Scientists Teach Robot to Independently Navigate Maze

While processors act as the brain of the computer, they are fundamentally different from the actual human brain. Transistors use electronic signals to perform logic operations, while the brain relies on nerve cells connected via synapses. The brain uses this signaling to control the body and perceive the surrounding environment, and it uses a learning process to trigger the reaction of the body/brain system when certain stimuli are perceived. 

Steering a Robot Through a Maze

A group of scientists have now applied this basic principle of learning through experience in a simplified form, and they used it to steer a robot through a maze using an organic neuromorphic circuit. 

The team was led by Paschalis Gkoupidenis, who is group leader in Pail Blom’s department at the Max Planck Institute for Polymer Research. The work was the result of collaboration between the Universities of Eindhoven, Stanford, Brescia, Oxford and KAUST.

The paper was published in the journal Scientific Advances. 

Imke Krauhausen is a doctoral student in Gkoupidenis’ group and at TU Eindhoven, as well as first author of the published paper.

“We wanted to use this simple setup to show how powerful such ‘organic neuromorphic devices’ can be in real-world conditions,” Krauhausen said. 

The researchers fed the smart adaptive circuit with sensory signals coming from the environment in order to achieve the navigation of the robot inside the maze. At each maze intersection, the path towards the exit was indicated, but the robot usually misinterprets the visual signals and makes the wrong decision before getting lost. 

Applying Corrective Stimuli

As the robot makes these wrong decisions and ends up at dead-end paths, it is discouraged to take them by receiving corrective stimuli. For example, if the robot hits a wall, the corrective stimuli are directly applied at the organic circuit via electrical signals induced by a touch sensor attached to the robot. 

The robot then gradually learns to make the right decision with each execution of the experiment. This helps it avoid receiving corrective stimuli, and it eventually finds the right way out of the maze. The learning process takes place exclusively on the organic adaptive circuit. 

“We were really glad to see that the robot can pass through the maze after some runs by learning on a simple organic circuit. We have shown here a first, very simple setup. In the distant future, however, we hope that organic neuromorphic devices could also be used for local and distributed computing/learning. This will open up entirely new possibilities for applications in real-world robotics, human-machine interfaces and point-of-care diagnostics. Novel platforms for rapid prototyping and education, at the intersection of materials science and robotics, are also expected to emerge.” Gkoupidenis says.