For the second year, the State Department has recognized employees who enable better use of data in the art of diplomacy. Data for diplomacy is both an awards program and a part of the department’s modernization plan. This year’s group award went to people in the Bureau of Conflict and Stabilization Operations for their work in a program called the Conflict Observatory. Federal Drive with Tom Temin got the details from team lead Susan Wolfinbarger.
Susan Wolfinbarger So I am in the Bureau of Conflict and Stabilization Operations, Office of Advanced Analytics. And LOCATE is just a quick way of conveying what we’re trying to do. We’re trying to locate things around the world. So it stands for location based observations for conflict analysis, trends and evidence. And what we do is focus on how we can take advantage of all of the new open source and commercial data feeds and technologies and platforms that are out there to design foreign assistance and overseas programs to help the department become more data driven in the work that we’re doing.
Eric White Gotcha. And so what kind of platforms are you utilizing for this kind of work?
Susan Wolfinbarger Well, we have one flagship program right now, I will say it’s called the Conflict Observatory. And that program is focused on documenting Russia’s war crimes and other abuses in Ukraine. And the backbone of that platform is an ArcGIS enterprise. It’s a platform from a software company called Esri, and they’re the premier geospatial software provider in the world. And so we’re using that system as the backbone for all of the analysis that’s happening as part of the conflict observatory. So thinking about satellite imagery analysis and open source investigations and how we can feed all of these different data products into a common repository so that a whole team has a centralized location to work together and produce analysis.
Eric White And when you say analysis, break it down for us there. I’m thinking geospatial. It’s a satellite imagery and faraway pictures. How is this analysis used to formulate State Department action or policy?
Susan Wolfinbarger Sure. So all of the work, as I said, was commercial or open source. So all of this data can be a whole variety of different types of information. A lot of it is satellite based, and everything that the conflict observatory is doing is remote. So you can do this analysis from anywhere in the world. And the data feeds largely do cover most areas of the world as well. And so thinking about satellite imagery, there are a whole suite of different products that are out there all the way from thermal detections that we’re getting from NASA satellites to very high resolution optical visible imagery through companies like Max R and Planet and Black Sky. And then we’re also using some kind of mid-range imagery from small sats that collect imagery every single day covering most of the surface of the earth. So we’re really thinking about how do we deal with just such a huge amount of information. And that’s just from the satellite imagery that’s not thinking about in the in the context of Ukraine, all of the information that’s being volunteered online on social media, in chat rooms, all of these things. The teams are also pulling that type of information. So it can be photos and videos that they take identifying landmarks and geolocate those to help understand what’s happening on the ground to go alongside perhaps the satellite imagery of a situation. And so we’re really thinking about triangulating with multiple types of information to understand what has happened in specific incidents.
Eric White The conflict in Ukraine is unprecedented for several reasons, but one of them is the amount of help that the you mentioned a lot of the commercial companies and open source places that you’re able to go to for this kind of information. I mean, has it has there ever been something like this where you all are able to take this kind of view of a conflict happening on the ground and have all these sources to pull from?
Susan Wolfinbarger I think we really have hit a new point. I’ve been doing this for more years than I will admit to at this point, both the nonprofit sector before coming to the State Department. So I can remember, when you would be lucky to get one satellite image in a couple of weeks time frame. And now we can get daily satellite imagery. So just on the availability of the commercial imagery front, it’s a huge innovation over the past couple of years for the availability of information that we have at our hands. And then with the proliferation of smartphones and good Internet connections all over the world, people are taking photos, taking videos and sharing that information publicly. And so I think really for what’s happening in Ukraine, it’s really hit a critical juncture for the availability of these different sensing technologies, because, cell phones are sensors just like satellites. And so I think going forward and thinking about potential other things that are happening in the world, we’re going to start seeing this level of documentation happen as more of a matter of course in the in the coming days and years.
Eric White And how crucial has industry’s role been in this? Famously, StarLink has been providing Internet for Ukrainian citizens during the conflict when they lost access to it. But what other ways has your work been greatly improved because commercial companies have stepped up in and in providing these this crucial software and imaging?
Susan Wolfinbarger Well I’ll start out by saying I’m only operating in the fully remote realm for this work. But thinking about all of the commercial innovations that are happening and just the satellite field is a huge innovation. The availability of small sats that are collecting daily imagery is a huge innovation because that level of coverage used to be – I’m going to get a get a little deep diving on the tech for just a moment.
Susan Wolfinbarger Get large areas covered with satellite imagery. We used to have to use data from sources like Landsat that would be 30 meter meters per pixel. So you could see kind of the houses and things like that. But now with some of these daily products, they’re about 3 to 5 meters. So you can easily pick out buildings and monitor things in a way that we haven’t been able to do across very large areas. So it’s really both just the increase of the types of data that’s being collected, but also how broadly it’s being collected. And I think that’s super critical for Ukraine because the things that are happening is over a very large area and you have to think about how you can triage such a large area and so much data. And so I think that that is one of the innovations that the conflict observatory team is really bringing to us. They’re taking advantage of A.I. and machine learning techniques to to really help us go from looking at eastern Ukraine to being able to pinpoint the sites of specific incidents. And that’s really critical because you can’t monitor such a large area manually. I’m thinking about when you’re on Google Earth or Google Maps, you zoom in down to the level of your house and you look at your block. You can’t scroll through that for something the size of eastern Ukraine. And so in addition to the availability of the data, I think it’s also important to think about the processing of that data and how those advances in AI and machine learning are helping small teams cover very large areas.
Eric White Right. And because the data zoning is is only good to you if you have the analytics behind it. So can you just describe a little bit more about that process and what you all have in store for analyzing all this array of data?
Susan Wolfinbarger So I’ll give a example of how we’re taking advantage of all the different data feeds. And it’s a really allowing the teams to quickly go from a whole country’s worth of data down to an individual site. So the Smithsonian Cultural Rescue Initiative is one of the partners on the conflict observatory, and they are monitoring cultural sites for damage. And what they are doing is using multiple different data feeds to get from what we want to know what’s happening to cultural sites across the country, down to identifying damage at specific sites. So the first step in the process is that they’re using, and I mentioned this earlier, NASA data that is actually thermal detections because a lot of things that happen in the conflict zone produce heat, such as burning. And so that data is collected every day. That will help us get a sense of where things are happening. We also have a database of all the cultural sites. And so when you overlay those two pieces of information and you also narrow it down to just urban areas. That can give you a good idea of where you should be looking. And we’re really excited about a new data feed that’s coming in from the partner Planet Scape AI. They are doing automatic damage assessments for buildings across the whole of eastern Ukraine, so they’re using daily satellite imagery and running some AI and machine learning processes on that imagery to automatically find damaged and destroyed buildings across a very large area. So then you add that piece of information into it and you have multiple pieces of information. You have heat signatures, you have damaged buildings that have been detected. We have the locations of the cultural sites. And then that allows the team to then go to high resolution satellite imagery, because that’s a very limited resource comparatively to these other sources of data. And then they can actually zoom in and look in detail at the damage that they’ve identified through these other processes. So it’s really that method of triaging and working through vast amounts of data to get down to those incidents of interest. That’s, I think, one of the biggest innovations that the conflict observatory team has been able to achieve in this program.
Eric White Got it. And so obviously they’re pleased about it. The higher ups, they’re giving you an award. But what have you heard from State Department leaders about how crucial this has been in them formulating policy and responses to what’s going on in Ukraine?
Susan Wolfinbarger Yeah, I think that this program has been hugely helpful. When you have a conflict situation, there’s often just a real lack of credible information out in the public sphere. And so the fact that we have this phenomenal team of subject matter experts that are part of the conflict observatory doing this research and putting out publicly is so critical to our efforts to share information about Russia’s abuses in Ukraine. I think we’ve seen lots of statements from both US leaders and from people and leadership from overseas citing Conflict Observatory reports, particularly one of the recent reports on the relocation of Ukrainian children across Russia. The work that’s been done with these completely remote data sources is really giving the public and policymakers critical information that they can trust and rely on as they’re making decisions and operating in their positions. I think one of the last things that I would say and I’ve hinted about this a couple of times is that the Conflict Observatory right now is focused on Ukraine, but this is a capability that is available globally. All the data feeds almost are available covering the entire world. And what I hope as part of my my role in the LOCATE team is that we can take advantage of what we’ve learned in documenting war crimes in Ukraine and apply that to other conflict related situations. In Ukraine, as I said, we’re only looking at war crimes. But this platform, this technology, these data sources can be used to monitor a whole range of other events like cease fires and other types of events on the ground. So I think and I hope that you’ll be hearing more about conflict observatories in the future.