guide that can help you identify the most common types of munitions that you might come across,
particularly explosive ordinances. You can find that guide on the Bellingcat website, but we're actually going to have a chat from him in two weeks' time and digging in a little bit more on the regional differences between weaponry and munitions. and what kind of keys and guides you can create yourself or the other community to try and help identify these things. So please join us for that call. But yeah, thank you so much, Doe, Doe and Nick for your time today.
It was a pleasure. Thanks, Charlie. Thank you. Thank you. Thank you so much all. Thank you for listening to the stage talk. If you'd like to catch a stage tour live where you can ask the guest questions, join the Bellingcat Discord server by visiting www .discord .gg slash Bellingcat. The music you've heard is titled Dawn by Newer Self and is courtesy of Artlist.