A main aspect of analyzing sensor data is to find dependence structures or to verify the lack of those. In most cases, the first idea is to use Pearson’s correlation coefficient. However, this correlation coefficient is restricted to linear dependencies, which only account for one variety in a large field of possible interrelations. Moreover, sensor data is not always stationary. The concept of distance correlation as introduced by [1] and its extension by [2], the so-called local distance correlation, respond to both of these problems. They are not only not restricted to linear dependence but are also able to detect independence and do not require normality. We propose different examples of application in the field of bridge monitoring from finding similarities and anomalies in sensor outputs over testing for independence up to possible alarm concepts for long term surveillance.