Like a school of fish or a flock of birds, computer robots can now organize themselves into a formation to achieve a specific goal – all without human involvement.
Harvard devised a swarm of self organizing Robots
Harvard University researchers devised a “swarm” of 1024 tiny robots that, “like a mechanical flash mob,” can assemble themselves into various shapes without requiring detailed instruction, a Wall Street Journal report noted.
To enlighten robots with a kind of “hive intelligence,” Dr. Michael Rubenstein and his research colleagues developed a computer algorithm allowing a very large group of robots to locate each other and collaborate on a task without detailed human involvement or instructions.
“No one had really built a swarm of this size before, where everyone works together to achieve a goal,” Rubenstein was quoted as saying. “It could automatically change shape to adapt to the task at hand. “You could have them build other robots out of themselves.”
Robots building other robots and self multiplying is the stuff of science fiction movies, the reality is just a single command, beamed to all robots simultaneously, sets in motion a process where the machines organize themselves and each participant plays a role.
Self organizing Robots given three primary bits of intelligence for task completion
To accomplish these tasks, each robot is given three primary bits of intelligence: how to maintain a sense for its relative position in the group; how to follow the edge of a group; and how to track the robot’s distance from where it started.
From these informational building blocks the computer brains, the size of approximately a penny in this experiment, create a system similar to driver ants, who live together in colonies of 20 million or more. While the ants are blind and don’t overtly have a leader giving instructions, they worth together to find food and are directed by chemical signals, smell and touch – all with a team spirit hard wired into their genetic code.
In a similar fashion, robots can communicate to one another, have an awareness of their surroundings, including the location of other robots in a formation.
“The beauty of biological systems is that they are elegantly simple and yet, in large numbers, accomplish the seemingly impossible,” Harvard computer scientist Radhika Nagpal, who worked on the project, was quoted as saying.
The report said the relatively new experiment one day may aid in oil spill cleanups, deep-sea ventures, military surveillance and planetary exploration. The report did not mention the use for security in repressing a domestic population or any of the realistic negative consequences.