Controlling Fish Schools via Reinforcement Learning of Virtual Fish Movement

2026-03-17

AI Summary

To overcome the durability and mobility limitations of physical robots in underwater fish schooling scenarios, this study proposes a screen-based 2D virtual fish agent approach. For the first time, model-free reinforcement learning is applied to modulate the collective behavior of real animal groups, enabling virtual agents to learn effective herding strategies without requiring an accurate biological behavior model. By integrating visual stimuli with statistical analysis, both simulated and real-world experiments demonstrate that the learned policies significantly outperform baseline conditions—namely, no stimulus and a heuristic “edge-holding” strategy—in achieving directional control of fish schools. These results validate the feasibility and superiority of virtual agents for regulating biological group behavior.

Technology Category

• Intelligent Robots: Behavior Learning & Control
• Machine Learning: Reinforcement Learning
• Multiagent Systems: Agent-Based Simulation and Emergent Behavior

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

This study investigates a method to guide and control fish schools using virtual fish trained with reinforcement learning. We utilize 2D virtual fish displayed on a screen to overcome technical challenges such as durability and movement constraints inherent in physical robotic agents. To address the lack of detailed behavioral models for real fish, we adopt a model-free reinforcement learning approach. First, simulation results show that reinforcement learning can acquire effective movement policies even when simulated real fish frequently ignore the virtual stimulus. Second, real-world experiments with live fish confirm that the learned policy successfully guides fish schools toward specified target directions. Statistical analysis reveals that the proposed method significantly outperforms baseline conditions, including the absence of stimulus and a heuristic "stay-at-edge" strategy. This study provides an early demonstration of how reinforcement learning can be used to influence collective animal behavior through artificial agents.

Problem

• fish schools
• collective behavior
• behavioral control
• virtual agents
• reinforcement learning

Innovation

• reinforcement learning
• virtual fish
• collective animal behavior
• model-free control
• bio-inspired robotics