There are four main types of condition-based maintenance (CBM): vibration analysis, thermal imaging, oil analysis, and ultrasonic testing. Each technique monitors a specific physical or chemical signal that changes as equipment degrades, allowing maintenance teams to act before failure occurs rather than after. The right combination depends on asset type, failure mode, and how critical the equipment is to your operation.
For industrial manufacturers managing complex, high-value assets, understanding these techniques is not an academic exercise. Unplanned downtime costs manufacturers an average of 27 hours per month, and every hour offline cascades through production schedules and service contracts. The sections below break down how each CBM method works, how to choose the right one, and how field technicians actually put this data to use on the plant floor.
What are the main types of CBM techniques?
Condition-based maintenance relies on four core monitoring techniques: vibration analysis, infrared thermography (thermal imaging), oil and fluid analysis, and ultrasonic testing. Some programs also incorporate electrical signature analysis and acoustic emission monitoring as supplementary methods, but these four form the foundation of most industrial CBM programs.
- Vibration analysis: Sensors measure oscillation patterns in rotating machinery. Changes in frequency or amplitude indicate bearing wear, imbalance, or misalignment before any visible damage occurs.
- Infrared thermography: Thermal cameras detect heat anomalies in electrical panels, motors, and mechanical components. Overheating is one of the earliest indicators of impending failure.
- Oil and fluid analysis: Laboratory or on-site testing of lubricant samples identifies metal particles, viscosity changes, and contamination. This reveals internal wear that no external sensor can see.
- Ultrasonic testing: High-frequency sound detection identifies leaks, friction, and electrical discharge in equipment that operates too quietly or too loudly for standard acoustic monitoring.
These techniques are not mutually exclusive. Most mature CBM programs layer two or more methods on the same asset, since different failure modes produce different signals. A gearbox, for example, benefits from both vibration analysis and oil sampling because each reveals a different class of problem.
How does each CBM type detect equipment failure?
Each condition-based maintenance technique detects failure by tracking a measurable physical or chemical property that shifts as the asset degrades. The key is establishing a baseline for healthy operation, then monitoring for deviations that exceed defined thresholds. Here is how each method works in practice:
Vibration analysis
Rotating equipment generates a characteristic vibration signature when operating normally. As components wear, that signature changes in predictable ways. Bearing defects produce high-frequency spikes; imbalance shows up as elevated amplitude at the rotational frequency; misalignment creates harmonic patterns. Vibration analysis catches these changes weeks or months before catastrophic failure, giving maintenance teams a defined intervention window.
Infrared thermography
Heat is a byproduct of friction, electrical resistance, and mechanical stress. Thermal imaging captures temperature distribution across a component’s surface and compares it to baseline readings. A motor running 15 degrees hotter than normal at the same load is a clear signal that something has changed internally. This technique is especially effective for electrical switchgear, motor windings, and heat exchangers.
Oil and fluid analysis
As internal components wear, they shed microscopic metal particles into the lubricant. Spectrometric analysis identifies the type and concentration of these particles, pointing directly to which component is degrading. Changes in viscosity or the presence of water and contaminants indicate a different class of problem: seal failure or coolant ingress. This is the only CBM method that reveals internal wear without disassembly.
Ultrasonic testing
Ultrasonic detectors convert high-frequency sound (typically 20 kHz to 100 kHz) into audible signals or waveform data. Compressed air leaks, steam trap failures, and bearing lubrication deficiencies all produce distinctive ultrasonic signatures. This technique is particularly useful in noisy manufacturing environments where audible sound is masked by background machinery.
Which CBM method is best for manufacturing equipment?
For most industrial manufacturing equipment, vibration analysis is the most broadly applicable condition-based maintenance technique because the majority of critical assets involve rotating components: motors, pumps, compressors, gearboxes, and fans. However, the best method depends on the specific asset and its dominant failure mode.
- Rotating machinery (motors, pumps, compressors): Vibration analysis is the primary tool, supported by oil sampling for gearboxes and large compressors.
- Electrical infrastructure (panels, switchgear, transformers): Infrared thermography is the standard approach, with ultrasonic testing added for partial discharge detection.
- Hydraulic and lubrication systems: Oil analysis is the most direct method, often combined with vibration monitoring on the pump itself.
- Pneumatic systems and steam traps: Ultrasonic testing identifies leaks and blockages that other methods miss entirely.
The practical answer for most industrial manufacturing operations is a layered program: vibration analysis as the backbone, thermal imaging for electrical and heat-generating assets, and oil analysis for high-value gearboxes and compressors. Ultrasonic testing fills the gaps. A single-method program will always have blind spots.
What’s the difference between CBM and predictive maintenance?
Condition-based maintenance (CBM) triggers a maintenance action when a monitored parameter crosses a defined threshold. Predictive maintenance (PdM) uses machine learning and statistical modeling to forecast when a failure will occur, even before a threshold is breached. CBM reacts to current condition; PdM anticipates future state.
In practice, the two approaches exist on a continuum rather than as opposites. A CBM program that collects vibration data over time and uses trend analysis to project when a bearing will fail is functioning as predictive maintenance. The distinction matters most when organizations are deciding what technology investment to make:
- CBM requires sensors, defined thresholds, and a process for acting on alerts. It is accessible to most mid-market manufacturers without advanced data science capabilities.
- PdM adds machine learning models trained on historical failure data. It delivers earlier warning and fewer false positives, but requires more data history and analytical infrastructure.
For organizations moving away from reactive maintenance, CBM is typically the right first step. It delivers measurable outcomes, builds the data foundation that PdM models need, and does not require a full digital transformation to implement. Predictive maintenance is the logical evolution once that data foundation exists.
How do field technicians apply CBM data on the job?
Field technicians apply condition-based maintenance data by using it to prioritize work orders, guide inspection checklists, and make informed decisions at the asset rather than relying on scheduled intervals or reactive calls. The value of CBM data is only realized when it reaches the technician at the point of service, in a usable format, at the right moment.
In practice, this means the CBM system needs to connect to the work order management platform. When a vibration sensor on a compressor crosses its alert threshold, that signal should automatically generate a work order, attach the relevant asset history, and route the task to a technician with the right skills. A technician arriving on-site without that context is starting from zero, which drives return visits and erodes first-time fix rates.
Offline access is a non-negotiable requirement on the plant floor. Mechanical rooms, basements, and remote production areas frequently have no reliable network signal. If the technician cannot access asset documentation, CBM trend data, and job checklists offline, the data collected by the monitoring system becomes useless at the exact moment it is needed most. This is one of the most common failure points in CBM implementations that rely on generic enterprise software not built for industrial field operations.
How Gomocha Helps Manufacturing Teams Act on CBM Data
Collecting condition data is only half the equation. The other half is getting that data into the hands of the right technician, with the right asset context, at the right time. That is where most CBM programs break down, and where we built Gomocha to close the gap.
Our field service platform connects CBM alerts directly to workflow-driven work orders, so technicians arrive on-site with full asset history, inspection checklists calibrated to the specific failure mode, and offline access to every document they need. No connectivity required on the plant floor.
Specifically, Gomocha helps manufacturing service teams by:
- Automatically generating work orders when CBM thresholds are breached, with asset history and priority routing built in
- Delivering fully offline-capable mobile access to asset documentation, safety procedures, and PM checklists, so technicians are never blocked by poor signal
- Enabling ops teams to configure inspection workflows per asset type using our no-code Workflow Designer, without waiting on IT
- Integrating natively with AFAS and Microsoft Dynamics, and connecting to SAP and JDE via connectors, so CBM data and work order history stay in sync with your ERP
Manufacturing teams using Gomocha have reduced unplanned downtime by up to 41% and improved first-time fix rates by up to 19%. If you want to understand where your current operation is losing time and margin, start with our Efficiency Assessment to identify the highest-impact opportunities specific to your asset base.
Frequently Asked Questions
How do I know which CBM techniques to prioritize when starting a program from scratch?
Start by auditing your most critical assets — the ones where unplanned failure causes the greatest production impact or safety risk. Map each asset to its dominant failure mode (mechanical, electrical, lubrication-related) and match that to the CBM technique most sensitive to that failure type. For most manufacturers, this means launching with vibration analysis on rotating equipment first, since it covers the broadest range of critical assets, then layering in thermal imaging and oil analysis as the program matures.
How often should CBM data be collected for each monitoring technique?
Collection frequency depends on the technique and the asset’s criticality. Vibration analysis on critical rotating equipment is typically performed monthly or continuously via permanently mounted sensors, while less critical assets may be surveyed quarterly. Oil sampling is usually done every 250–500 operating hours or quarterly, whichever comes first. Thermal imaging inspections are commonly scheduled semi-annually for electrical infrastructure, though high-criticality panels may warrant quarterly scans.
What are the most common mistakes teams make when implementing a CBM program?
The most common mistake is collecting condition data without a clear process for acting on it — sensors and sampling programs get set up, but alerts sit in a dashboard with no automatic link to work order generation or technician dispatch. A second frequent error is skipping baseline establishment: without a documented ‘healthy’ signature for each asset, it is impossible to determine whether a deviation is meaningful. Finally, teams often underestimate the need for offline data access in the field, which causes CBM insights to break down at the exact moment a technician needs them most.
Can CBM be applied to older equipment that wasn't designed with sensors in mind?
Yes — most CBM techniques are non-invasive and can be retrofitted to legacy equipment without modification. Portable vibration analyzers, handheld thermal cameras, and ultrasonic detectors require no permanent installation and can be deployed on virtually any asset during routine rounds. For higher-value legacy equipment, wireless vibration sensors can be surface-mounted without disassembly. The key is establishing a clean baseline reading early in the retrofit process, before further degradation occurs.
How much condition data history is needed before CBM alerts become reliable?
For threshold-based CBM, reliable alerts can be configured after as few as 3–6 months of baseline data collection on a healthy asset, assuming consistent measurement conditions. However, the more historical data you accumulate, the more accurately you can distinguish normal operational variation from genuine degradation trends. Organizations aiming to evolve toward predictive maintenance (PdM) with machine learning models typically need 12–24 months of labeled failure history per asset class before model training produces dependable results.
What's the best way to get maintenance technicians to trust and act on CBM data?
Technician buy-in is largely an information design problem: if CBM alerts arrive without context — no asset history, no explanation of what the threshold breach means, no guidance on what to inspect — technicians will default to their own judgment, which may or may not align with the data. Presenting CBM alerts alongside trend charts, failure mode context, and asset-specific checklists directly in the work order makes the data actionable rather than abstract. Early wins matter too — documenting cases where CBM data caught a failure before it happened builds credibility with the team faster than any training program.
How does CBM interact with existing preventive maintenance (PM) schedules?
CBM does not replace preventive maintenance entirely — it refines it. Time-based PM tasks that are justified by regulatory requirements or OEM warranty terms should remain on schedule, but CBM data can be used to extend or compress intervals on tasks where condition is a better trigger than calendar time. For example, if oil analysis consistently shows clean, in-spec lubricant at a 500-hour change interval, that interval can often be extended safely, reducing labor and material costs. The goal is a hybrid program where PM handles compliance-driven tasks and CBM handles condition-driven ones.