From the last two decades, lots of research have been done on the non-destructive testing (NDT) and structural health monitoring (SHM) techniques. Though the conventional NDT methods are effective, the complete reliability on them is not possible due to the countable time period. For all the catastrophic failure, time plays an important role. If appropriate time of structural and mechanical failure will not be estimated, it will be disastrous. So, in this paper, method describing on online technique has been proposed. An evolutionary algorithm has been used to generate an online method to identify and detect the damage. In the application of any evolutionary algorithm, large numbers of data are required. It has been observed that during the collection of data, error is accumulated. Error needs to be reduced. So, to minimize the error amount during the data collection, the relation between dependent variables and independent variables has been figured out. In this work, the result from the genetic algorithm (GA) and the auto-fitted GA has been compared. From the comparison, it can be clearly noticed that the results from the auto-fitted GA are more convergent and the percentage of error is less. The novelty of this study is the application of hierarchical GA with the statistical bootstrap method for damage detection and its severity measurement.