An important step in predictive maintenance is to algorithmically prioritize the predictions so that a guided scaling of predictive actions can be followed. Also, if impact of a predicted fault can be estimated, sufficient importance or urgency can be associated with predictive action.
We have developed a policy based algorithmic prioritization of predicted faults along with qualitative or quantitative assessment of impact.
Algorithmic prioritization of predicted faults removes the barrier of closed loop predictive operation by automating the selection decision through algorithm. This is an important milestone in predictive operation since it directly affects organizational ability to extract business value by focusing on prevention of most important problems.
Since different organizations can have different goals for predictive maintenance, algorithmic prioritization provides policy based driven selection criteria. One may want to reduce incidents, another may want to improve customer experience or to prevent hardware related faults. Setting the correct focus is just a matter of choosing one among the many off-the-shelf policies available in the application.
Additional policies can also be configured to handle specific operational requirements, thus providing a flexible yet automated way of scaling predictive operation. It’s also possible to select a combination of multiple policies to provide a granular control on the focused area of operation.
