Harvard robot inspired by fish swims in a school

Shane McGlaun - Jan 18, 2021, 5:57am CST
Harvard robot inspired by fish swims in a school

Fish swimming in a massive school have complex and synchronous behaviors to help them find food, migrate, and evade predators. The school of fish isn’t coordinated by a single fish or a team. Fish also don’t communicate with each other about what to do next. How this swarm moves comes from something known as implicit coordination happening when an individual fish makes decisions based on what their neighbor is doing.

A team of researchers from the Harvard John A. Paulson School of Engineering and Applied Sciences and the Wyss Institute for biologically Inspired Engineering developed robots inspired by fish able to synchronize their movements like a real school of fish would without external control. Researchers say this is the first time they have demonstrated complex 3D collective behaviors and implicit coordination in underwater robots.

Researchers say that robots are frequently deployed in areas inaccessible or dangerous to humans, where human intervention may not be possible. There is a significant benefit in those situations to having autonomous robots that are self-sufficient. The team used implicit rules and 3D visual perception to create a system with a high degree of autonomy and flexibility to operate underwater where GPS and Wi-Fi aren’t available.

The robotic swarm is called Blueswarm. Researchers used blue LED lights for their vision-based coordination system. Each underwater robot, known as a Bluebot, is equipped with two cameras and three LED lights. Each robot has onboard fish-lens cameras that can detect the LEDs of neighboring robots using a custom algorithm to determine distance, direction, and heading.

The system allows the robots to exhibit complex self-organized behaviors, including aggregation, dispersion, and circle formation. Scientists say that each robot implicitly reacts to its neighbors positions. The robots can perform a simple search mission using a red light in a tank. The dispersion algorithm allows the robots to spread out across the tank until one comes in contact with the light source, flashes its LEDs, and triggers the aggregation algorithm with all robots converging on the target.


Must Read Bits & Bytes