Condition Monitoring Strategies

Condition monitoring gets treated, too often, as a single technique — “we added sensors” — when it’s actually a toolkit of several distinct methods, each suited to catching different failure modes, and choosing the wrong one for a given asset is a common reason condition monitoring programs underperform without anyone quite understanding why.

The Techniques, and What Each One Actually Catches

Vibration analysis is the workhorse for rotating equipment — bearings, motors, pumps, gearboxes. It’s excellent at catching mechanical wear, imbalance, and misalignment well before failure, because these failure modes produce a distinctive, measurable vibration signature that changes predictably as wear progresses.

Thermal imaging catches failure modes that manifest as abnormal heat — electrical connection degradation, insulation breakdown, certain friction-related mechanical issues. It’s typically a periodic, point-in-time technique (a scheduled thermal scan) rather than continuous monitoring, which makes it well-suited to failure modes that develop gradually enough for periodic checks to catch reliably.

Oil analysis monitors lubricant condition and the wear particles suspended in it — catching internal wear on gearboxes, engines, and hydraulic systems that wouldn’t be visible from the outside at all. The lubricant itself becomes the diagnostic sample.

Acoustic monitoring listens for the sound signatures of specific failure modes — air or gas leaks, certain types of mechanical friction, electrical arcing — that don’t necessarily show up strongly in vibration data but produce a distinctive acoustic pattern.

Current signature analysis monitors the electrical current drawn by motors, catching electrical faults (winding issues, rotor bar problems) and certain mechanical issues that manifest as a load change, often without needing any sensor on the mechanical equipment itself — just an instrumented connection to the motor’s power supply.

Why Matching Technique to Failure Mode Is the Whole Game

A vibration sensor on a motor with a developing winding fault, and no current signature monitoring, may catch the fault late or not at all — vibration isn’t the primary signal that failure mode produces. This is the single most common reason a condition monitoring investment underperforms: the technique deployed doesn’t match the failure mode the asset is actually most prone to. Before specifying sensors, the more useful question is which failure modes this specific asset class has historically experienced, and which technique actually has a track record of catching that failure mode early.

Continuous vs. Periodic, Matched to Failure Speed

Not every technique needs to run continuously. A failure mode that develops over months (many types of lubricant degradation) is well served by periodic oil sampling. A failure mode that can develop and become critical over days or hours warrants continuous monitoring, because a periodic check might simply miss the window between one reading and the next. Matching monitoring frequency to how fast the underlying failure mode actually develops avoids both the risk of missing a fast failure and the unnecessary cost of continuous sensors on slow, predictable wear patterns.

Combining Techniques for Complex Assets

Critical, complex equipment often warrants more than one technique simultaneously — vibration and current signature analysis together on a critical motor-driven asset, for example, catching both mechanical and electrical failure modes that a single technique would miss on its own. This isn’t over-engineering; it’s matching monitoring coverage to the actual range of ways that specific asset is known to fail.

From Reading to Health Assessment

A condition monitoring reading, on its own, is a data point. Turning a stream of readings — across whichever techniques are appropriately deployed — into an actual asset health assessment and a maintenance trigger is a separate, necessary step, covered in Using Sensor Data for Asset Health.

Choosing Techniques Deliberately, Not by Default

The plants getting real value from condition monitoring didn’t default to whichever sensor was easiest to install. They matched the monitoring technique to the specific failure modes their critical assets actually experience. SG2’s Manufacturing & Industry 4.0 practice scopes condition monitoring technique selection against real failure history and criticality, not a generic sensor package applied uniformly across every asset.