Project preview ... more soon.
Exploring the Hidden Life of Our Oceans
Oceans cover over 70% of Earth’s surface, yet we know more about the Moon and Mars than the deep sea. Scientists estimate that only 10-25% of marine species have been identified, with many remaining hidden in ocean sediments and deep waters. But species are disappearing faster than we can discover them—on average, it takes 13.5 years to classify a new species, meaning many will go extinct before we even know they existed.
Despite their size, the smallest ocean organisms—like plankton—are essential to life on Earth. They form the base of the marine food web, generate more oxygen than all forests combined, and capture up to 40% of global carbon, making the ocean our largest carbon sink.
A New Approach to Ocean Exploration
Digital Ocean Engine is a tool that lets us explore the microscopic world of the ocean in high resolution. It combines high-throughput microscopy, image processing, machine learning, and gaming engines to create interactive 3D avatars of marine life—bringing unseen organisms to life.
Instead of manually analyzing individual specimens under a microscope, we scan entire water samples using an imaging flow cytometer. This device rapidly captures images of organisms as they pass through a flow cell, allowing us to process hundreds of specimens per minute instead of spending hours on a single sample.
Once scanned, each image is analyzed using image processing and machine learning. We segment individual organisms and use platforms like Scikit and Ecotaxa to identify species, training an AI model to classify marine life automatically—far faster than traditional methods.
From Data to Digital Life
While the initial images are low-resolution, they provide enough detail to capture key biological traits like size, gender, life stage, chlorophyll content, and reproductive status. Using light-sheet and confocal microscopy, we then create high-resolution 3D avatars—digital replicas of each species, complete with internal organs.
These detailed 3D models are integrated into a gaming engine, where we customize each avatar based on the individual specimens detailed features. But we don’t stop at static models—our goal is to make these digital organisms move, feed, and interact just like their real-world counterparts.
We simulate realistic behaviors based on scientific studies and video footage. Instead of simply viewing a static image, you can watch it swim and interact with its environment in a virtual marine world.
A Living Digital Archive
The Digital Ocean Engine creates more than just a snapshot of biodiversity—it builds a growing digital archive of marine species. By capturing data over time, we can begin to model ocean life across different time periods, helping us understand how species evolve, respond to environmental changes, and interact within ecosystems.
But what if we could go further?
Imagine a simulation that not only shows the present but allows us to visualize the past and predict the future of our oceans. By combining real-time data with historical records and climate models, we could gain deeper insights into marine biodiversity and the critical role oceans play in sustaining life on Earth.
The Digital Ocean Engine is a step toward that future—where we don’t just explore the ocean, but truly understand and protect it.