Railway companies and electronics manufacturers have been dealing with various energy-saving activities, such as improving efficiency of regeneration brakes that uses the braking force as the power source, and modification of vehicle configurations along with various efforts for safer, higher speed and further comfortable railway system(Fukazawa, et al., 2010 and Seki, 2011). Since 2012, Tokyo Metro and Mitsubishi Electric have carried out field tests repeatedly for energy-saving operation control using the Automatic Train Operation system (ATO) and ensured that the power consumption reduced at the maximum by approx. 18% during the test run according to the speed profile, and no less than 10% in commercial operation which involves the approaching and recovery operations. Based on the result, we are currently implementing energy-saving activities to some of the commercial lines, and in the future we are planning to further analyze long term field data (Yoshimoto, et al., 2009 and Yamamoto, et al., 2012). In addition, authors considered proceeding with technical-field tests to optimize the train control mechanisms' performance using the self-learning function of the vehicle characteristics as an attempt to measure higher operation performances. This self-learning function aims to follow up to the changes of braking characteristics of the vehicle due to external factors, and to maintain the stop position accuracy despite such changes. This function also improves the tracking performance of the speed profiles by considering the energy-saving effect, providing further time allowance, and making passengers feel more comfortable while boarding The summary outlines of energy-saving operation control and the self-learning function of the vehicle characteristics are introduced in this paper.