22 Aug 2026 · Predictive Maintenance
Most facilities management programmes are built around Planned Preventive Maintenance (PPM): inspect an asset every month, service it quarterly, overhaul it annually and respond when it fails. PPM remains essential, but calendar-based maintenance alone cannot tell an FM team when an individual asset has actually started to deteriorate.
The P-F Curve provides the reliability-engineering logic for closing that gap. It describes the period between the point at which a potential failure first becomes detectable (P) and the point at which the asset can no longer perform its required function (F). Understanding this interval allows FM teams to intervene based on condition rather than waiting for functional failure.
Understanding the P-F Curve
An asset rarely moves from healthy operation to complete failure instantaneously. In many failure modes, degradation develops progressively and creates detectable indicators before the functional failure occurs.
Normal Condition → P: Potential Failure Detected → Degradation → F: Functional Failure
The time between P and F is the P-F interval. The engineering objective is to select a monitoring technique and inspection frequency capable of detecting the failure early enough to plan an intervention before the required function is lost.
Why Conventional PPM Can Miss Developing Failures
Fixed-frequency PPM assumes that maintenance at predetermined intervals will control failure risk. This is appropriate for many statutory, age-related and routine maintenance activities, but not every failure mode is strongly correlated with calendar age.
- Bearings may deteriorate between scheduled services.
- Electrical connections can develop abnormal resistance and heat while equipment remains operational.
- Pump cavitation can damage an impeller despite completed PPM.
- Chiller performance can progressively deteriorate through fouling or refrigerant-side issues.
- Motor insulation can degrade without an obvious mechanical symptom.
- Control sensors can drift and create significant energy waste without generating a conventional alarm.
A completed PPM work order therefore confirms that a maintenance activity was performed; it does not automatically confirm that the asset is healthy.
The P-F Interval Determines the Monitoring Strategy
The P-F interval differs by asset and failure mode. Some degradation mechanisms develop over months, while others can progress rapidly. Condition-monitoring frequency should therefore be shorter than the expected P-F interval and should provide sufficient time for diagnosis, parts procurement, shutdown planning and repair.
Detection Frequency < P-F Interval → Time to Diagnose + Plan + Procure + Intervene
Condition Monitoring Across the P-F Curve
| Technique / Data | Typical FM Application | What It Can Reveal |
|---|---|---|
| Vibration Analysis | Pumps, motors, fans, compressors, rotating plant | Imbalance, misalignment, bearing defects, looseness and resonance. |
| Ultrasound | Bearings, compressed air, steam traps, electrical systems | Early friction, leakage, arcing and abnormal ultrasonic signatures. |
| Oil / Lubricant Analysis | Engines, gearboxes and lubricated rotating assets | Wear particles, contamination, viscosity change and lubricant degradation. |
| Infrared Thermography | LV panels, DBs, MCCs, motors, bearings and mechanical systems | Hot connections, overload, phase imbalance, friction and abnormal heat patterns. |
| Motor Current / Electrical Testing | Motors and electrical equipment | Electrical imbalance, loading abnormalities and selected motor/electrical deterioration. |
| BMS Trend Analytics | Chillers, AHUs, pumps and controlled MEP systems | Setpoint deviation, valve behaviour, temperature drift, abnormal runtime and performance loss. |
| Performance Trending | Chillers, pumps, fans and energy-intensive plant | Efficiency deterioration, abnormal kW/TR, flow/pressure deviation and loss of capacity. |
Example: Bearing Failure in a Chilled-Water Pump
Consider a critical chilled-water pump. The bearing does not necessarily fail without warning. A progressive failure may first create subtle vibration-frequency changes. As degradation advances, vibration amplitude may increase, followed by temperature rise, audible noise and eventually loss of function.
- P1, early detectable condition: Specialist vibration analysis identifies a developing bearing defect.
- P2, developing degradation: Vibration trend increases and ultrasound may identify abnormal friction.
- P3, advanced condition: Bearing temperature rises; thermography and BMS/field readings may show abnormal behaviour.
- P4, obvious deterioration: Noise, excessive vibration or seal/alignment effects become apparent to operators.
- F, functional failure: Bearing seizure or severe damage causes pump trip, loss of output or consequential equipment damage.
The earlier the defect is identified, the greater the opportunity to schedule the intervention, verify standby capacity, procure parts and avoid an emergency shutdown.
Predictive Maintenance Is Not Just Sensors
A common misconception is that predictive maintenance is achieved simply by installing IoT sensors. Sensors create data; they do not automatically create an engineering decision. A functioning predictive-maintenance programme requires a complete technical workflow.
Asset Criticality → Failure Mode → Detectable Parameter → Baseline → Threshold → Trend → Diagnosis → Work Order → Verification
Without failure-mode knowledge, correctly selected measurements, baselines, alarm logic and a maintenance response process, large volumes of condition data can become another dashboard that technicians ignore.
Linking P-F Analysis with FMEA and Asset Criticality
Not every asset requires continuous condition monitoring. The business case becomes stronger when P-F analysis is combined with asset criticality and FMEA/FMECA.
- What are the credible failure modes?
- What is the consequence of each failure?
- Can the failure mode be detected before functional failure?
- What parameter changes as degradation develops?
- What is the approximate P-F interval?
- Is the monitoring technique technically and economically justified?
- What action should be triggered when deterioration is detected?
This prevents predictive maintenance from being applied indiscriminately and focuses investment on failure modes where early detection creates measurable operational value.
From Calendar-Based PPM to a Hybrid Maintenance Strategy
The objective is not to eliminate PPM. A mature FM strategy uses the maintenance method best suited to each asset and failure mode.
| Strategy | Best Applied When | FM Example |
|---|---|---|
| Statutory / Mandatory | Maintenance frequency is legally, regulatorily or contractually required. | Fire and life-safety inspections and mandatory testing. |
| Time-Based PPM | Failure behaviour or OEM requirement supports scheduled intervention. | Routine filter changes, lubrication and prescribed servicing. |
| Condition-Based Maintenance | A measurable condition indicates degradation. | Vibration-based bearing intervention or filter differential-pressure replacement. |
| Predictive Maintenance | Trend and data analysis can forecast or identify developing failure sufficiently early. | Chiller performance degradation or rotating-equipment condition analytics. |
| Run-to-Failure | Failure consequence is low and replacement is economically preferable to preventive intervention. | Selected low-value, non-critical components after risk assessment. |
Using BMS and CAFM as Part of the P-F Strategy
Existing FM technology can already provide useful condition information. BMS trend logs may reveal excessive equipment runtime, abnormal valve position, temperature deviations, pressure instability or declining plant efficiency. CAFM/CMMS provides the maintenance history required to understand whether those conditions are translating into repeat failures.
BMS / IoT Condition Data + CAFM Failure History + Asset Criticality = Better Maintenance Decisions
When an abnormal condition crosses a defined engineering threshold, the workflow should generate an inspection or work order, capture findings and verify performance after rectification.
Technical KPIs for Predictive FM
Moving toward predictive maintenance also changes what FM should measure:
- MTBF trend for critical assets.
- MTTR and emergency restoration time.
- Percentage of failures detected before functional failure.
- Condition-monitoring alerts converted into planned interventions.
- Unplanned versus planned maintenance ratio.
- Repeat failure and bad-actor asset rate.
- Critical asset availability.
- Emergency call-out reduction.
- Maintenance cost avoided through early intervention.
- Energy and performance recovery following corrective action.
Why the P-F Curve Matters for Gulf Facilities
Gulf operating conditions make early detection particularly valuable. High ambient temperatures, dust, humidity, long cooling seasons and intensive HVAC duty can accelerate deterioration or magnify the consequences of equipment underperformance. Critical facilities such as hospitals, aviation assets, data centres, hotels and major commercial developments also have low tolerance for unplanned downtime.
For these environments, waiting for an obvious fault can be expensive. Detecting degradation earlier creates time, and that time allows the FM team to move a repair from an emergency event into a controlled maintenance activity.
How Orion Venture Can Support the Shift to Predictive FM
At Orion Venture Facility Services, predictive maintenance can be developed as an engineering strategy rather than a technology purchase.
- Critical asset identification and risk ranking.
- FMEA/FMECA and failure-mode identification.
- P-F interval and detectable-condition assessment.
- Condition-monitoring strategy development.
- Vibration, thermography, ultrasound and other specialist predictive-maintenance planning.
- BMS trend-point identification and Fault Detection and Diagnostics requirements.
- IoT sensor selection based on failure mode and business case.
- CAFM/CMMS integration of condition alerts and maintenance actions.
- Baseline, threshold and exception-reporting development.
- Bad-actor, MTBF and repeat-failure analysis.
- Verification of performance after corrective intervention.
From Maintenance Completion to Failure Prevention
The maturity of an FM operation should not be judged only by how many PPM work orders it closes. A stronger measure is whether the maintenance strategy can identify degradation, intervene at the correct point and prevent high-consequence functional failures.
The P-F Curve gives FM teams a practical reliability framework for making that transition. It connects asset criticality, failure modes, condition monitoring, BMS data, CAFM history and engineering judgement into a maintenance strategy based increasingly on the actual condition of the asset.
PPM asks: "When is the next service due?"
Predictive FM asks: "What is the asset telling us now?"
People Centric. Technology Driven. Internationally Aligned Facilities Management.
