Contents
What is condition monitoring?
Condition monitoring is the continuous measurement and analysis of machine health indicators — vibration, temperature, pressure, current, acoustics — to detect degradation before it causes unplanned downtime. Unlike time-based maintenance (replace every X hours) or reactive maintenance (fix after failure), condition monitoring triggers action based on the actual state of the equipment. It is the foundation of predictive maintenance and a core capability of any modern Manufacturing Execution System (MES).
How does condition monitoring work in practice?
Sensors capture machine parameters in real time. Edge or cloud systems analyze the data against baselines. Deviations trigger alerts, maintenance orders, or automatic process adjustments — depending on the maturity level of the implementation.
| Step | What happens | Technology | Example |
|---|---|---|---|
| 1 | Capture | Vibration sensors, temperature probes, current transducers, acoustic emission sensors | Accelerometer on a press spindle bearing samples at 10 kHz |
| 2 | Transmit | OPC UA, MQTT, digital I/O via DI-Gateway | Edge gateway pushes data to Azure Cloud every 1 s |
| 3 | Analyze | Threshold monitoring, trend analysis, FFT, ML anomaly detection | MES correlates rising vibration amplitude with alarm codes |
| 4 | Act | Alert |
What are the 4 main condition monitoring methods?
| Method | What it measures | Best for | Detects |
|---|---|---|---|
| Vibration analysis | Displacement, velocity, acceleration (FFT spectrum) | Rotating equipment: motors, spindles, pumps, fans | Bearing wear, imbalance, misalignment, looseness |
| Thermography | Surface temperature via infrared camera or contact sensor | Electrical cabinets, friction contacts, hydraulic systems | Overheating, insulation failure, blocked cooling |
| Oil / particle analysis | Metal particles, viscosity, contamination in lubricant | Gearboxes, hydraulic systems, large drives | Internal wear, contamination, degradation |
| Electrical analysis | Current signature, power consumption, harmonic distortion | Electric motors, drives, servo systems | Rotor bar defects, winding insulation, load anomalies |
In practice, most discrete manufacturers start with vibration analysis (highest detection rate for rotating equipment) and PLC-based alarm correlation (captures downtime causes automatically from the machine controller).
How does condition monitoring compare to other maintenance strategies?
| Strategy | Trigger | Downtime risk | Cost profile |
|---|---|---|---|
| Reactive (fix when broken) | Failure event | High — unplanned stops | Lowest upfront, highest total (emergency repair + lost production) |
| Preventive (time-based) | Calendar or runtime interval | Medium — over-maintains or under-maintains | Moderate (30–40 % of parts replaced still had usable life) |
| Condition-based (this article) | Measured degradation exceeds threshold | Low — intervention before failure | Optimal — maintains only when needed |
| Predictive (ML-driven) | Algorithm predicts remaining useful life | Lowest — planned well in advance | Higher sensor/IT investment, highest long-term savings |
Condition monitoring is the practical middle ground: it delivers 80 % of the value of full predictive maintenance at a fraction of the complexity.
What role does an MES play in condition monitoring?
Standalone condition monitoring captures data. An MES makes it actionable. The MES connects sensor data with production context — which order was running, which operator was on shift, which process parameters were active — so that deviations can be traced to root causes, not just flagged as anomalies.
- Alarm correlation: MES links machine alarms to downtime events and quality defects.
- Automatic downtime classification: PLC-based alarm capture classifies stops without operator input — the machine explains itself.
- When a threshold is exceeded, the MES can notify the responsible technician.
FAQ
What is the difference between condition monitoring and predictive maintenance?
Condition monitoring measures current machine state and triggers action when thresholds are exceeded. Predictive maintenance uses historical condition data + ML to forecast when a failure will occur. Condition monitoring is the prerequisite — you cannot predict without first monitoring.
Which machines benefit most from condition monitoring?
Any machine where unplanned downtime is expensive: bottleneck machines, high-speed assembly lines, presses with critical bearing loads, CNC spindles. The ROI is highest where downtime cost per hour is highest.
Can condition monitoring work on brownfield machines?
Yes. Modern IoT gateways (IXON, DI-Gateway) capture digital signals from machines without modifying PLC logic. SYMESTIC connects both OPC-UA-capable modern machines and 10–30-year-old brownfield equipment across 15,000+ connected machines.
Related: Predictive Maintenance · MES: Definition & Functions · OEE Explained · MES Use Cases · OPC UA · Smart Factory
- Live demo
- 30 minutes
- Free
See machine states and process values live
In a live demo we show how machine states and process values are recorded and an alarm notifies your team when a value crosses a limit you set.
- Live OEE
- Downtime and reasons
- Connecting machines and ERP
About the author
Uwe KobbertFounder & CEO
In manufacturing IT since 1989; since 1995 he has been building MES solutions for controlling and digitalising production. Dipl.-Ing. in communications engineering and electronics.
All articles by Uwe Kobbert