Time-synchronized whole-body vibration and vehicle speed data from 11 urban bus routes in Haikou, China
Description
Research hypothesis We hypothesized that driving behavior—particularly stop–start frequency—would be a major determinant of whole-body vibration (WBV) exposure among urban bus drivers, and that vehicle speed alone would explain only a limited portion of the variation in vibration dose. We further hypothesized that a wide “safe speed zone” would exist rather than a strict optimal speed. Data overview This dataset contains time-synchronized WBV and vehicle speed data from 11 urban bus routes in Haikou, China, collected between 7 December 2021 and 13 January 2022. WBV was measured at the driver–seat interface using a six-channel human vibration analyzer (SV-106A, SVANTEK, Poland) with a tri-axial seat-pad accelerometer (SV-38A), following ISO 2631-1. Tri-axial acceleration was sampled with Wd frequency weighting for the X- and Y-axes and Wk weighting for the Z-axis. Data were logged at approx 8 s intervals. Vehicle speed was recorded independently by the on-board GPS terminal at 5 s intervals. Raw geographic coordinates were excluded to comply with national geographic information security regulations; only speed and time are provided.The vibration and speed records were collected asynchronously. They were processed in MATLAB R2026a: timestamps were unified, the overlapping period was extracted, speed was linearly interpolated to the vibration time base, and duplicate or missing values were removed. After processing, 9,548 time-aligned paired observations were obtained across the 11 routes. Each route file contains time, speed, and tri-axial frequency-weighted acceleration (X, Y, Z). A summary file provides route-level metrics, including the overall z-axis weighted RMS (W_KZ) and the frequency concentration index (F_Z). Notable findings Multiple linear regression(MLR) showed that stops per kilometer was the strongest positive predictor of vertical vibration dose (W_KZ). The model explained 85.3% of the variance in W_KZ. In contrast, vehicle speed alone explained only 12.7% of the variance in the shock deviation (SD)index, and the fitted quadratic curve indicated a wide safe speed zone (approx 20–65 km/h) rather than a strict optimal speed. A SD index was developed to identify high-impact vibration periods during driving. Notably, stops per kilometer was negatively correlated with W_KZ in univariate analysis but positively associated after controlling for speed and other variables, reflecting a suppression effect: congested routes with frequent stops tend to have lower speeds and lower overall vibration magnitudes, but each stop–start cycle adds transient shocks that contribute to vibration dose once speed is controlled for. How the data can be interpreted and used The dataset can be used to reproduce the analyses in the associated manuscript, including the shock deviation index, MLR of driving behavior features, and the speed–vibration relationship. It can also support further research on WBV exposure, driving behavior, and occupational ergonomics.
Files
Steps to reproduce
This dataset contains whole-body vibration (WBV) and vehicle speed data collected from 11 urban bus routes operated by Haikou Public Transport Group in Haikou, Hainan Province, China. Data collection was conducted between 7 December 2021 and 13 January 2022. The dataset was generated to support research on the contribution of driving behavior to occupational WBV exposure among urban bus drivers. Instrumentation and measurement protocol WBV was measured at the driver–seat interface using a six-channel human vibration analyzer (SV-106A, SVANTEK, Poland) with a tri-axial seat-pad accelerometer (SV-38A). The accelerometer was securely mounted on the driver's seat pan at the ischial tuberosity position. Tri-axial acceleration was measured along the fore-aft (X), lateral (Y), and vertical (Z) axes. The sampling rate was 6000 Hz. Frequency weightings were applied according to ISO 2631-1: Wd for the X- and Y-axes and Wk for the Z-axis. The RMS detector time constant was set to 1.0 s. Data were logged at approximately 8 s intervals. Each measurement covered one complete round trip (outbound and inbound), and the minimum recording duration exceeded 30 min per trip. Vehicle speed data were collected independently by the on-board GPS terminal installed on each bus. Speed and timestamp were recorded at 5 s intervals. Raw geographic coordinates (latitude and longitude) were not included in this dataset to comply with national geographic information security regulations; only vehicle speed and time are provided. If users require geographic information, they should obtain it through appropriate legal and ethical channels. Data processing and time alignment The vibration and speed records were collected asynchronously and had different sampling intervals. All processing was performed in MATLAB R2026a . The following steps were applied: Timestamps were converted to MATLAB datetime format. The overlapping time period between the vibration and speed datasets was extracted using synchronize(..., 'intersection'). Because the vibration data had a coarser time resolution (approximately 8 s) than the GPS speed data (5 s), the speed signal was linearly interpolated to the vibration time base using retime and interp1. Duplicate timestamps were removed, and rows containing missing values were deleted using rmmissing. After processing, 9,548 time-aligned paired observations were obtained across the 11 routes. Data structure The dataset is organized by route. Each CSV file contains time-synchronized variables, including: Time: timestamp of the vibration record Speed_kmh: vehicle speed (km/h) X,Y,Z: X-axis,Y-axis, Z-axis frequency-weighted acceleration (m/s²) A summary file is also provided with route-level metrics, including the overall z-axis weighted RMS (WK_Z) and the frequency concentration index (F_Z). The MATLAB scripts used for time alignment and feature extraction are provided as supplementary files.
Institutions
- Hainan Center for Disease Control & PreventionHainan, Haikou
- Chinese Center For Disease Control and PreventionBeijing, Beijing
- Guangdong University of TechnologyGuangdong, Guangzhou
Categories
Funders
- Chinese Center For Disease Control and PreventionMinistry of Agriculture and Rural AffairsBeijing, BeijingGrant ID: 131031109000160004