We review the current state in developing intelligent
methods to diagnose and maintain mobile mechanical
machinery. At first the problem area is characterized and
the different architectural layers are identified. Then
we review necessary preprocessing methods and methods for
fault detection and identification and outline how an
active database platform, called RapidBase, could be used
for this task. Next we address the task of building the
diagnostic models. We review methods to estimate the
remaining life time of components. We address different
methods for industrial maintenance management from
failure-based maintenance to condition-based maintenance.
Next we identify potential remote diagnostic services.
Finally a roadmap is suggested prioritizing the industry
needs in this domain.
|Series||VTT Research Report|