Skip to main content
  1. Home
  2. Emerging Tech
  3. News

MIT’s ocean-exploring robotic fish takes a test swim through a Fiji reef

Add as a preferred source on Google

It’s often said that we know less about the ocean depths than we do about the surface of the moon. That fact was highlighted again this week when Smithsonian explorers published a paper about a new ocean reef zone, aptly called the rariphotic, where about half of the fish were previously undiscovered.

Recommended Videos

Even with 21st-century advances in small submersible vehicles, accessing and documenting marine life still proves to be a challenge. But now engineers at the Massachusetts Institute of Technology hope to have a solution with a soft robotic fish named SoFi, which they say can get closer to real fish by imitating their behavior.

A Robotic Fish Swims in the Ocean

“We’ve developed a soft robotic fish that mimics the movement of real fish and has now, for the first time, demonstrated an ability to swim alongside real fish in the actual ocean,” Robert Katzschmann, an MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) doctoral candidate who lead the project, told Digital Trends. Katzschmann and his team think this is the first robot of its kind that can swim freely underwater.

From the outside, SoFi looks like a rather unassuming rubber fish but it’s what’s inside the robot that matters. Electronics, including sensors and a Linux pc, are packed within its brain case, urethane foam offers buoyancy, and a hydraulic pump near its chest helps move the tail. By pumping water alternatively into two ballon-shaped chambers, the robot’s tail bends back and forth, powering the robot along. A camera at the front records it’s journey through the water.

SoFi can be remote controlled via a waterproof Super Nintendo controller, adapted to communicate using ultrasonic signals. As it swims, it look a lot like a real fish gliding through the water — albeit a bit less gracefully. After testing the robot at Rainbow Reef in Fiji, the researchers said fish appeared to be unbothered by SoFi’s presence.

Moving forward the researchers want to make SoFi faster and more autonomous, with onboard sensors that allow it to follow other fish. Although SoFi would make a nifty toy, Katzschmann and his colleagues hope the technology finds more a professional applications, giving marine scientists and engineers better access to underwater worlds.

“The ability to more easily get water samples using robots would enable us to learn a lot more about more remote parts of the ocean,” he said. “Using them to inspect underwater infrastructure would allow for the more regular maintenance of hardware that may not otherwise be kept up to code … I could imagine one day being able to use several of these fish to record data from the ocean, including even taking water samples. I could also imagine SoFis someday being used for underwater inspection of critical infrastructure like oil rigs and pipelines.

Dyllan Furness
Former Contributor
Dyllan Furness is a freelance writer from Florida. He covers strange science and emerging tech for Digital Trends, focusing…
Researchers hid a prompt injection inside a PNG, and AI fell for it
Hacker

AI coding assistants like Claude are becoming every developer's favorite coworker. They can review code, explain confusing functions, and even write entire features with a single prompt. But new research suggests that this growing trust could also become their biggest weakness.

A team of security researchers (professor Sudipta Chattopadhyay and researcher Murali Ediga) has demonstrated an unusual attack that doesn't target the AI model directly. Instead, it targets what the AI doesn't pay enough attention to during code reviews. Rather than hiding malicious instructions in lines of code, the researchers tucked them inside an image file. Since many AI review tools treat images as decorative assets rather than as something worth inspecting, the pull request can appear perfectly harmless and sail through the review.

Read more
AI has already fallen into the wrong hands and they’re using it to make bombs
Logo, Text

Artificial intelligence has quickly become the go-to tool for everything from writing emails and summarizing meetings to helping students study or developers debug code. But the same technology that saves people time can also be misused, and a new report suggests that terrorist organizations are finding ways to do exactly that.

According to a research paper shared with The New York Times ahead of its publication, researchers found evidence that members of Boko Haram have been using popular AI chatbots to support both day-to-day activities and combat-related tasks. Interviews with 27 former members conducted in Nigeria over the past two years suggest that tools such as ChatGPT, Gemini, Claude, Grok, Meta AI, and DeepSeek were used to gather technical information, troubleshoot weapons, and even assist with planning attacks.

Read more
Claude Code can now browse the web without opening Chrome
The desktop app now includes an in-app browser that can read websites, click links, and interact with web apps.
Claude Code Featured

Developers spend a surprising amount of time bouncing between their code editor, browser tabs, API documentation, GitHub issues, and design files. Anthropic thinks Claude Code should simply do all of that without constantly asking users to switch windows. The company has announced a new in-app browser for Claude Code on desktop, allowing its AI coding assistant to open websites, read documentation, inspect designs, and interact with web pages directly from within the application.

A browser built into Claude Code

Read more