Abstract: The urban lifeline safety project is a major national livelihood project, which ensures the safe operation of urban gas, water, electricity and other systems. Nowadays, the project is confronted with three types of difficulties: Difficulties in mechanism discovery and experimental reproduction; difficulties in risk detection and early identification, and difficulties in accurate early warning and collaborative prevention and control. In order to reveal the mechanism and reproduce the experiments, a full-size urban lifeline multi-hazard large-scale scientific device is firstly developed in the world, a series of dis-aster modes with 5 categories and 169 kinds of disasters coupled are established, based on which a comprehensive risk dynamic assessment method coupled with underground pipelines and above-ground disaster-bearing body is presented. To improve the risk detection and early identification, highly sensitive gas sensing laser chip as well as detector with active grating and wavelength-biased are invented. We develop an intelligent ball for leakage located in water supply pipe network based on the inertial guidance, force balance, acoustic spectrum and BeiDou. Bridge overall modal (fingerprint) monitoring technology is break-through. In order to solve the problem of accurate early warning and collaborative prevention and control, an urban lifeline safety risk physical heuristic artificial intelligence early warning technology is presented, based on which a monitoring and early warning system is developed with collaborative sensing, fusion monitoring, targeting early warning, linkage response and other functions. The related studies result in the development of the safety and emergency response industry, followed by the pioneering and substantial contributions to the prevention and control of urban lifeline risks.