Fault Detection and Isolation (FDI) methods that monitor the navigation system for sensor faults in real-time, can contribute significantly to improve system reliability. Quick detection and isolation of sensor faults can prevent serious damages and irreparable consequences. In this book, remarkable model-based fault diagnosis techniques, such as the Structural Analysis and the Parity Space Approach, are presented and successfully applied to detect and isolate navigation sensor faults in an autonomous navigation robot. The dynamical model of the robot is derived in linear and nonlinear state-space form. Structural analysis is performed on the robot nonlinear model to analyse its structural properties, and thus to provide the set of redundancy relations necessary for fault diagnosis. The required knowledge about weak/strong detectability of sensor faults is obtained applying the Parity Space Approach to the robot linearized model. The effects of the sensor noise are also accounted. In this case, the fault detection reduces to the problem of detecting a change in the mean of a normally distributed random sequence. FDI applications on multi-sensor navigation systems are also provided.