A Study on Inference of Cognitive Load by Electroencephalography toward Skill Acquisition for Process Planning.

It is difficult to acquire the skills related to process planning from interviews with skillful operators due to complicated decision-making processes. On the other hand, electroencephalography (EEG) has attracted considerable attention recently to reveal complicated decision-making processes. Therefore, this study aims to infer the cognitive load of the operators by EEG for skill acquisition. For this purpose, multiple types of preliminary tasks are prepared with different difficulty levels. Subsequently, a machine-learning model using a linear support vector machine is constructed to infer cognitive load, which varies according to the difficulty level, from feature vectors calculated by common spatial pattern method. The machine-learning model is applied to time-series EEG data obtained when understanding a lot of mechanical drawings. Consequently, it is recognized that cognitive load varies according to the complexity of drawings. Therefore, the inference results demonstrate the ability to identify time-varying cognitive load and to support the interviews toward skill acquisition.