Due to the high degree of informatization in cigarette factory and the large amounts of collected production data, this thesis focuses on how to effectively use these production data to monitor the abnormal information during the process of production. In previous monitoring, managers make adjustments to abnormal situations by watching the production board or artificial feedback. In this way, the response speed of abnormal situations is slow and leaky. Therefore, this paper proposes a method of abnormal information analysis to obtain effective abnormal information during the production of rolling and packaging workshop, meanwhile using mobile platform to push these effective abnormal information to relevant manager. By this way, the response ability and the control level of production can be improved.