System Integration and Participant-Level Validation of a Gate-Based Contactless Multimodal System for Railway Pre-Duty Screening

Railway pre-duty assessments require rapid, objective measurements that do not disrupt operational workflows. This study evaluated the system-level integration and participant-level measurement performance of a gate-based Through-Pass platform for railway pre-duty screening. The platform combines radio-frequency identification, infrared temperature sensing, camera-based remote photoplethysmography, breath-alcohol sensing, signal-quality control, edge processing, and rule-based screening. Participant-level method comparison was performed in 300 adults using paired Through-Pass and reference measurements of body temperature, heart rate, systolic and diastolic blood pressure, and blood alcohol concentration. Agreement was assessed using Bland–Altman analysis, concordance statistics, error metrics, Deming regression, and multivariable models of measurement error. Mean biases were close to zero for all five outcomes, and Lin’s concordance correlation coefficients ranged from 0.938 to 0.995. After multiple-testing correction, measurement error showed no consistent associations with age, sex, body mass index, grouped Fitzpatrick skin type, or eyeglass use, whereas the reference-value magnitude was associated with error for diastolic blood pressure and blood alcohol concentration. These findings support the feasibility of integrating established sensing components into a unified pre-duty measurement platform under controlled indoor conditions; field and screening-classification validation remain necessary before operational implementation.