The increasingly complex technology of automobile makes the traditional automobile maintenance technology can not meet the requirements of the development of the times, so the use of new technology has become the inevitable choice of automobile maintenance. Compared with traditional detection technology, data flow technology has many advantages and will be widely used in vehicle maintenance. First, data flow technology overview of the so-called data stream technology mainly refers to the electronic control unit, the actuator and the sensor, the data between the 3 components of the data parameters through the diagnosis of the interface and the final detection of fault technology. Two. The common methods of data flow analysis are realized by computer analysis, causality analysis, comparative analysis, comprehensive analysis and numerical analysis in the process of automobile maintenance. To improve the level of vehicle maintenance, we must strengthen our understanding of these methods. 1. the computer analysis method called computer analysis mainly refers to the maintenance personnel analysis the data in the car through the computer. When using the computer analysis method to analyze the fault, the staff should consider both the value of the sensor and the rate of influence. 2. causal analysis method of causal analysis mainly refers to the analysis of the interrelated data. There is a corresponding causal relationship between data in every control system inside the vehicle. Causal analysis is needed to find out the relationship among them. The causality analysis is an effective way to find out the fault. 3. the comparative analysis method of comparative analysis mainly refers to the method of comparative analysis of components inside automobiles. This method is mainly used in the case that the technical data is not sound and the standard data can be referred to. In the process of comparison and analysis, it is often used to compare with the same type of vehicle in detail. This way can effectively improve the maintenance level. 4. comprehensive analysis method is mainly used in the vehicle internal system sensor after a few fault codes, not easy to judge the true fault position. It is difficult to judge the true situation after a few signals are sent out by the sensor. At this time, the relationship between them should be analyzed by the comprehensive analysis method to determine the range of the fault. 5. the so-called numerical analysis method of numerical analysis is mainly based on the law of numerical change, and carries out a specialized analysis method for vehicle speed, speed and other indicators. In the actual maintenance process, this technology is mainly realized by the diagnostic instrument. Three, the maintenance example 1 vehicles with L13A3 engine, CVT transmission in the wide range of transmission of the transmission does not shift, the engine in the 6000r/min, the speed of only 30km/h failure. The maintenance personnel carry out a serious analysis. In general, the car needs to meet 4 conditions at the same time: the ignition switch is in the "ON" state; the brake pedal is stepped down; the throttle opening should be kept in a certain position; the shift lever should be in the "P" state. If one of these 4 conditions does not meet, the shift can not be achieved. The maintenance personnel first checked the brake pedal through the KT600 tester, and found that the brake pedal was in the open position. After checking the engine system, it is found that the throttle is closed, and the opening of the throttle is 12.5%. This data is obviously larger than the requirement. After thorough analysis, the maintenance personnel think the throttle sensor is damaged. After changing the throttle sensor and adjusting the position of the throttle sensor, the car can shift the gear freely. After a detailed test, the system is found to be normal and the maintenance is finished. This paper briefly analyzes the application of data flow technology in vehicle maintenance, and illustrates the simplicity and reliability of data flow technology combined with practical cases. In the future vehicle maintenance, we must strengthen the research of data flow technology.