The U.S. Coast Guard (USCG) icebreaker Healy is embarking on a groundbreaking mission this summer, capturing images of the Arctic to facilitate the study of this rapidly changing region. In an effort to develop artificial intelligence (AI) tools for analyzing Arctic imagery, researchers from Lincoln Laboratory installed a camera system aboard the Healy before it set off on its three-month science mission. This innovative dataset will contribute to safer navigation and improved maritime domain awareness, while also shedding light on the effects of climate change in the Arctic.
With the warming climate and melting sea ice, previously inaccessible Arctic passages are opening up to more traffic. This includes military vessels and ships engaged in illegal fishing, which can pose national security challenges to the United States. The changing Arctic also raises questions about how its climate, wildlife, and geography are being impacted. However, there are currently very few imagery datasets available for studying these changes.
While satellite and aircraft imagery can provide some information about the Arctic, the addition of an outward-looking camera attached to the Healy provides a closer and more detailed view of the environment. The captured images can be used to train AI computer-vision tools, helping the USCG plan naval missions and automate analysis. This technology will be particularly valuable considering that USCG resources in the Arctic are limited.
The USCG icebreaker Healy was an ideal choice for equipping with a new sensor to gather this dataset. The Lincoln Laboratory research team collaborated with the USCG Research and Development Center to design the Cold Region Imaging and Surveillance Platform (CRISP). This platform consists of a long-wave infrared camera manufactured by Teledyne FLIR, built to withstand the harsh maritime environments of the Arctic.
The camera, mounted at the front of the ship’s fly bridge, provides stabilization during rough seas and can capture images in complete darkness, fog, and glare. Equipped with a GPS-enabled time-synchronized clock and a network video recorder, it records both video and still imagery along with GPS positional data. The CRISP system can be operated manually from the bridge or set to an autonomous surveillance mode, panning back and forth to record video every three hours and a still image every 15 seconds.
Despite some challenges during installation, including unexpected access points, the laboratory team successfully implemented the CRISP system. The resulting dataset, estimated to be around 4 terabytes in size, will be made publicly available once the USCG science mission concludes in the fall. This open dataset aims to foster collaboration among researchers and expedite the development of effective AI tools for operating in the Arctic.
In addition to providing the dataset, the laboratory team plans to release a baseline object-detection model. This will serve as a starting point for other researchers to build upon, enabling advancements in classifying and tracking objects within the Arctic images. Beyond assisting with USCG missions, this ambitious project has the potential to contribute to the combat against climate change by leveraging AI to analyze Arctic data.
The Lincoln Laboratory research team is honored to have been a part of this project and looks forward to the progress that can be achieved by applying AI to the unique challenges facing the United States. They take pride in their ability to apply AI to critical national priorities, from combating Covid-19 to enhancing maritime awareness in the Arctic. By sharing their datasets and fostering collaboration, they hope to make a significant impact in the field of AI research and the understanding of the rapidly changing Arctic region.
