Article Preview
TopIntroduction
Robots, considered as a key technology of the fourth industrial revolution, are being developed across diversified areas in response to the demand for improvements to the quality of life in contemporary society. The robot industry can be classified largely into 3 key areas, such as manufacturing robots, specialized service robots, and personal service robots (Kim, 2016). Among these, personal service robots are designed to provide services to individuals, with the principal agent of operation being the individual themselves. As a result, convenient interface methods, as well as accurate motions desired by the user are important (Chu et al., 2010).
“Interface” refers collectively to physical media or software programs that transmit information between two systems. Namely, the interface is a concept covering all physical and virtual media that are produced to enable communications between things, or between humans and machines (computers) (Chun, 2011). Interface technology is being developed in diverse ways. However, most of the interface technologies currently in development are studies involving healthy members of the general public as subjects. For this reason, most interfaces currently in use cannot operate efficiently for the users with physical impairments. Thus, separate interface methods are required for the physically disabled, or those with a physical impairment.
It is efficient to utilize interface technology for users with physical limitations to assist these users to better manage their handicaps. That is why interface technology utilizing bio signals is generally utilized for people with inborn physical impairments. From such a viewpoint, BCI (Brain Computer Interface) can be used very efficiently. BCI refers collectively to the interface technologies that directly connect the human brain with computers (Park et al., 2016), for control of the computer through EEG signals, and can be considered the ultimate interface operated by mind control (Chun, 2011; Park et al., 2016). Interfaces using the BCI technology also provide a convenient control method to not only the physically impaired, but also to general users.
EEG (electroencephalography) an electrophysiological monitoring method to record electrical activity of the brain (Wikipedia, 2020). EEG procedures have an advantage in that measurement costs for bio signals are reasonable, and brain wave information can be provided in real time without harming the human body.
EEG signals are displayed in diverse ways, depending on the user’s emotions or thoughts. When the user maintains specific thoughts and emotions, similar signals are repeated, and such characteristics can be used for robot control (Lee & Shi 2020). Induction methods for the EEG signals are diversified, different characteristics corresponding to every signal-inducing method exist, and the measured signals vary depending on the method used. Despite this fact, nothing has been presented regarding signal induction methods used for interfaces utilizing EEG signals. Hence, accuracy tends to be reduced, since signals are induced by their own methods (Jeon, 2014).
Thus, as outlined in this article, EEG signals are extracted by utilization of the image association technique, based on vision, hearing, gestures, the phonation association technique, and movement association technique, to standardize the induction method for EEG signals in the use of the interface utilizing EEG signals. An analysis is performed as to which method EEG signals should be induced by among the extracted signals to be most efficient for the robot control. When the analysis according to each method is performed, only the basic frequency conversion without special pre-treatment for the extracted EEG signals is performed to eliminate the factors exercising a subsidiary influence, and the analysis of recognition performance utilizing this is implemented. In this way, the derived results of recognition performance are made to reflect only the differences as a function of the induction technique for EEG signals. The induction technique derived in this article is expected to serve as a guideline to be presented to the subject when the interface utilizing the EEG signal induction technique optimized throughout diverse areas is realized in the future.