Development of a robot that cognizes and learns unknown events
Kushiro Kyohei, Junichi Takeno · Frontiers in artificial intelligence and applications · 2011
The present paper proposes a system to identify and learn about unknown events using the Module of Nerves for Advanced Dynamics (MoNAD) system. This system identifies unknown events using the difference in MoNAD convergence velocities between known and unknown events. The system was mounted on a robot and color identification tests were conducted to evaluate the feasibility of the system. In experiments, the robot autonomously identified unknown events, and automatically captured them as new information by itself. The present paper introduces how the robot proactively identifies unknown events, and captures them as new information. The system presents a solution to the problem of “Why do humans always learn new events and adapt to them?”