Structural Health Monitoring — Continuous Condition Assessment of Engineered Structures
Structural Health Monitoring · Also known as: SHM, damage detection monitoring, condition monitoring of structures, vibration-based structural monitoring
Structural Health Monitoring (SHM) is a process-based engineering methodology used in civil, mechanical, and aerospace engineering to continuously assess the condition of structures — bridges, buildings, dams, pipelines, and aircraft — through embedded or attached sensor networks. By acquiring real-time or periodic measurement data and applying signal processing and statistical pattern recognition, SHM aims to detect, locate, classify, and quantify damage before it reaches a critical state, enabling evidence-based maintenance decisions.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
SHM is appropriate for structures where failure consequences are severe, inspection is difficult or expensive, or continuous service is critical — long-span bridges, high-rise buildings, offshore platforms, wind turbines, aircraft fuselages, and nuclear facilities. It suits situations where periodic visual inspection is insufficient to detect sub-surface or fatigue damage, and where real-time condition data can meaningfully improve maintenance decisions. SHM is not well-suited to simple, inexpensive structures where the cost of a sensor network exceeds the benefit, or where damage is clearly visible and simple scheduled inspection suffices.
Strengths & limitations
- Enables early detection of damage before it propagates to critical levels, reducing catastrophic failure risk.
- Provides continuous, objective condition data that replaces or augments subjective visual inspection.
- Supports evidence-based maintenance scheduling, shifting from time-based to condition-based strategies.
- Applicable across diverse structure types and engineering domains — civil, mechanical, aerospace, offshore.
- Integration with machine learning allows automated, scalable damage detection across large sensor networks.
- Long-term monitoring datasets increase understanding of actual load environments and structural behaviour.
- Sensor installation, data acquisition infrastructure, and data storage are capital-intensive, particularly for large or remote structures.
- Environmental and operational variability — temperature, humidity, traffic patterns — can mask or mimic damage signatures, requiring careful normalisation.
- Prognosis and remaining-useful-life estimation remain challenging; physics-based and data-driven models both carry significant uncertainty.
- High-level damage classification (type and severity) is substantially harder than simple detection and often requires validated physics-based models.
- Long-term sensor reliability and calibration drift introduce data quality issues over multi-year monitoring periods.
Frequently asked
What sensors are most commonly used in SHM?
Accelerometers are the most widely deployed sensor type because vibration data are rich in damage-sensitive features. Strain gauges (foil and fibre-optic FBG types) are used where static and dynamic strain is the primary diagnostic variable. Displacement transducers, acoustic emission sensors, piezoelectric patches, and fibre-optic distributed sensing systems are used for specific applications. The choice depends on the damage type anticipated, the structure's dynamic characteristics, and the monitoring frequency required.
How is SHM different from non-destructive testing (NDT)?
NDT (ultrasound, radiography, dye penetrant, etc.) is typically a periodic, offline inspection technique applied at discrete points in time and specific locations. SHM is a continuous or near-continuous, in-situ monitoring process using permanently attached sensors that captures the global structural response. NDT generally achieves finer local resolution but requires access and interruption of service; SHM provides ongoing global assessment without interruption but with lower spatial resolution unless dense sensor networks are used.
Can machine learning replace physics-based models in SHM?
Machine learning is increasingly powerful for feature extraction and damage classification, but it does not replace physics-based models for prognosis — predicting remaining useful life requires understanding of degradation mechanics (fracture, fatigue, corrosion) that purely data-driven models cannot reliably extrapolate beyond their training distribution. Hybrid approaches that combine physics-based priors with machine learning uncertainty quantification are the current state of the art.
What is the Rytter classification?
Rytter (1993) proposed a four-level hierarchy for damage assessment: Level 1 — Detection (does damage exist?); Level 2 — Localisation (where is the damage?); Level 3 — Classification (what type of damage?); Level 4 — Severity/Prognosis (how severe is it and what is the remaining useful life?). Each level is more demanding than the previous; most deployed SHM systems operate reliably only at Levels 1–2, with Levels 3–4 remaining active research challenges.
Sources
- Farrar, C. R., & Worden, K. (2007). An introduction to structural health monitoring. Philosophical Transactions of the Royal Society A, 365(1851), 303–315. DOI: 10.1098/rsta.2006.1928 ↗
- Farrar, C. R., & Worden, K. (2012). Structural Health Monitoring: A Machine Learning Perspective. Wiley. ISBN: 978-1119994336
How to cite this page
ScholarGate. (2026, June 3). Structural Health Monitoring. ScholarGate. https://scholargate.app/en/civil-engineering/structural-health-monitoring
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Finite Element AnalysisMaterials Science↔ compare
- Modal AnalysisManufacturing↔ compare
- Reliability AnalysisReliability↔ compare