Highlights Measuring crowding through archived and real-time data using object detection tools. Object detection tools with surveillance video to quantify transit performance. Logistic regression models with automatic passenger counts and train operations data. Estimation of left-behind passengers at stations with farecard use only at entrance. Estimation and evaluation of waiting time as a transit service reliability measure.

Thirty-four drivers participated in a driving simulator experiment which investigated time and speed perception as it related to cognitive workload resulting from secondary tasks. Each participant drove the virtual drive twice, once with either an audio or a map task and again with no distractions as a control. Participants knew from a practice drive that they would be asked to estimate their speed and time duration of driving, thus this study used the prospective paradigm. Based on previous literature it was expected that there would be an underestimation of time and an overestimation of speed. The reverse occurred, participants overestimated time and underestimate their speed. This suggests that drivers may have found the drive unstimulating, despite the secondary tasks, and that the rural environmental may have impacted speed perception. In addition, a large group of participants, nine out of 34, crashed the virtual vehicle at a horizontal curve that was not problematic in previous simulator studies. When investigating these crashes further, it was found that drivers who crashed in the second drive had significantly worse time perception in the first drive as compared to drivers who did not crash in the second drive. This finding suggests that current time perception may be a predictor of future speed selection.