During past ten years, airtraffic in Ch na has experienced a rapid rise. As a result, the work loads of air traffic controllers increased greatly and therefore controllers’ mistakes that threaten fight safety happened more frequently. Although TCAS is a successful technical innovation for pilots to enhance flight safety, it is still lack of anautomatic system for air traffic contro lers, which can monitor real time operations of air traffic control and give an alert to human controllers on probab ly control errors, act ing as an in telligent supervisor. The two main reasons are the complexity of some core technology and lack of participation of system scientists and control engineers. This paper presents the architecture of Air Traffic Control Comm nd Monitoring System (ATCCMS). This system is designed fo rair traffic controllers to reduce wrong commands and to enhance flight safety.The core design idea is as follows: Firstly, ATCCMS in tegrates several kinds of fundamental information, including radar data, flight plan, ground?a irvoice communication,meteoro logical data and en route situation data. Secondly, correlating to radar data and expert know ledge,ATCCMS acquires the content of ground?a irvo ice communication based on up todate speech recognition and natural language processing techno logy. Thirdly,ATCCMS estimates futuret rajectory of aircraft and detect spotential conflicts based on the recognition of human controller’s voice command. Simulation experiments show that ATC safety w ill be effectively improved through information integration approach rather than upgrading individual technology. The proto type of ATCCMS is under developing cooperated with North China Air traffic Management Bureau, CAAC