Owing to the cruising scope of the mobile robot is conspicuously restrained by the heavy weight and the discharge capacity of the battery pack, power management and charge/discharge reliability are getting widespread attentions on the intelligent robots in the past years. In general, battery capacity and the state-of-discharge (SOD) are pessimistically estimated based on the upper bounds of each robot modules. This paper is aimed to develop an inexpensive, stable, and distributed power scheduling algorithm (PSA) to evaluate the on-line performance on the humanoid robot. A PC-based platform with the human-machine interface (HMI) will be applied to detect each of the modulized servo-motors and sensors. Finally, off-line simulations and experiment of the on-line data acquisition are also carefully examined and successfully demonstrated the advantages of the proposed PSA system.