Predictive maintenance

Predictive maintenance uses equipment-condition data to estimate when failure or servicing is likely, so maintenance can be scheduled before breakdown while avoiding unnecessary routine work.

Condition monitoring: The continuous or periodic measurement of equipment parameters to assess its operating condition. It supplies the observations from which predictive maintenance detects developing faults.
Industrial Internet of Things: Networks of connected sensors, machines, and software used to monitor and control industrial operations. It provides the connected infrastructure for collecting equipment-condition data at scale.
Condition-based maintenance: Maintenance performed when monitoring indicates that equipment condition has reached a defined threshold. It uses observed condition to trigger work, while predictive maintenance additionally estimates future needs.
Preventive maintenance: Maintenance performed at planned intervals or usage thresholds to reduce the likelihood of failure. It schedules work by elapsed time or use rather than forecasting need from condition data.
Vibration analysis: The measurement and interpretation of mechanical vibration to identify equipment faults and operating changes. Changing vibration patterns can reveal imbalance, misalignment, and bearing damage before failure.
Time-series analysis: The analysis of observations ordered in time to describe patterns and forecast future values. Equipment sensor readings form time series whose trends can precede failure.
Reliability-centered maintenance: A strategy for selecting maintenance tasks according to how equipment functions can fail and affect operations. Its failure-focused task selection can incorporate predictions of when intervention is worthwhile.
Reactive maintenance: Maintenance performed after equipment fails or an unacceptable fault becomes apparent. Predictive maintenance seeks to intervene before the breakdown that triggers reactive work.
Remaining useful life: An estimate of how long an asset or component can continue operating before it fails or no longer meets requirements. Maintenance plans often use this estimate to decide when intervention is due.
Survival analysis: Statistical methods for analyzing time until an event, while accounting for observations whose event time is unknown. It can estimate failure risk when many assets have not yet failed during observation.

Linked from 9 pages