Scientists use digital actuality for fish to show robots the best way to swarm

April 30, 2025

The GIST Editors' notes

This text has been reviewed in line with Science X's editorial course of and insurance policies. Editors have highlighted the next attributes whereas guaranteeing the content material's credibility:

fact-checked

peer-reviewed publication

trusted supply

proofread

Scientists use digital actuality for fish to show robots the best way to swarm

Fish are masters of coordinated movement. Faculties of fish haven’t any chief, but people handle to remain in formation, keep away from collisions, and reply with liquid flexibility to modifications of their surroundings. Reproducing this mixture of robustness and suppleness has been a long-standing problem for human-engineered programs like robots.

Now, utilizing digital actuality for freely-moving fish, a analysis crew based mostly in Konstanz has taken an essential step towards that objective.

The analysis is revealed within the journal Science Robotics.

"Our work illustrates that options advanced by nature over millennia can encourage sturdy and environment friendly management legal guidelines in engineered programs," mentioned first writer Liang Li from the College of Konstanz. Co-author Máté Nagy from Eötvös College underscores this, saying, "The invention opens up thrilling potentialities for future purposes in robotics and autonomous car design."

Deciphering nature's hidden algorithm

Utilizing a digital actuality (VR) setup that mimics pure education, researchers positioned particular person juvenile zebrafish into networked arenas the place every fish may freely work together with "holographic" digital conspecifics. Every digital fish was a projection of an actual fish from one other enviornment, that means that fish may swim and work together collectively in the identical digital world. The absolutely immersive 3D surroundings lets researchers exactly manipulate visible stimuli and file how the fish reply.

This excessive degree of management allowed the scientists to isolate precisely which cues the fish have been utilizing to information their interactions with different fish. In different phrases, they might reverse-engineer the conduct of education in zebrafish to grasp how fish clear up the complicated downside of coordinating their movement.

The answer they found was a easy and sturdy legislation based mostly solely on the perceived place, not the velocity, of their neighbors to manage their following conduct.

"We have been stunned by how little data the fish have to successfully coordinate actions inside a faculty," says Iain Couzin, senior writer on the research and Director of MPI-AB and speaker on the Cluster of Excellence: Collective Habits. "They use native guidelines which might be cognitively minimal, however functionally wonderful."

To see simply how lifelike the management legislation was, the crew examined it with actual fish. They carried out a VR "Turing check," based mostly on the idea of testing whether or not folks can inform if they’re interacting with an actual human or with synthetic intelligence.

Within the aquatic Turing check, an actual fish would swim with a digital fish that switched between being actual and being managed by the algorithm they found. The true fish couldn’t inform the distinction. They behaved the identical whether or not interacting with an actual conspecific or the digital follower ruled by the algorithm.

From fish to machines

To check the broader utility of their discovery, the crew embedded it in swarms of robotic vehicles, drones, and boats. The robots have been tasked with following a transferring goal utilizing both parameters from the zebrafish algorithm or from a state-of-the-art methodology utilized in autonomous automobiles known as Mannequin Predictive Controller (MPC).

Throughout all assessments, the pure management legislation that fish have advanced delivered efficiency that was practically indistinguishable from MPC by way of accuracy and vitality consumption—however at a fraction of the complexity.

Oliver Deussen, a co-author on the research and Professor of laptop science on the College of Konstanz and speaker on the Cluster of Excellence: Collective Habits:

"This work highlights the reciprocal relationship between robotics and biology—utilizing robotics to discover organic mechanisms, which in flip can encourage new and efficient robotic management methods."

Extra data: Liang Li et al, Reverse engineering the management legislation for education in zebrafish utilizing digital actuality, Science Robotics (2025). DOI: 10.1126/scirobotics.adq6784. www.science.org/doi/10.1126/scirobotics.adq6784

Journal data: Science Robotics Offered by College of Konstanz Quotation: Scientists use digital actuality for fish to show robots the best way to swarm (2025, April 30) retrieved 30 April 2025 from https://techxplore.com/information/2025-04-scientists-virtual-reality-fish-robots.html This doc is topic to copyright. Aside from any honest dealing for the aim of personal research or analysis, no half could also be reproduced with out the written permission. The content material is supplied for data functions solely.

Discover additional

Discovering the beat of collective animal movement: Scientists present reciprocity is essential to driving coordinated actions 0 shares