Damage that originates from abnormal wear in the cylinder liner of marine diesel engines causes a considerable loss of property to manage shipping vessels. However, because multiple factors contribute to such damage, it is difficult to anticipate the damage that is caused by abnormal wear, such as scuffing, by investigating the mechanism of abnormal wear. Therefore, several methods have been explored for predicting the abnormal wear between the cylinder liner and the piston ring in marine diesel engines. For example, methods that are based on an analysis of the temperature, vibration, or cylinder drain oil have been researched. However, the response time of such methods is too slow for an operator to have enough time to promptly cope with severe damage. The implementation of such methods also requires prior modifications to the engine, which cost time and money. To overcome such problems, methods of prediction that use AE (acoustic emission) have been widely investigated. Studies on the relationship between abnormal wear and AE signals have demonstrated that severe damage is preceded by a change in the RMS value of AE and the FFT amplitude of a specific frequency. However, previous studies on the AE technique were based on the offline analysis of stored data due to their focus on the relationship between abnormal wear and AE signals. Thus, for direct implementation of the AE technique in industry, a realtime diagnostic system is needed. This paper focuses on the development of a real-time diagnostic system for analyzing high speed AE signals and examining the wear status of cylinder liners in marine diesel engines.