Abstract: Existing early-warning systems of coal mine earthquake run independently in different mines. There is serious repeated investment and no correlation between these similar systems, and on-site analysis ability on massive coal mine earthquake data is clearly insufficient. Key technologies of a SaaS based remote early warning system of coal mine earthquake was researched based on Sensor Mine Internet of Things Demonstration Project(Jiahe Coal Mine). Overall architecture of the system was given and three key technologies of the system were discussed including structure of SaaS-based expert analysis software, construction of multi-source SaaS database and MapReduce-based parallel data filtering and receiving. The application showed that the system basically has structure of remote early-warning system of coal mine earthquake in the whole country and can meet basic requirements of remote early-warning of multi-mine earthquake.