Prediction of high-risk operation time for new on-the-job training in anthropocentric cell manufacturing system

This paper aims to propose the core concept of a high-risk operation time prediction based on human cognitive science and information technologies for new OJT using an artificial neural network. The prediction intends to notice an educator who sits in the remote site and alerts him to concentrate on a learnerpsilas operation on the shop floor. The high-risk operation in New OJT is analyzed and integrated into the human tension extent. Then alpha-brain wave, a frequency level of the highest attention to stimulant, of the educator is measured by an electroencephalography in order to investigate educatorpsilas tension extent. The result shows that educatorpsilas tension extent is 60 seconds after stimulation. Finally, the neural network is designed for the prediction of high-risk operation time in new OJT. Twenty eight samples are used for neural network training and experiments during new OJT. The results reveal that the proposed neural network model performs high precision of prediction with 97.65% confidence and all of the predicted high-risk operation times are within the interval of educatorpsilas tension extent.