Real-time problem solving is an interesting topic in planning in the recent years. Besides discussing the differences between real-time scheduling and off-line scheduling, a real-time scheduling framework is proposed and some pivotal technologies are discussed in this paper. This framework can deal with uncertainty in real-time environment and output approximate solution in given time. Based on this framework, an intelligent assembly robot system is built, and through some experiments this framework is proved to do well in real-time environment. After all, the conclusion is given out, and the policy of real-time planning problem solving algorithm simply and the possible developments in the future are discussed in this paper.