For engineering systems, a diagnosis of the asset's state of health is performed. cbmLAD uses the data already gathered during usage, analyzes it carefully and efficiently, and gives fact based diagnosis of the state of degradation.  cbmLAD searches for recurrent patterns of bad performance and identifies the responsible parts by their serial numbers. For situation where a controlled design of experiments cannot be performed, the inter-reactions between different factors affecting productivity are understood and managed. cbmLAD is a powerful tool for learning the hidden relations between system's elements, and how they inter-react.