Data-driven preliminary dynamics analysis of cranes using rope angle and load measurements

Abstract

Overhead cranes play a crucial role in industrial operations, yet oscillatory motions during handling tasks may threaten both safety and efficiency. This study proposes a simple but systematic approach for conducting an initial assessment of crane dynamic behavior by jointly analyzing rope angle measurements and load data transmitted via a controller area network (CAN) bus system. The dataset, sampled every five seconds, was cleaned through the removal of missing entries and outliers, followed by descriptive analysis, correlation evaluations, histogram and boxplot inspections, and time-series exploration. The results indicate that the X and Y rope angles exhibit low variability with relatively uniform distributions, while the load data are right-skewed, dominated by low values but occasionally reaching very high magnitudes. A moderate positive relationship was observed between the two angle axes, whereas their associations with load remained weak. Time-series trends and moving-average assessments showed that sudden increases in load can trigger noticeable angular deviations, making the crane more sensitive under heavier operating conditions. Overall, the system demonstrates stable behavior under normal loads, yet risks become more significant as load levels rise. The study aims to outline dominant oscillation components in the frequency domain, establish a reproducible and straightforward data processing workflow, and provide preliminary indicators that may support future anti-sway optimization and predictive maintenance strategies.