A scuba diver swimming in the Great Barrier Reef in 1985 would have seen something completely different than a diver today. In the last few decades, thanks to factors like warming seas and pollution, more than half of the coral has died. But more subtle changes are happening so quickly—and in hard-to-reach areas all over the world—that they're hard for researchers to track.
Until recently, studying coral meant diving down with a tape measure, calculating the size of the tip of a branch, and then trying to make some crude estimations about the rest of the reef. But now 3-D scanning is making it possible to follow detailed shifts in coral health.
"Underwater is like the Wild West," says Sly Lee, founder of The Hydrous, a nonprofit pioneering 3-D scanning of reefs. "It's like it's stuck in the early 1900s—the technology used is so antiquated."
After Lee dove with researchers using tape measures, he decided to test out reality computing instead. "People didn't know it was possible," he says. "Programmers are in their own world, in labs. They didn't think reality capture underwater would work, because there are so many variables."
It turned out that it worked. "You take a ton of pictures from a regular camera, and then the software we use is smart enough to triangulate and find where points match up," Lee says. The software, from Autodesk, turns 200,000-plus shots of a reef into a virtual 3-D model. Right now, because the software is still in development, Autodesk does some of the processing by hand, but that should soon change.
"If you trace this technology back to when it first began, maybe around 30 years ago, and put that progress on a logarithmic curve, it's rising exponentially," says Lee. "Virtual reality 3-D scanning has been growing for decades, but we're just now at the knee of the curve. It's just now picking up."
Robots may soon be cheap enough to take the underwater pictures, too. "Right now we're using a diver method, and I'm physically going underwater, scanning by hand," he says. "But we're also seeing exponential drops in costs for robots."
With the current method The Hydrous uses, it takes around 15 to 30 minutes to take thousands of photos around a reef. After the model is created, researchers can go back to the same spot later to repeat the process and compare differences. The team also plans to use the data to create virtual reality reefs for anyone to visit.
"Right now you can watch a video or look at pictures, which is beautiful, but it's not interactive," says Lee. "Reefs are pretty difficult to access since they're underwater, so we're developing virtual reality content that will be ready later this year."
Ultimately, he hopes to map out the whole ocean. "At first, we wanted to scan a subset of coral reefs," he says. "But that's when I was thinking linearly." Lee, who is currently part of a startup accelerator at Singularity University, is now obsessed with with the idea of exponential growth in technology.
"We're going to be able scan all of the reefs, easily," he says. "In the next five years—that's conservative, probably way faster than that."
The team is planning to enter an upcoming XPrize for ocean scanning. "It's $10 million to scan the bottom of the ocean," Lee says. "I was like, holy shit, we're already doing that."
Eventually the same technology could also be used in space. "Exponential technology meant to work underwater, like what we're doing, will also work in space," he says. "What we're doing now underwater in remote areas around the world is kind of an analogue to what we could do later in space."