Robot swarm from Harvard uses a 3D vision‑based coordination system to move in sync while swimming.


  • Collaborative
  • Autonomous
  • Dynamic


  • Search and rescue
  • Surveillance
  • Security

UN Sustainable Development Goals Addressed

  • Goal 9: Industry Innovation & Infrastructure

The Challenge

Robots are often designed to replace or help humans perform tasks. Robots can be programmed to complete these tasks, but still need input from the human to make dynamic decisions. Although many robots are able to move on their own in two-dimensions, they encounter issues when moving through three-dimensional spaces such as air and water. This is because of the extra energy required for sensing and locomotion in these areas. Additional human input reduces the number of tasks a robot can perform on its own in remote, hard-to-reach places.



Innovation Details

The robot swarm is made up of individual robots, called Bluebots, that have three blue LED lights and two cameras on each ‘fish’. The cameras can detect the LED lights from surrounding robots and use an algorithm to determine the distance, direction, and heading of neighboring robots and actively react to their positions. This allows the Bluebots to autonomously self-organize similarly to schools of fish. They can aggregate by calculating the positions of their neighbors to move towards the center, or can do the opposite to disperse. They can also follow the lights directly in front of them in a clockwise direction to swim together in a circle.

video thumbnail

Biomimicry Story

Certain species of fish are able to swim together in groups known as ‘schools’, and can perform complex, synchronized behavior without following a leader. Instead, they use signals from their neighbors to make decisions and coordinate their movement. Some fish vibrate their fins in certain subtle ways to signal danger. Other fish use the glimmer of light off their scales to signal a change in direction.