SOINN robot, developed by a research group at the tokyo institute of technology to model human thought and decision-making

a research group led by osamu hasegawa at the tokyo institute of technology has presented a robot programmed with reasoning processes akin to those of human thought and capable of learning new tasks on its own. it uses an algorithm called SOINN (self-organizing incremental neural network’), a subset of pattern-based artificial intelligence, to adapt to situations and continually learn new information. the robot demoed here takes visual, auditory, and tactile data as input. when confronted with a new task, it uses its past experiences, in conjunction with sensory input, to determine how best to behave in the situation. it can also ask for help and communicate via wireless internet with other robots to share knowledge.

associate professor and research group leader osamu hasegawa asserts that the SOINN algorithms are very light on computation, despite the possibilities that the technology opens up. by way of example, hasegawa offers:

‘for example, suppose this robot doesn’t know how to make tea, and it’s sent to an elderly person who lives alone. and suppose that person asks it to make a cup of green tea. the robot doesn’t know how, so it asks robots around the world how to make tea. suppose, for example, that a robot in the UK tells it how to make bitish-style tea. we think this robot will become able to transfer that knowledge to its immediate situation, and make green tea using a Japanese teapot.’

read more information on SOINN, the robot’s pattern-based artificial intelligence (PBAI) system.

video of the SOINN robot in action, plus a discussion with osamu hasegawa video © diginfo

SOINN robot mimics human reasoning the robot in the midst of a water-pouring task

SOINN robot mimics human reasoning detail of head

SOINN robot mimics human reasoning detail of hands

SOINN robot mimics human reasoning the algorithms (left) and visual input (right) at work in the robot’s ‘thought’

via diginfo