Japanese researchers have developed a robot with neurons similar to those found in the human brain and developed them in a laboratory to teach them to “think like humans”.
During a test at the University of Tokyo, the Daily Mail reported that the small robotic vehicle on wheels was small enough to fit in the palm of a hand, placed inside a simple maze.
The robot connected the network of brain neurons that grew from the cells, and when these artificial neurons were electrocuted, the machine hit its target – a black circle. When the robot goes in the wrong direction or runs in the wrong direction, the neurons in the cell culture will jam with electrical stimulation to bring them back into the path.
These experiments, described in a new research paper published in the Applied Physics Letters, are a major step in the effort to teach robots intelligence, according to researchers, especially as it is the first time intelligence has been “taught” to someone. Robotic robot using neurons grown in the laboratory from cells.
In their study, the authors stated: “We created a closed-loop system to automatically generate a synchronous signal from the live neural network, which is automatically activated, and adapted to the network using a mobile vehicle robot when the robot strikes obstacles or is not within 90 degrees in front of it. Electrical stimulation was used from one electrode so that the robot could successfully reach its target in four different areas. “
Artificial neurons grown from cells acted as a ‘body repository’ for robot decisions.
During the test, the robot is given internal balance signals, everything is going to be planned and it is advancing towards the goal.
However, if the robot encounters an obstacle, this balance is disturbed by the disturbance signals, causing the robot to vibrate and reset.
During the tests, the robot was constantly given constant symmetrical signals that were interrupted by interference signals so that it could successfully solve the maze task.
The robot was unable to see the environment or receive other sensory information, so it relied entirely on trial and error electrical stimuli.
Researchers have demonstrated that intelligent task-solving skills can be developed by “physical reservoir computers” – a body that performs calculations based on brain signals.
Hirokasu Takahashi, Associate Professor in the Department of Mechanical Information, University of Tokyo, said, “Our experiments have inspired me to consider that intelligence in the living system arises from a mechanism that extracts a coherent output from an unorganized or chaotic state.
Advances in the computation of physical reservoirs may contribute to the development of artificial intelligence machines that think like us.
The team hopes that the use of physical reservoir computing in this context will contribute to a better understanding of the brain’s mechanisms and lead to the development of the neurological system.
A neurological system can simulate the neural structures found in the human nervous system.
“Award-winning beer geek. Extreme coffeeaholic. Introvert. Avid travel specialist. Hipster-friendly communicator.”