How Collective Intelligence Shapes Animal Behaviour

Animals frequently tackle problems collectively. Ants locate food, birds sweep across the sky, and fish travel as though they were one organism. There is no leader controlling these actions: each animal simply responds to straightforward rules in its immediate surroundings.

When large numbers of animals behave together, intelligence can emerge at the level of the group. Researchers at the Technical University of Berlin investigated this form of group intelligence to understand how co-operation produces organised behaviour in nature and can support the design of improved human systems.

Collective intelligence of animals

An individual ant possesses only a small amount of information. Its field of vision is restricted and its memory is brief. Nevertheless, an ant colony can locate food, choose effective routes and adjust when conditions change suddenly.

No central figure gives instructions. Instead, every ant obeys simple principles, including following chemical trails and preventing collisions. These modest individual behaviours can produce problem-solving at the group level.

Scientists refer to these results as collective intelligence. Solutions at group level develop through local interactions, rather than through advance planning or central command.

Studies of collective behaviour examine how uncomplicated rules generate intricate patterns over time and across space. Ant trails, bird murmurations and fish schools are all governed by related principles.

A field shaped by many disciplines

Collective animal behaviour brings together biology, physics, mathematics and artificial intelligence. A central area of research is self-organisation, where order develops without any central control.

A recent perspective paper maintains that future advances will require closer collaboration between these disciplines.

Its authors, Valentin Lecheval and Pawel Romanczuk, are specialists at the Institute for Theoretical Biology (ITB) at Humboldt University of Berlin in Germany.

They also belong to Berlin's Science of Intelligence Cluster of Excellence (SCIoI).

Research at this centre considers intelligence to be an outcome of interactions among animals, groups and other systems, rather than a quality confined within individual brains.

Why new tools are transforming research

The rapid development of technology is taking collective behaviour research into a new stage. Automated tracking systems can monitor every animal within a group simultaneously.

Drone footage records movement across extensive areas. Artificial intelligence (AI) identifies concealed patterns within enormous data sets. Virtual reality experiments make it possible to control environments precisely during animal research.

These technologies enable researchers to study complete collectives rather than individual animals in isolation. Detailed data now shows how local interactions create patterns on a larger scale.

Technology by itself, however, cannot resolve the more fundamental questions. Lecheval and Romanczuk state that progress requires common concepts and models spanning different disciplines.

“The many challenging questions of collective behavior can only be tackled with multiple theoretical frameworks and novel empirical approaches,” noted the authors.

Understanding animal interaction

One significant challenge is identifying the rules that shape interactions between individuals. Fish schools offer a useful illustration.

Their movement is driven by attraction, alignment and repulsion. Recent approaches use AI to derive interaction rules from movement data.

Most research, though, examines only a limited number of species living in laboratory settings. This narrow species coverage makes comparisons more difficult and limits insights into ecology and evolution.

Future studies seek to use common methods across a wider range of species. This would permit comparisons between different habitats and evolutionary histories.

Combining fieldwork with laboratory experiments is also becoming increasingly important, particularly for systems that are difficult to reproduce indoors, such as extensive ant networks and huge fish schools.

Variation within groups

Members of a group are seldom identical. They may differ in size, energy level, social role and capacity to learn. These differences can shape both group movement and decision-making.

Baboon research indicates that smaller animals face greater energy costs when travelling with a group. Bird studies show that pair bonds affect movement patterns and the flow of information.

Future models aim to account for this diversity instead of treating all members as the same. Creating models with this degree of realism demands close co-operation between data gathering and theoretical development.

Connecting movement and decision-making

Collective movement and collective decision-making are often studied as separate issues. Movement models explain how groups remain together.

Decision models address how groups select routes or shelters. In reality, animals combine these processes. Searching for food, avoiding predators and migrating all involve movement directed by goals and information.

Lecheval and Romanczuk advocate integrated models that link movement and decision-making over different time scales. Artificial intelligence has an expanding role in this work.

AI models can learn movement strategies according to their performance in activities such as foraging or survival. These models assist in testing theories about animal memory, perception and learning.

Beyond animals alone

Concepts from collective behaviour are now influencing numerous areas. Swarm robotics employs basic robots that co-operate without leaders.

Potential future uses include environmental monitoring, agriculture, search and rescue, and removing microplastics. Studies of collective cell movement also contribute to cancer research.

Architects apply collective behaviour models to enhance crowd evacuation and building safety.

Intelligence in animal interactions

Research into collective behaviour now provides a wide-ranging framework for examining co-ordination and intelligence in both living and engineered systems.

Ant colonies, robot swarms, groups of cells and human crowds all operate through common principles of interaction and self-organisation.

By combining biology, physics, mathematics and artificial intelligence, future research offers the prospect of greater insight into how simple actions can create intelligent outcomes without leaders.