What Is Condition-Based Maintenance?
Condition-based maintenance (CBM) is a maintenance strategy that triggers action based on the measured physical condition of an asset โ not on a calendar date or running-hour counter. Instead of replacing a bearing every 12 months (preventive maintenance) or waiting for it to fail (reactive maintenance), CBM asks a different question: What is the machine telling us right now?
The core principle: maintain when needed, not when scheduled. This shifts maintenance from interval-driven cost centers to data-driven resource allocation โ you intervene exactly when degradation begins, no sooner and no later.
CBM sits between preventive and predictive maintenance on the reliability maturity spectrum. It provides the sensing layer โ the actual condition data โ that makes predictive analytics possible. Without CBM, predictive maintenance has nothing to predict from.
Key Condition Indicators
CBM relies on measurable physical parameters that change as an asset degrades. Each indicator captures a different failure mode at a different point along the failure timeline:
Vibration โ the earliest indicator of rotating machinery faults. Unbalance, misalignment, bearing defects, and mechanical looseness all produce distinct vibration signatures in the acceleration (g), velocity (mm/s), and displacement (ยตm) spectra before any other symptom appears.
Temperature โ surface or fluid temperature measured via thermocouples, RTDs, or thermal imaging. Detects blocked cooling, overloading, lubrication failure, and insulation degradation. Temperature is a lagging indicator โ it measures the consequence of damage, not its onset.
Pressure โ differential pressure across filters, pumps, and heat exchangers. A rising ฮP across a lube oil filter ahead of schedule means the filter is clogging or the oil is dirtier than expected โ actionable information that triggers inspection before the bypass valve opens.
Oil debris and analysis โ particle count, ferrous debris concentration, viscosity, water content, and elemental spectroscopy. Catches internal wear (gear scuffing, bearing spalling) before it becomes visible in vibration. Especially valuable for gearboxes and reciprocating engines where debris is trapped in the oil circuit.
Current draw (motor current signature analysis) โ electrical anomalies caused by mechanical faults. A cracked rotor bar, eccentric air gap, or driven-load change all modulate the current waveform. Non-intrusive โ measured at the motor starter panel without contacting the machine.
Visual inspection โ the most accessible CBM technique. Cracking, corrosion, leakage, misalignment marks, loose fasteners, belt wear, fluid discoloration, and unusual noise all qualify as condition indicators. A trained eye with a structured checklist remains one of the most effective CBM tools in any plant.
The P-F Curve: Why CBM Catches Failures Before They Happen
The P-F curve is the conceptual foundation of condition-based maintenance. It describes the progression of a failure from the earliest detectable point (P) to functional failure (F).
Condition
โ
| Potential Failure (P)
Normal | โข โข
operating | โข โข
range | โข โข Functional Failure (F)
| โข โข
| โข โข
| โข โข
+--------------------------------โ Time
โ P-F interval โ
-
Point P (Potential Failure) โ the earliest moment a fault can be detected using your chosen monitoring technique. For a bearing spall, this is a burst of high-frequency energy in the acceleration envelope spectrum, visible weeks before the defect reaches the raceway surface.
-
Point F (Functional Failure) โ the asset can no longer perform its required function. The motor has tripped, the pump has seized, or product quality is out of spec.
-
P-F interval โ the window between P and F. This is your opportunity window. Every CBM technique is chosen to maximize the P-F interval while balancing measurement cost and complexity.
Different indicators catch the failure at different points along the curve. Oil analysis detects wear particles earliest (longest P-F interval), followed by vibration, then temperature, then physical effects. A layered CBM strategy โ combining multiple indicators โ gives the widest coverage.
How to Implement Condition-Based Maintenance
Implementing CBM follows a structured six-step process:
1. Select critical assets. Not every asset needs CBM. Screen your register using criticality criteria: safety impact, production loss cost, repair cost, and detectability. Focus on assets where failure consequence justifies the monitoring investment. Pareto principle applies โ 20% of assets cause 80% of downtime.
2. Define limits. For each critical asset, establish four thresholds:
- Baseline โ normal operating condition, measured during commissioning or after a known-good overhaul
- Alert โ deviation warrants investigation at next opportunity
- Alarm โ deviation warrants planned intervention
- Shutdown โ deviation warrants immediate stop
3. Install sensors or define routes. For continuous monitoring: mount vibration sensors, temperature probes, pressure transmitters at documented locations. For route-based CBM: define measurement points, instrument setup, and collection frequency. Document each point's exact location and orientation so readings are repeatable.
4. Collect data. Establish a consistent cadence. For continuous monitoring this is automatic. For route-based, enforce schedule compliance โ missed rounds create blind spots. Standardize how readings are taken to eliminate operator variability.
5. Analyze. Compare readings against baselines and limits. Prioritize trend analysis over single-point thresholds โ a consistent upward slope matters more than today's absolute value. Many faults announce themselves through rate of change before crossing any alarm threshold.
6. Act. Translate analysis into action: inspect, lubricate, adjust, repair, or replace. Close the loop by recording the outcome so the next analysis cycle benefits from what was found.
Low-Cost CBM: Start Before You Invest in Online Sensors
Continuous online monitoring is powerful but expensive to deploy across an entire plant. Most organizations should start with route-based CBM:
- A technician walks a defined route weekly or monthly
- Uses handheld instruments at each measurement point
- Records readings on a tablet or structured form
- Compares new readings against stored baselines and trends
Minimum equipment to begin:
| Tool | Approximate cost | What it measures |
|---|---|---|
| Handheld vibration meter | $500 โ $2,000 | Overall velocity, acceleration, bearing condition |
| Infrared thermometer | $100 โ $300 | Surface temperature at bearing housings, motor frames |
| Ultrasonic detector | $500 โ $2,000 | Airborne/structure-borne noise โ leaks, steam traps, bearings |
| Stroboscope | $200 โ $800 | Visual speed measurement, belt inspection |
Route-based CBM catches roughly 80% of developing faults at a fraction of the cost of permanent online sensors. Once route discipline is established, online sensors can be justified for specific assets where the P-F interval is too short to cover with monthly rounds.
CBM vs. Predictive Maintenance
These terms are often used interchangeably, but they describe different layers:
| Dimension | Condition-Based Maintenance | Predictive Maintenance (PdM) |
|---|---|---|
| Scope | Data collection + threshold alerts | Data collection + trend projection + RUL estimation |
| Question | "Is the asset healthy?" | "When will it fail?" |
| Techniques | Vibration RMS, temperature limits, pressure limits | Weibull analysis, regression models, ML-based prediction |
| Output | Alarm / alert / work order trigger | Probability distribution over time, recommended action window |
| Reliability maturity | Entry-level | Advanced |
Think of CBM as the sensing and alerting layer โ it tells you that something has changed. Predictive maintenance is the analytics layer โ it tells you how much time remains before failure.
You cannot do PdM without CBM. Predictive models need the condition data that CBM provides. But CBM alone is already a significant improvement over time-based preventive maintenance for assets with random or wear-out failure patterns โ which describes the majority of industrial equipment.
How a CMMS Powers CBM
A computerized maintenance management system turns CBM from a collection of disconnected readings into an integrated operational workflow:
- Route management โ define inspection routes, assign to technicians, track completion rates, auto-reschedule missed rounds
- Data collection forms โ structured forms with dropdowns, numeric entry, pass/fail criteria, and photo attachments. Standardizes capture so readings are comparable across rounds and technicians
- Baseline and limit storage โ stores design limits, alarm thresholds, and historical baselines per asset. New readings are compared automatically
- Trend visualization โ plots any numeric reading over time. Rate-of-change alerts trigger earlier than absolute-level alarms
- Automated work order generation โ when a reading crosses the alert threshold, the CMMS creates a follow-up work order with the reading context
- Closed-loop learning โ work order outcomes link back to condition readings, improving alert accuracy over time
OpexMX implements this entire workflow. Condition readings โ entered manually from a handheld device or streamed from IoT sensors โ feed directly into asset health scores, work order triggers, and trend charts. The platform handles route scheduling for technicians and escalates readings that exceed configured limits.