P300-based Control for Assistive Robot for Habitat

The lack of caregivers in the field of nursing care is a serious problem in Japan, given the progressively declining birthrate and aging population. Because the lack of personnel in nursing care also affects qualitative aspects, careless care will lead to a decline in quality of life (QoL) of the care recipients. Therefore, with logistical support, such as from robots and ICT, it is expected that the burden on caregivers can be reduced by increasing the instances where assistance of caregivers is not required, and that the QoL of the care recipients themselves, can be improved by increasing the number of things that they can do by themselves. In this paper, we propose a nursing care support robot system, especially for care recipients who have lost motor function of their limbs or have disabilities. This system adopts a BMI (Brain Machine Interface) using Electroencephalography (EEG), and by using "without moving limbs" brain wave information of disabled care recipients, the robot's arm can be controlled to perform simple movements such as picking up objects. According to performance evaluation results of the system with 5 healthy 21-24 years old males as the test subjects, a high recognition accuracy of 80% on average was obtained, when the stimulus count was 10 or more. By developing this system, it could become possible to switch on or off or control household electrical appliances or electronic devices in the vicinity and even receive meal assistance through brainwaves.