This thesis aims to identify and analyze specific patterns and temporal events at the rail-road tracks of the Viennese metro lines using the accelerometer sensor installed in smartphones. Examples of this are driven bends, acceleration, switches and track damage. For this purpose eight smartphones were purchased in different qualities, in order to also analyze the thereof resulting differences in their measurements. These eight mobile phones were used to carry out a series of test runs in Vienna's metro network. Their recorded sensor readings were then analyzed via various statistical methods, including correlation analysis, multiple linear regression and SCARM, a function in R for signal extraction from noisy and outlier-interfered data streams. The smartphone sensors all measure ap-proximately the same, but differ in their maximum sensor sampling rates as well as in their variances. Also, the gyroscope data of the mobile phones, which was also recorded, is very similar to the curvature recorded by Wiener Linien's rail test car. Based on the acceleration data, the speed can be estimated, which, together with the bends, allows a reconstruction of the driven path via sensor fusion and thus an approximate localization in the metro net. Switches can be detected with the accelerometers of the smartphones. However, this was possible only if the smartphone was rigidly installed in the train or on the floor of the passenger compartment. If the smartphone was held in the hands, switches and also slight track damage could not be identified, as the human body absorbs the vibrations too strongly.