Robot Mother Builds Children and Mimics Natural Selection

Cambridge University scientists have made a “mother” robot that creates cube robot children, evaluates them based on task performance, and improves the next generation accordingly.

ROBOT TESTING: IF YOU THOUGHT YOUR MOM WAS BAD…

Equipped with a Universal Robots UR5 robotic arm and gripper, the mother robot built and tested ten generations of “cube babies” in five experiments. Designing and creating each robot took about 10 minutes for the mother, who designed them based on five rules guiding their construction, motor command and shape.

ONLY BEING HERCULES IS GOOD ENOUGH?

Each cube robot was tasked to travel the longest possible distance from a starting point in a certain amount of time. Depending on their performance evaluations, some of the “children” underwent no changes in the successive generation and others had their “genomes” altered and mutated.

The researchers found that the mother was adept at improving each generation so that the later generations would not have the genomes deemed defective. The last robots the mother created performed the task twice as quickly as the first generation.

As its lead researcher pointed out, the study effectively imitated natural selection.

PROGRAMMABLE NATURAL SELECTION TO MAKE DARWIN PROUD

“Natural selection is basically reproduction, assessment, reproduction, assessment and so on,” Dr. Fumiya Iida, the project’s lead researcher, explained in a press release. “That’s essentially what this robot is doing – we can actually watch the improvement and diversification of the species.”

The project was completed at Cambridge in collaboration with researchers from ETH Zurich. It’s an example of evolutionary robotics, “a growing field which allows for the creation of autonomous robots without human intervention” according to Cambridge University.

The origin of intelligence is a mystery to many biologists, and a question that Iida and his team were interested in exploring. “One of the big questions in biology is how intelligence came about – we’re using robotics to explore this mystery,” said Iida, a lecturer at Cambridge’s Department of Engineering. “We think of robots as performing repetitive tasks, and they’re typically designed for mass production instead of mass customisation, but we want to see robots that are capable of innovation and creativity.”

You can read the full results of the study in the open-access journal PLOS ONE.


 

ON INTELLIGENCE… PLAY WITH A.I. WITH MUSIO: