The best listening experience is on Chrome, Firefox, or Safari. Subscribe to Federal Drive’s daily audio interviews on Apple Podcasts or PodcastOne.
Illegal fishing is bad, not just for fish populations. It can threaten the food supply, the marine ecosystem, and even geopolitical stability. The latter issue is why the Pentagon’s Defense Innovation Unit launched a competition for ways to use technology to detect and stop illegal fishing. For more, Federal Drive with Tom Temin turned to DIU’s chief strategist for artificial intelligence and machine learning, Bryce Goodman.
Tom Temin: Mr. Goodman, nice to have you.
Bryce Goodman: It’s good to be here.
Tom Temin: And I guess I said geopolitical stability, it can be affected by illegal fishing, that’s really what the DIU says. Is that why the Pentagon is interested in it and not just the Commerce Department, for example?
Bryce Goodman: Yeah, absolutely. Fishing provides a large amount of protein for a large part of the world’s population. Thus, undermining illegally fishing stocks can give rise to local conflicts over scarce resources. And then, of course, there is the problem of state actors who use their fishing fleet to expand their presence in the waters, and sometimes use fleets that nominally fish but do not necessarily actually fish.
Tom Temin: A little fishing and a little espionage at the same time.
Bryce Goodman: That’s right.
Tom Temin: OK. And tell us about this challenge. What are you trying to do, because a nation’s warship meets a fishing boat, and they want to? How can AI and machine learning help in this whole process?
Bryce Goodman: The ocean is massive on the right, and covers more than two-thirds of the Earth’s surface. And our ability to detect ships in the ocean is incredibly limited right now. It’s limited in the sense that ships are supposed to have what’s called an AIS, an automatic identification signal. And it’s a ping that says, this is where I am, this is where I am going, this is who I am. And that’s great when ships turn them on. The problem is that many ships do not light them, especially when they are involved in illicit activity. And so in this case, you don’t have an AIS signal, and you need other means to be able to locate and identify these vessels. You can use satellite imagery to do this. Optical imagery is great, the challenge is if you have a cloud in the way or if it’s dark you’re out of luck. And you also have the problem that, like I said, the ocean is huge. So trying to get a person to look at pictures of the whole ocean is totally infeasible. So what we are doing is we have set up a computer vision challenge with artificial intelligence, where the participants will receive a dataset consisting of a synthetic aperture radar. These are therefore images derived from a satellite equipped with a radar, and this radar is capable of penetrating through cloud cover and can operate day and night. So this avoids some of the problems associated with nighttime issues and clouds. Teams go, or individuals will have access to this data. And this data can be used as a basis for training machine learning algorithms. And these algorithms should be able to automatically identify vessels, whether or not those vessels have their AIS enabled.
Tom Temin: So it’s not something you could see from naked eye imagery, but there are anomalies in the data or something that could give birth to a vessel, in other words.
Bryce Goodman: Yes, therefore, synthetic aperture radar is quite difficult for human analysts to understand. It is possible, but it is also quite heavy. So what we do is literally find needles in the haystack. And what we can possibly do is cross-correlate and say, okay, there’s a bunch of vessels inside an area that we know is closed to fishing, and then we can send a patrol over there to check on what’s going on.
Tom Temin: We chat with Bryce Goodman, he is chief strategist for artificial intelligence and machine learning at the Defense Innovation Unit. Well, is it really fishing, or is it just all kinds of boats that are in the territorial waters of a country, I guess it’s the United States that you don’t know all about? just not? I mean, are you really looking for fishing boats or boats of any type that shouldn’t be there?
Bryce Goodman: Yes, that’s a great question. So the first thing to say is that we are not just looking at US waters. It is truly a global competition in terms of reach. And one of the great things about these prize contests is that all the data is made available for free and openly so that researchers all over the world can build that capacity, and we have a prize track that encourages people to make their algorithms completely open. The source. Which means any group, whether you are the Ecuadorian Fisheries Authority or the Fisheries Authority of Japan, would be able to download and use these algorithms. So that’s something that I think is really unique about these prize contests. And in a challenge like tracking illegal fishing, that really requires multinational cooperation, doesn’t it? The high seas are international. And many of the countries that suffer the most, like Palau, simply do not have navies capable of enforcing their rights and protecting their exclusive economic zone. And you ask if it’s really fishing, and I would say it’s illegal fishing first and foremost, and illegal fishing has things that go way beyond damage to fish. . We know that forced labor and human slavery are a huge component of the world’s fishing fleet. We know there are other things like drug trafficking and human trafficking that these fleets can be involved in. And we also know that some armies have used the cover of fishing fleets to carry out aggressive actions.
Tom Temin: Like China, for example.
Bryce Goodman: It’s true, like China.
Tom Temin: OK. And how do you keep the datasets out of the hands of North Korea, out of the hands of the Chinese, or out of the hands of the Russians, so that they can figure out how to make their ships for espionage and human trafficking? , etc., even less detectable?
Bryce Goodman: So that’s also an excellent question. And that’s, of course, one of the risks when you launch these challenges is to make things open and accessible. And we mitigate that risk in a number of ways, but at the end of the day you always give others access to data at some level, and when you open something, that means it’s really open. However, our assumption is that in a world where the oceans are completely transparent, America and its allies win, and China and other nations that commit crimes at sea, lose. So, I am more than happy for the Chinese, North Korean competitors, well the North Korean competitors can not compete because of the sanctions, but, if the Chinese competitors want to build algorithms that allow us to track the fleets of Chinese peaches, we will definitely take them. We win when the oceans are more transparent. And I think China can take advantage of it. There are elements in China, I hope, that would be concerned about the problem of illegal fishing and where we might be able to cooperate.
Tom Temin: And what are the details of this program? When are our bids due? And how does it work ? What types of prizes are awarded? Is this a first floor, second floor type door? How does it all work?
Bryce Goodman: The competition is therefore organized in partnership with the non-profit association Global Fishing Lodge. And we also have the support of the US Coast Guard, the National Oceanic and Atmospheric Administration (NOAA), the National Maritime Intelligence-Integration Office (NMIO), and we will also integrate into the US Navy. The competition will be open in mid-August, and it will run for three months. Those interested in competing in the [competition] can search for “iuu.xview.us”, they can register and they will be able to download data. And then they will have three months from mid-August to submit algorithms. These algorithms will automatically be noted on the data that we have kept in reserve. These scores will show up on a leaderboard, so you’ll be able to see where you stack up and stack up against everyone else. And then we’re going to give awards to various individuals or groups depending on where they’re from. So there are two prize pools – there is $ 100,000 which is provided by our NGO partner, Global Fishing Watch. And it’s money that anyone in the world can make. And then there’s an additional $ 50,000 that the Defense Innovation Unit is putting in place, and that’s specifically for organizations based in the United States.
Tom Temin: And so intellectual property, as you said, will be put in the public domain so that people can develop products around it quite freely.
Bryce Goodman: Yes, so definitely, if you are on the Global Fishing Watch Tour and want to be eligible for the $ 100,000 in prize money, you need to make your algorithms open source. On the Defense Innovation Unit component, we have a component called the closed source component which allows companies to compete. These are much smaller cash prizes, but it allows these companies to retain all intellectual property rights to their algorithms.
Tom Temin: Bryce Goodman is an Entrepreneur as Chief Strategist for Artificial Intelligence and Machine Learning at the Defense Innovation Unit. Thank you very much for joining me.
Bryce Goodman: My pleasure.