In the case of some elderly or disabled persons, not only the motor ability, but also the environment perception ability is sometimes deteriorated. To assist the daily living motion of those people, power-assist robots with the perception-assist have been proposed. The power-assist robot with the perception-assist assists not only the user's motion but also the user's interaction with an environment, by applying the modification force to the user's motion if it is necessary. Since it is difficult for the robot to prepare all proper perception-assist for every task, tool, and environment previously, the robot needs to learn the proper perception-assist for each task, tool and environment by itself. The effectiveness of the performed perception-assist by the robot has been judged by the EMG signals. However, if the EMG signals do not change enough for the judgment, the learning of the robot might not succeed. In this paper, both EMG signals and EEG signals are measured at the same time to observe the features of these signals when users use the power-assist robot. EEG signals are used as the criteria of the effectiveness of the performed perception-assist in addition to EMG signals.