A unique collaboration between MDA and two other companies will show how satellite surveillance technology can – in real time – ferret out “dark vessels” that may be associated with illegal fishing.
MDA, Starboard Maritime Intelligence and Unseenlabs will perform a live demonstration of their work during the Indo Pacific 2022 International Maritime Exposition being held May 10 – 12 in Sydney, Australia. In this example, they will focus on illegal fishing for southern bluefin tuna, using RADARSAT-2 data and other products brought into the collaboration.
Dark vessels are ships that travel without broadcasting the internationally required automatic information system (AIS) data that provides statistics such as position, course, speed, nationality or the destination port – which helps manage maritime traffic worldwide, especially those AIS that broadcast to satellites.
RADARSAT-2, however, provides C-band radar coverage meant for maritime surveillance of dark vessels, across hundreds of thousands of square kilometres of ocean and using several different imaging beam modes. (There are 20 available, and most can be repurposed for maritime, according to MDA.)
The aim of MDA is in part, to showcase how optical and synthetic aperture radar (SAR) from RADARSAT-2 may be used to provide real-time information even on a vessel choosing not to broadcast its presence, the company’s Mark Carmichael told SpaceQ. Carmichael is MDA’s director of product management (geointelligence) and has been with the company since 2000.
In many cases, information can be delivered in under an hour to the customer, which Carmichael said makes MDA distinctive among the competition – and which puts its products before government customers across the “Five Eyes,” an intelligence alliance including Australia, Canada, New Zealand, the United Kingdom, and the United States.
“That kind of reliable, very fast turnaround is what people are ready to pay for, and it’s very valuable,” Carmichael said. MDA has an advanced analytics, multi-sensor data fusion platform, which brings together source data and machine learning analytics. “It’s presented to the user to do maritime intelligence gathering and interpretation,” he added of the platform.
The new collaboration comes from the dark vessel detection project, with the Department of Fisheries and Oceans (DFO) and Defence Research and Development Canada. (MDA announced a contract in February 2021 in association with that project, to address illegal, unreported and unregulated or IUU fishing.)
“DFO is championing this counter-illegal fishing capacity building activity around the world, in part for themselves, but also for other governments,” Carmichael said, citing examples such as the Pacific Islands Forum Fisheries Agency, a 17-member group that fights illegal fishing and does monitoring, control and surveillance of maritime fishing.
“Out of that program,” he added, “dark vessel detection program is coming. The maritime insights products – and all the IP [intellectual property] built in there – is owned by MDA. We’re rolling it out in part with our own investment money, but also with new customers.”
With this new project, Carmichael said, the benefit for customers is “they get to use all the enhancements that we built,” which serves as a backbone for collaborations between the Canadian government and jurisdictions such as New Zealand, Ecuador and the Pacific Islands Forum Fisheries Agency.
The new collaboration includes the French company Unseenlabs – which is dedicated to tracking dark vessels – along with Starboard Maritime Intelligence, built by the Xerra Earth Observation Institute. Xerra is a New Zealand Earth observation and remote sensing technologies research institute that builds products and services in these areas, and MDA has been collaborating with Xerra for about three years.
The new illegal fishing partnership among MDA, Unseenlabs and Starboard uses machine learning (ML) in two principal ways, when speaking exclusively of MDA’s work. (Carmichael clarified that the two partners have their own products as well, with potentially distinct processes.)
The first ML application is to offer systematic classification from RADARSAT-2’s extra-fine mode. The algorithm can extract the class of the vessel from the imagery, along with other information such as velocity. Additionally, Carmichael said, ML – starting in the most recent release of the product – can detect vessels in high-resolution optical imagery.
He added MDA does not, at this time, anticipate bringing on the next-generation RADARSAT Constellation Mission (RCM) into the collaboration, because the tasking of that set of satellites is under the Government of Canada.
The government, Carmichael said, is releasing a set of open data from RCM under the Earth Observation Data Management System (EODMS) website. For the moment, dark vessel detection “is not an area that’s getting covered in the open data realm,” he said. Furthermore, the website does not provide anywhere near real-time access in the sense that MDA’s new collaboration would, which is “critical” to timely dark vessel detection, Carmichael said.
Looking ahead, Carmichael said MDA anticipates seeing this collaboration develop along two different tracks. The first, regarding maritime insights more generally at MDA, shows how the products to fight dark vessels and illegal fishing “is super powerful stuff for Canada,” Carmichael said. MDA’s work, he said, is removing the difficulties of “data wrangling” – meaning the set of processes to transform and clean up raw data to useable insights.
The second track is linkages with Starboard and Xerra. “We’re supporting them to develop their business that addresses New Zealand. That has strategic value because we have Five Eyes countries seeing the timeliness and the responsiveness RADARSAT-2, which is a very strong differentiator compared to other folks in the market.” Carmichael added MDA hopes to keep developing the relationship with Xerra in the future.