What is the difference between predictive and preventive maintenance?

Preventive maintenance (PM) is scheduled at fixed intervals regardless of actual machine condition — e.g., replace bearings every 2000 hours. Predictive maintenance (PdM) uses real-time condition data (vibration, temperature, current draw) to identify degradation before failure — only intervening when data indicates an issue is developing. PM prevents some failures but causes unnecessary maintenance actions; PdM optimises timing but requires more sophisticated monitoring infrastructure.

What is the P-F curve in maintenance?

The P-F curve (Potential Failure to Functional Failure curve) shows how a failure develops over time. 'P' is the point where a potential failure can first be detected (by vibration analysis, thermography, oil analysis, etc.). 'F' is the point of functional failure. The P-F interval is the window between detection and failure — the opportunity for predictive maintenance. Longer P-F intervals allow more planning time for the maintenance response.

How does reactive maintenance increase costs?

Reactive maintenance (run-to-failure) typically costs 3–10 times more than planned maintenance because: (1) failures occur at the worst possible time, causing maximum production disruption; (2) equipment damage is more extensive, requiring more expensive repairs and parts; (3) secondary systems may also fail; (4) technicians may need to work overtime; (5) emergency spare parts procurement is more expensive than stock replenishment.

Blog

Predictive vs Preventive vs Reactive Maintenance: Which Strategy Wins?

February 28, 2026 · 8 min read · Shopfloor Copilot Team

Every manufacturer chooses a maintenance strategy — consciously or by default. The choice has a direct, measurable impact on OEE Availability, maintenance cost per unit produced, and equipment lifecycle. Here's a practical breakdown of each approach, when to use which, and how to use condition data to move up the maturity curve.

The Three Maintenance Strategies

Strategy	Trigger	Cost multiplier	OEE impact	Best for
Reactive (Run-to-Failure)	Equipment fails	3–10×	Major Availability loss	Non-critical, cheap-to-replace assets
Preventive (Fixed Schedule)	Calendar or meter	1× (baseline)	Planned Availability loss	Equipment with predictable wear patterns
Predictive (Condition-Based)	Condition data	0.3–0.5×	Minimal Availability loss	Critical equipment with monitorable degradation

Reactive Maintenance: The Hidden Cost

Reactive maintenance (run-to-failure) isn't always wrong — for a £10 proximity sensor, replacement on failure is perfectly rational. The problem is when reactive maintenance becomes the default for critical equipment by neglect rather than design.

When critical equipment fails unexpectedly, the true cost includes:

Lost production at full margin (not just cost of repair)
Emergency spare parts at 2–5× normal cost
Overtime labour for emergency repair
Secondary damage from running to full failure (e.g., bearing seizure destroying a shaft)
Downstream effects on customer commitments

Industry data: The U.S. Department of Energy estimates that reactive maintenance costs 3–10× more per maintenance event than a properly planned preventive task on the same asset.

The P-F Curve: Time Between Detection and Failure

The P-F curve illustrates how failures develop. The key insight: most failures don't happen instantaneously. They develop over time, and if you're monitoring the right indicators, you can detect them at point P (potential failure) — long before they reach point F (functional failure).

Vibration analysis detects bearing defects weeks to months before failure (long P-F interval)
Thermography detects electrical hot spots hours to days before failure
Oil analysis detects contamination and wear metals weeks before bearing failure
OPC UA signals detect cycle time deviation and micro-stoppages in real time

The longer the P-F interval, the more time you have to plan the maintenance response — ordering parts, scheduling downtime in a low-impact window, preparing the repair team.

Preventive Maintenance: Necessary but Imperfect

Fixed-interval preventive maintenance is better than reactive for anything critical, but it has two fundamental weaknesses:

Over-maintenance: If the interval is set conservatively (common when you don't have failure history data), you replace components that still have 30–50% service life remaining. This creates unnecessary planned downtime and labour cost.
Under-maintenance: If the interval is set too long, or if operating conditions vary (duty cycle, environment), failure can still occur before the next scheduled PM.

Preventive maintenance works well for wear-based failure modes with consistent patterns — timing belts, filters, lubrication. It doesn't work as well for random failure modes (electrical components, bearings under variable load), where condition monitoring is more appropriate.

Choosing the Right Strategy by Equipment Criticality

Critical, constraint resource (no bypass): Predictive maintenance — invest in OPC UA condition monitoring, vibration sensors, thermal imaging schedule
Important, has backup: Preventive maintenance at conservative intervals — the backup absorbs any failure during the PM window
Non-critical, low-cost asset: Reactive maintenance — planned replacement on failure is cost-effective

Frequently Asked Questions

What is condition-based maintenance (CBM)?

Condition-based maintenance (CBM) is another term for predictive maintenance — maintenance performed when condition monitoring data indicates an asset is degrading toward failure. CBM uses continuous or periodic monitoring (vibration, temperature, current draw, oil contamination) to detect the P (potential failure) point on the P-F curve and trigger a maintenance response before functional failure occurs.

How does predictive maintenance improve OEE?

Predictive maintenance improves OEE by reducing unplanned downtime — directly increasing the Availability component. By catching degradation before it causes failure, predictive maintenance converts unplanned (worst-case OEE impact) downtime into planned (scheduled, minimised) downtime. A 50% reduction in unplanned downtime on a bottleneck line typically translates to 2–5 OEE percentage points of improvement.

Move from Reactive to Predictive — Starting Today

Shopfloor Copilot tracks equipment health scores, predicts failure dates with Prophet time-series forecasting, and raises prioritised maintenance alerts — all from existing OPC UA signals.

Explore Predictive Maintenance →