Josephine Condemi
Observation of the deep is expensive and concentrated in the exclusive economic zones of the USA, Japan and New Zealand. Katy Croff Bell with the Ocean Discovery League is working to make it more accessible and widespread: here’s how.
Gerridae, or water striders, glide on the water’s surface thanks to their long legs covered in hydrophobic microhairs that don’t break the surface tension, allowing them to stay afloat. These insects have inspired a new generation of ultralight floating robots that can be used as mobile stations for collecting and processing environmental data on the ocean-atmosphere system.
A prototype of these biomimetic robots, with a a bacterial activity-powered battery, was published in June 2024 in Advanced Materials Technologies1 and developed within the Ocean of Things programme by the United States Defense Advanced Research Projects Agency, dedicated to creating a distributed network of environmental sensors.
But beyond the surface, there are still unexplored abysses, even though remotely controlled vehicles and then underwater drones have joined submarines since the 1980s. Why? «It’s really hard to see through water», explains researcher Katy Croff Bell.
Katy Croff Bell, born in 1978, is a deep-sea explorer with more than 30 oceanographic expeditions to her credit. In 2021, she founded the non-profit organisation Ocean Discovery League, of which she is president. She holds a BSc in Ocean Engineering from MIT, an MSc in Maritime Archaeology from the University of Southampton and a PhD in Geological Oceanography from the Graduate School of Oceanography at the University of Rhode Island. He founded and directed the Open Ocean Initiative at the MIT Media Lab, a 2017-2019 Vice-Chair of the Federal Advisory Committee on Marine Protected Areas and a 2017-2020 National Geographic Fellow.
«When you position the cameras, you can see only a very, very small part of the abyssal landscape. It’s like being on the side of a mountain, at night, during a snowstorm, with just a flashlight: you have a very, very narrow view and need a lot of time to get a picture of the entire mountain. Another reason is that, in the last 60-70 years, exploration tools have been very expensive and, consequently, few and concentrated only in some places like the United States, Europe, Japan, and New Zealand. Most coastal countries around the world, especially those with large amounts of deep ocean within their exclusive economic zones, do not have access to this type of equipment. For the past 70 years, exploration has been concentrated in the exclusive economic zones of the United States, Japan, and New Zealand», Croff Bell emphasises. «How can we make these types of systems more available, more accessible, so that they reach more people and places?».
Not a game for everyone
Croff Bell founded the Ocean Discovery League in 2021, a non-profit aimed at accelerating deep-sea exploration through low-cost and open-source systems. Among the institute’s first achievements is The 2022 Global Deep-Sea Capacity Assessment, which mapped the state of global ocean exploration through online surveys and manual research data across 186 geographic areas in six main regions: Europe, Asia, North America, Africa, Oceania, Latin America, and the Caribbean. «We tried to reach as many people and countries as possible», Croff Bell explains. «At first, we relied on our mailing list, then used a database called Ocean Expert: for all the countries that did not respond to the survey, we contacted local scientists and invited them to participate. In the end, we got more than 300 responses: wonderful, but not enough. We then hired an additional team of researchers in each region of the world with deep waters to supplement that survey».
What were the results? «One of the most important was to document consistently and in detail the inequalities between countries. An aspect that encouraged me greatly was seeing how in many places where there are no tools, there are experienced and skilled people who would like to use them. I see it as a great opportunity to work with people who have the skills and interest but lack access and funding: the traditional vehicles and systems we use cost millions of dollars to purchase, not to mention the operation of research vessels, when perhaps their country has other priorities. That’s why we’re trying to reduce costs: if people want to do it, we can help them».
An unexpected result of the study? «When we asked people if the exploration of the ocean or the deep sea was important for their country, the response in the United States was lower than in many other places, even though investments in this area are much higher there than elsewhere».
Sponsored ad
Sponsorizza con noi
Open-source AI systems
How can Artificial Intelligence accelerate the processing of data from ocean explorations? «There are two main challenges right now», explains Croff Bell. «One is the amount of data, the fact that people have piles of hard drives on their desks that are not in a shareable and accessible format. And the other is that usually, when people analyse their data, they are conditioned by their level of expertise and their field of specialisation. So, a biology student will annotate or analyse videos very differently from a geology professor: they will look for different things, starting from different levels. We hope that by using artificial intelligence, there will be a more consistent ability to analyse data, so as to identify new discoveries that might not fall within the researcher’s expertise: a geologist might only look for rocks and not be interested in coral, and therefore not notice a new species on the rock they are observing or want to collect. This will be a multi-year effort. But we are working with the Monterey Bay Aquarium Research Institute on the “Ocean Vision AI” project, which consists of building an online platform that allows people to upload and analyse their data using machine learning algorithms to make data analysis faster and more efficient. As part of this project, we have an open-source annotated image database called FathomNet2, which is a start».
How was the process of collecting and uploading this database? What is it made of? «It is an ongoing process, starting with data from the Monterey Bay Aquarium Research Institute, which has been one of the leaders in annotating ocean videos for decades, with a proper room dedicated to this activity, which has allowed us today to have vast archives needed to train the algorithms. We also have a fair amount of data from the Office of Ocean Exploration of NOAA, the United States National Oceanic and Atmospheric Administration, which has conducted expeditions worldwide. It is an open database, so anyone can upload their data. And it continues to grow over the years».
What are the Institute’s medium- and long-term goals? «In the long term, first, we are working on a project to identify points where we have already seen the seabed. We have collected a metadata database from nearly 50,000 dives. Have there been changes over time? This is something we are working on right now, to establish what we have seen in the last 70 years. Then we are working on finding ways to address deep-sea exploration more strategically and globally: I think research is important and valuable, but much of the ocean has remained neglected, and we need to find a combination between current very specific scientific interests and the exploration of the other 99% of the ocean, so that we can ask more questions and formulate other scientific hypotheses in the future. For example, all deep-sea creatures have much to teach us about adaptation to their unique environments, whether it be muddy abyssal plains, or adaptations to chemosynthesis and life near hydrothermal vents: the biological world of the ocean is incredible for the way different organisms have adapted in so many different ways. Therefore, we want to develop low-cost and more user-friendly tools so that people can access them, have more explorers from more places around the world, and develop training programmes to enable them to use them. These are the main areas we are working on right now.
The story this article is about was discovered using an artificial intelligence tool, Asimov, developed by ASC 27, especially for Mangrovia. The tool helped us find the story, but the rest of the content you read and see is the outcome of creative processes and human sensibilities and is in no way generated by artificial intelligence. Follow us to find out the details of how we use artificial intelligence in the newsroom!
- For more on this prototype, see Elhadad, A., et al. (2024) Revolutionizing Aquatic Robotics: Advanced Biomimetic Strategies for Self-Powered Mobility Across Water Surfaces. Advanced Materials Technologies. doi.org/10.1002/admt.202400426 ↩︎
- On FathomNet, see Katija, K. et al. (2022). FathomNet: A global image database for enabling artificial intelligence in the ocean. Sci Rep 12, 15914, https://doi.org/10.1038/s41598-022-19939-2 ↩︎
