ISO 24032:2020 Crane Health Monitoring Systems Standard Explained

Health Monitoring System Architecture

ISO 24032:2020 defines a four-layer architecture for crane health monitoring systems. The sensing layer comprises sensor groups installed at critical crane locations, including vibration sensors (accelerometers that measure RMS and peak acceleration values on gearbox bearings and motors), temperature sensors (thermocouples or PT100 RTDs that monitor motor winding, bearing housing, and hydraulic oil temperatures), strain sensors (resistance strain gauges that measure stress at critical cross-sections of the main girder), displacement sensors (laser displacement meters that measure mid-span deflection of the main girder), force sensors (which measure hoisting rope tension and outrigger ground pressure), and acoustic emission sensors (which detect high-frequency stress wave signals released during crack initiation and propagation in steel structures). The data acquisition layer uses data acquisition modules (high-precision ADC plus signal conditioning circuitry) with sampling frequencies of ≥10 kHz for vibration signals and ≥1 Hz for slow-varying signals such as temperature and strain. The data transmission layer sends data to the monitoring center via industrial Ethernet or 5G/4G wireless networks. The data analysis layer, a cloud-platform-based real-time analysis engine, processes sensor data using signal processing algorithms (FFT time-frequency analysis, wavelet analysis, envelope analysis) and AI models (deep learning anomaly detection and fault classification models) to output real-time health condition assessments. The standard also recommends minimum sensor configurations for each crane type: bridge cranes require at least 8 sensors (2 strain gauges at mid-span of the main girder, 2 vibration sensors on the gearbox, 2 temperature sensors on the motor, and 2 vibration sensors on the hook block), while tower cranes require at least 12 sensors (4 strain gauges on tower main chords, 2 vibration sensors on the slewing bearing, 2 vibration sensors on the hoist gearbox, 1 anemometer, 1 horizontal displacement sensor at the tower top, and 2 settlement sensors at the foundation).

Data Analysis and Fault Diagnosis

The standard recommends the following data analysis and fault diagnosis methods. Vibration signal analysis includes time-domain analysis (computing RMS value Vrms, peak value Vp, and kurtosis factor — when the RMS value exceeds ISO 10816 limits or the kurtosis factor exceeds 4, early-stage faults are indicated), frequency-domain analysis (applying FFT to convert time-domain signals into spectra — searching for sidebands around the gear mesh frequency (GMF) to confirm gear wear or tooth breakage), and envelope analysis (demodulating vibration signals to extract bearing fault characteristic frequencies (BPFI/BPFO/BSF/FTF) — identifying fatigue spalling on inner or outer raceways, rolling element wear, or cage fracture). Temperature analysis monitors temperature trends at each measurement point — when the temperature under identical operating conditions exceeds the normal operating temperature by more than 10°C, it indicates bearing damage, poor heat dissipation, or lubrication failure. Strain analysis evaluates stiffness changes in the main girder structure by tracking historical strain data trends, and compares strain differences between full-load and no-load conditions to determine whether fatigue damage has occurred in the main girder. Acoustic emission analysis uses count and energy parameters detected by acoustic emission sensors to identify micro-crack initiation and propagation in steel structures. The standard specifies that health monitoring systems shall implement three alarm thresholds: the Attention Level — parameters slightly abnormal (20%–30% above normal values) — the system automatically notifies maintenance personnel to schedule an inspection; the Alarm Level — parameters significantly abnormal (50%–80% above normal values) — the system automatically notifies the maintenance supervisor and recommends stopping the crane for inspection; and the Danger Level — parameters critically exceeded (more than 100% above normal values) — the system automatically triggers an emergency shutdown signal and cuts off the power supply.

Condition Assessment and Prediction

The standard specifies the following condition assessment and prediction methods. The Health Index (HI) compares the current value of each monitored parameter against its normal baseline value (the initial value of new equipment or after overhaul) to calculate a composite index from 0 to 100. An HI of ≥80 indicates a healthy condition and normal operation can continue. An HI between 60 and 80 indicates an attention state requiring scheduled inspection. An HI between 40 and 60 indicates an abnormal state requiring prompt shutdown for inspection. An HI below 40 indicates a dangerous state requiring immediate shutdown. Remaining Useful Life (RUL) prediction applies to components with progressive degradation trends (such as gear and bearing wear) — trend prediction models (linear regression, exponential regression, or machine learning models) based on historical data forecast the time until the component reaches its failure threshold (i.e., remaining safe operating hours). The standard recommends scheduling component replacement when the RUL falls below 500 hours. Predictive maintenance recommendations — the health monitoring system should automatically generate maintenance suggestions based on condition assessment and life prediction results, such as "The vibration trend on the hoist gearbox input shaft bearing is rising significantly — recommend replacing the bearing within the next 200 hours" and "The mid-span strain value of the main girder has increased by 8% over the baseline — recommend scheduling a structural inspection." The standard emphasizes that health monitoring systems do not replace mandatory periodic statutory inspections (such as annual inspections and load tests) but rather fill the "blind spots" between inspections — where condition changes may go undetected during the up-to-12-month interval between inspections — by providing more timely safety warnings through continuous real-time monitoring.

System Integration and Management

The standard provides recommendations for integrating and managing health monitoring systems. Networked management of multiple cranes — dozens of cranes in large factories and ports can be centrally managed through a unified platform, with each crane's health index displayed on a main monitoring screen as an instrument panel — green (healthy), yellow (attention), red (alarm) — and clicking through provides access to detailed monitoring data and historical trends for that crane. Alarm management — the system should provide alarm grading and alarm push notification functions: attention-level alarms are pushed to the equipment maintenance department (as daily summary reports); alarm-level alarms are pushed in real time to supervisors and operators (via SMS or mobile app notifications); danger-level alarms automatically trigger remote shutdown commands (through a remote cut-off relay in the crane control system). Data storage and management — raw monitoring data should be retained for at least 3 months, while feature data and alarm records should be retained until equipment scrapping. The data generated by the health monitoring system serves as the fundamental input for ISO 12482 life assessment — the actual load spectrum and damage state continuously monitored and recorded by the system enable accurate remaining life assessment. The standard also requires that health monitoring systems provide open data interfaces (via REST API or OPC UA protocols) for integration with factory MES (Manufacturing Execution Systems) and CMMS (Computerized Maintenance Management Systems), enabling seamless integration of crane health management with factory production management. Kelude provides complete crane health monitoring system solutions for users.

FAQ: Health Monitoring System Essentials