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

Secret agent fish? Robot infiltrates aquatic schools, influences their behavior

Add as a preferred source on Google

Robots may be getting more realistic but most of us can still discern a human from a humanoid. Fish, on the other hand, seem easier to fool.

Researchers at École Polytechnique Fédérale de Lausanne (EPFL) in Switzerland have developed an aquatic robot that can infiltrate schools of zebrafish, going so far as to influence their activities. By designing robots to integrate with animal communities, the roboticists hope to unravel these animals’ behavioral and social structures by learning more about how they communicate.

Recommended Videos

“We wanted to study how a robotic agent can be inserted in a fish shoal, and, when inserted, how this robotic agent could learn how the fish interact, communicate, and take decisions,” Frank Bonnet, an EPFL postdoctoral researcher who worked on the project, told Digital Trends.

The robot has two parts — a magnetic fish and a motor, which is positioned underneath the aquarium and helps propel the robot through the water.

To infiltrate the school, the EPFL team identified characteristics that real fish use to distinguish themselves and designed the robot accordingly. These included physical traits (such as shape, color, and markings) and behavioral traits (including acceleration speeds, size of schools, and how the fish move their tails in transit). Taking the project one step further, the researchers designed the robot to learn and adapt from the behavior of the real fish.

“In somes studies, we compared groups composed of only fish with groups composed of fish and robots,” Bonnet said. “We could show that the two are behaving similarly, which shows that the robots are able to act like fish, and integrate groups of fish. We also used other setups, such as a ring-shaped setup, to show that the robotic agents can influence on the swimming direction preference of the fish.”

EPFL researchers have previously done similar work on cockroach communities, which are willing to accept practically anything that emits specific cockroach pheromones. Fish posed a much more difficult task, but Bonnet already sees how the work will shed light on the nuances of their societies.

“We could already learn which features are relevant for the fish for social integration,” he said. “We can also learn the hierarchy in fish societies, if for instance there are some leaders, and what are the important aspects to become a leader in fish groups, and how the fish are collectively taking decisions.”

A paper detailing the research was recently published in the journal Bioinspiration & Biomimetics.

Dyllan Furness
Former Contributor
Dyllan Furness is a freelance writer from Florida. He covers strange science and emerging tech for Digital Trends, focusing…
AI chatbots will happily create fake news articles, and tests show ChatGPT is the worst at it
Electronics, Phone, Mobile Phone

The biggest AI chatbots are supposed to help people find information, not manufacture misinformation. But a new investigation suggests that separating those two jobs isn't as straightforward as many AI companies would like users to believe. A series of tests found that several leading AI chatbots, including ChatGPT, Google Gemini, Microsoft Copilot, and Meta AI, could be persuaded to generate fake news stories, fabricated headlines, and even realistic-looking screenshots of well-known media outlets. While every chatbot showed some weaknesses, ChatGPT reportedly performed the worst, producing the most convincing fake content with surprisingly little effort.

The findings come from an investigation by German newsroom CORRECTIV, which tested how the AI assistants responded when asked to create false reports involving sensitive topics such as election fraud, government coups, war, vaccines, climate change, and conspiracy theories.

Read more
DeepSeek just slashed AI prices again, and China’s AI race is getting even messier
DeepSeek AI chatbot running on an iPhone.

China's AI race has never really been short on competition, but it's increasingly looking like a battle over who can charge the least rather than who can build the best model.

DeepSeek has unveiled its latest open-weight AI model, called V4-Flash, alongside a dramatic price reduction that makes it significantly cheaper for developers to use. The company has cut token pricing by 50% and has also decided against introducing a previously announced dynamic pricing system that would have increased costs during periods of heavy demand. The new changes paint a clear picture: DeepSeek appears more interested in winning over developers and expanding its user base than maximizing revenue in the short term.

Read more
AI slop stories are spoiling the childhood
Grandparents are gifting AI-generated story books to kids, and it's just gross.
A child reading a book.

AI slop, the colloquial term for low-effort AI-generated content, has emerged as a huge problem across different industries. Music labels are fighting streaming services to put an AI label on such tra/cks. Publishing houses are wary of AI-written drafts. Research journals are buried under a deluge of AI-generated papers. The software industry is reeling under the pressure of vibe-coded apps brimming with security flaws. Even Apple is struggling against a tide of bug reports created using AI. Of course, the disdain against AI is pretty obvious, but there's now a new dimension to it.

Boomers are gifting AI-generated books to their grandkids.

Read more