Structural Health Monitoring (SHM) has recently emerged as a useful tool for tracking the performance parameters of a structure such as strain, deflection, and acceleration through a series of sensors installed on them. The signals produced from these sensors are the main performance indicator of the structure. In assessing the condition of the structure, the proper analysis and the evaluation of changes in pattern of signals are the most important tasks in SHM. Another important aspect of SHM is the detection of the defective sensors. And it is very difficult to identify it manually from a series of sensors. Although it is an important task in SHM but no straightforward method exists currently to carry out this task. In this study, the sensor data from a Canadian bridge have been utilized here to develop Artificial Neural Network (ANN) and Wavelet Transform (WT) based methods for tracking the changes in sensor data pattern and detecting the defective sensors in SHM. The ANN structures are constructed with input nodes accepting data from selected strain gauges and a target selected from the remaining strain gauges. The data collected at different time periods are de-noised by WT and tested against the trained network to find the pattern of differences between the input and output data series. The proposed methods have been validated with the available data and are found to be effective in tracking the data patterns and detecting defective sensors.