A real-time scheduling method considering human operators has been proposed to generate suitable production schedules for human operators, manufacturing equipment and jobs real-timely in the autonomous distributed manufacturing systems (ADMSs). The proposed scheduling method gives the higher priorities to the human operators, and generates the production schedules based on preference values, in order to improve the human operator's motivation. It was shown, through case studies, that the proposed real-time scheduling method is effective to satisfy the preferences of the human operators. However, the make span is deteriorated especially for the cases where the human operators have similar preferences. Therefore, a new method is proposed to calculate the preference values considering not only the preferences of human operators but also improvement of productivity.