Use of a multihazard sensor system for understanding rider experience and environmental conditions on urban rail

The range of potential environmental hazards to which public transit riders and operators are exposed include vibration, noise, and air pollution. Understanding these conditions—and their interactions—at a granular level across geographically expansive networks is critical to assessing, designing, and evaluating options for infrastructure improvements. However, the majority of studies to date have focused on a single or few parameters at varied spatial and temporal scales. This work presents a cost-effective and semi-automated approach to simultaneously measuring vibration, noise, and air pollution (particulate matter) at a resolution sufficient to resolve differences in rider experience on a segment-by-segment basis ( ⩾ 1 sample·min −1 ). The system was validated through a study of the Massachusetts Bay Transportation Authority urban rapid transit system (Boston, MA). Vibration, noise, and particulate matter hotspots were identified. Vibration data grouped by line, perhaps pointing to differences in train or track characteristics within the transit network, while noise hotspots were not predicted by studied system characteristics and therefore are likely driven by track-specific issues (e.g. aging, curves). Air pollution levels were systematically higher (1) during peak ridership hours and (2) when trains traveled through underground sections of the system, with an abrupt change at the segment where trains transitioned between below and above ground. Correlations are observed between hazards, especially between air pollution and noise ( R = 0.54), implying that targeted interventions (e.g. related to braking or rail systems) could be especially effective in improving the experience of the rider, especially in underground sections where there is less ventilation and more reverberation. This study presents the utility of a multihazard approach for improving understanding of the rider experience in public transit systems and as a method for informing transition system management, suggesting an important next step of assessment by infrastructure management experts to develop processes for converting data-driven insights into candidate system improvements.