Lego robot with an organic “Brain” learns to navigate a maze by Saugat Bolakhe (Scientific American, Jan 28, 2022)

Link: Lego robot with an organic “Brain” learns to navigate a maze by Saugat Bolakhe (Scientific American, Jan 28, 2022)

1B40476A-F31B-4C43-90F946B930BABE16_source

Summary: In a recent investigation, a research team at the Max Planck Institute for Polymer Research in Mainz, Germany, has designed a carbon-based neuromorphic computing device—essentially an organic robot brain that functions like an animal’s—that can learn to navigate and solve a maze. The robot’s future steps were not pre-programmed, but with each wrong move, it could learn from its prior mistake and make a different decision to change its direction. After navigating different permutations of paths, the robot was able to escape. In the coming years, these low-power, organic neuromorphic systems could have many applications, like helping robots work for long hours in remote places on Earth without constantly needing to recharge – or even on other planets. They could help treat certain diseases and injuries to the nervous system, and neuromorphic implants may even enable humans to control powered exoskeletons.

Audience : This article may be interesting to Neuroscience students, as well as people interested in AI, ML, and Robotics.