Compare like with like: the same equipment references, the same measurement points and a recorded load for every image. Then a finding can be tracked year on year as a measured temperature, a difference in °C from a matching reference, and the load at the time. Repaired items should be re-imaged and closed out, not carried forward.
