Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Nov 4, 2025

Cardiac Muscle-cell Based Actuator and Self-stabilizing Biorobot - PART 1
11:22

Cardiac Muscle-cell Based Actuator and Self-stabilizing Biorobot - PART 1

Published on: July 11, 2017

8.3K

Reinforcement learning with artificial microswimmers.

S Muiños-Landin1,2, A Fischer1, V Holubec3,4

  • 1Molecular Nanophotonics Group, Peter Debye Institute for Soft Matter Physics, Universität Leipzig, 04103 Leipzig, Germany.

Science Robotics
|May 27, 2021
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Fourier Plane Tomographic Spectroscopy Reveals Orientation-Dependent Multipolar Plasmon Modes in Micrometer-Scale Janus Particles.

ACS nano·2026
Same author

Physical embodiment enables information processing beyond explicit flow sensing in active matter.

Science advances·2026
Same author

Thermofluidic Nonequilibrium Assembly of Reconfigurable Functional Structures.

ACS nano·2025
Same author

The 2025 motile active matter roadmap.

Journal of physics. Condensed matter : an Institute of Physics journal·2025
Same author

The ever-expanding optics of single-molecules and nanoparticles.

The Journal of chemical physics·2024
Same author

Harnessing synthetic active particles for physical reservoir computing.

Nature communications·2024
This summary is machine-generated.

Artificial microswimmers learned to navigate in noisy environments using machine learning. This study shows collective learning is possible, with noise impacting speed and behavior, revealing an optimal velocity for delayed responses.

Area of Science:

  • Physics
  • Robotics
  • Artificial Intelligence

Background:

  • Artificial microswimmers mimic biological organisms but lack adaptive capabilities.
  • Both microswimmers and organisms face Brownian motion, randomizing movement.
  • Existing artificial microswimmers have limited environmental adaptability and memory.

Purpose of the Study:

  • To explore adaptive behavior in artificial microswimmers using machine learning.
  • To investigate collective learning in noisy environments.
  • To demonstrate navigation capabilities under Brownian motion influence.

Main Methods:

  • Combined real-world artificial active particles with machine learning (reinforcement learning).
  • Utilized real-time control of self-thermophoretic active particles.

More Related Videos

Bioinspired Soft Robot with Incorporated Microelectrodes
08:24

Bioinspired Soft Robot with Incorporated Microelectrodes

Published on: February 28, 2020

9.1K
Preparation and 3D Tracking of Catalytic Swimming Devices
06:50

Preparation and 3D Tracking of Catalytic Swimming Devices

Published on: July 1, 2016

7.8K

Related Experiment Videos

Last Updated: Nov 4, 2025

Cardiac Muscle-cell Based Actuator and Self-stabilizing Biorobot - PART 1
11:22

Cardiac Muscle-cell Based Actuator and Self-stabilizing Biorobot - PART 1

Published on: July 11, 2017

8.3K
Bioinspired Soft Robot with Incorporated Microelectrodes
08:24

Bioinspired Soft Robot with Incorporated Microelectrodes

Published on: February 28, 2020

9.1K
Preparation and 3D Tracking of Catalytic Swimming Devices
06:50

Preparation and 3D Tracking of Catalytic Swimming Devices

Published on: July 1, 2016

7.8K
  • Applied external control for collective learning experiments.

    • Main Results:

    • Demonstrated successful navigation in a noisy environment despite Brownian motion.
    • Showed that collective learning is achievable with external control.
    • Identified that noise decreases learning speed but modifies optimal behavior and decision strength.
    • Discovered an optimal velocity due to feedback loop time delay, potentially universal for delayed systems.

    Conclusions:

    • Machine learning enables adaptive navigation in artificial microswimmers.
    • Noise influences learning dynamics and emergent behavior in active matter systems.
    • Delayed responses in noisy environments may lead to universal optimal velocity characteristics.