The research in the Adaptive systems group is centered around autonomous (i.e. freely moving) robots. The ultimate aim of our research is to generate autonomous robots capable of carrying out a variety of relevant tasks, particularly dangerous or tedious tasks that are presently carried out by people.
Research towards intelligent robots
The field of autonomous robots is rapidly growing, especially in Japan (and other East Asian countries) where giant leaps have been taken in recent years, regarding robotic hardware. However, the problem of providing autonomous robots with reliable overall behavior is very difficult indeed, and the problem is far from solved. In the Adaptive systems research group, our main consideration is decision-making in autonomous robots, using a decision-making method (based on the concept of utility) that has been developed within the research group.
Detecting drowsiness in drivers
In addition, our research also concerns stochastic optimization methods and their application in various fields. For example, in our current work, we are considering the problem of detecting drowsiness in drivers, based mainly on driving behavior. The problem has been formulated as a classification task, in which stochastic optimization methods are applied to generate a system capable of estimating whether or not a car or truck driver is drowsy, based on the variation in lane position and other indicators.